close

Вход

Забыли?

вход по аккаунту

?

DEVELOPMENT OF VISIBLE/INFRARED/MICROWAVE AGRICULTURAL CLASSIFICATION AND BIOMASS ALGORITHMS

код для вставкиСкачать
INFORMATION TO USERS
This was produced from a copy of a document sent to us for microfilming. While the
most advanced technological means to photograph and reproduce this document
have been used, the quality is heavily dependent upon the quality o f the material
submitted.
The follow ing explanation o f techniques is provided to help you understand
markings or notations which may appear on this reproduction.
1. The sign or “target” fo r pages apparently lacking from the document
photographed is “ Missing Page(s)” . I f it was possible to obtain the missing
page(s) or section, they are spliced in to the film along w ith adjacent pages.
This
may
have
necessitated cutting through an image and duplicating
adjacent pages to assure you o f complete continuity.
2. When an image on the film is obliterated w ith a round black mark it is an
indication that the film inspector noticed either blurred copy because o f
movem ent during exposure, or duplicate copy. Unless we meant to delete
copyrighted materials that should not have been film ed, you wili find a good
image of the page in the adjacent fram e. If copyrighted materials were
deleted you will find a target note listing the pages in the adjacent frame.
3. When a map, drawing or chart, etc., is part o f the material being photo­
graphed the photographer has follow ed a definite method in “sectioning”
the material. I t is customary to begin film in g at the upper le ft hand corner of
a large sheet and to continue from le ft to right in equal sections with small
overlaps. If necessary, sectioning is continued again—beginning below the
first row and continuing on until complete.
4. For any illustrations th a t cannot be reproduced satisfactorily by xerography,
photographic prints can be purchased at additional cost and tipped into your
xerographic copy. Requests can be made to our Dissertations Customer
Services Departm ent.
5. Some pages in any document may have indistinct print. In all cases we have
film ed the best available copy.
Universe
Microfilms
International
300 N. 2E E B RD.. A N N A R BO R , M l 48106
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
8206660
Rosenthal, Wesley Dean
DEVELOPMENT OF VISIBLE/INFRARED/MICROWAVE AGRICULTURAL
CLASSIFICATION AND BIOMASS ALGORITHMS
Ph.D.
Texas A & M University
University
Microfilms
International
1981
300 N. Zeeb Road, Ann Arbor, M I 48106
Copyright 1981
by
Rosenthal, Wesley Dean
All Rights Reserved
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
PLEASE NOTE:
In all cases this material has been filmed in the best possible way from the available copy.
Problems encountered with this document have been identified here with a check mark V
^
1.
Glossy photographs or pages
2.
Colored illustrations, paper or print
3.
Photographs with dark background
4.
Illustrations are poor copy______
5.
Pages with black marks, not original
6.
Print shows through as there is text on both sides of p ag e______
7.
Indistinct, broken or small print on several pages
8.
.
copy_
Print exceeds margin requirements______
9.
Tightly bound copy with print lost in spine______
10.
Computer printout pages with indistinct print
11.
Page(s)____________ lacking when material received, and not available from school or
author.
12.
Page(s)____________ seem to be missing in numbering only as text follows.
13.
Two pages num bered____________ . Text follows.
14.
Curling and wrinkled pages______
15.
Other__________________________________________________________________________
\/
University
Microfilms
International
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
R eproduced w ith perm ission o f the copyright owner. Further reproduction prohibited w itho ut perm ission.
DEVELOPMENT OF VISIBLE/INFRARED/MICROWAVE AGRICULTURAL
CLASSIFICATION AND BIOMASS ALGORITHMS
A Dissertation
by
WESLEY DEAN ROSENTHAL
Submitted to the Graduate College of
Texas A&M University
in partial f u lf illm e n t of the requirement for the degree of
DOCTOR OF PHILOSOPHY
December 1981
Major Subject:
Agricultural Engineering
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
DEVELOPMENT OF VISIBLE/INFRARED/MICROWAVE AGRICULTURAL
CLASSIFICATION AND BIOMASS ALGORITHMS
A Dissertation
by
WESLEY DEAN ROSENTHAL
Approved as to s ty le and content by:
_ _ _ _ _ _ _
'Co-Chairman of Committee)
y i,.... ... %L
,
(Co-Chairman of Committee)
y i4
lembfer.
(Member)
Head o f Department)
December 1981
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ABSTRACT
Development of Visible/Infrared/Microwave Agricultural
Classification and Biomass Algorithms (December 1981)
Wesley Dean Rosenthal, B, S. University of Nebraska
M. S ., Kansas State University
Co-Chairman of Advisory Committee:
Dr. Bruce Blanchard
Dr. John Nieber
Due to inadequate crop acreage and biomass estimates using sate l­
l i t e and a ir c r a f t v is ib le and infrared data, a study was conducted to
(1) develop and te s t agricultural crop c la s s ific a tio n models using two
or more spectral regions (v is ib le through microwave), and (2) estimate
biomass by including microwave with visible
study was conducted at two locations;
Dalhart, Texas in 1980.
and infrared data.
The
Guymon, Oklahoma in 1978, and
A irc ra ft multispectral data collected during
the study included visib le and infrared data (multiband data from 0.5
urn
- 12 pm), passive microwave data [C band (6 cm) v e rtic a l and hori­
zontal p o larizations, and L band
(20 cm) horizontal p o larizatio n ] and
active microwave data [K band (2 cm), C band (6 cm), L band (20 cm),
and P band (75 cm) lik e and cross p o larizatio n s].
from each f i e l d
at Dalhart.
consisted of soil
Ground truth data
moisture at both sites and biomass
The study was divided into four problems: (1) are d i f f e r ­
ences in individual band responses related to crop type differences?
(2) what is the most accurate multi frequency crop classifying dendro­
gram (tree c la s s if ie r )
at both locations?
(3) what is the u t i l i t y of
microwave data alone or in combination with other spectral bands for
classifying
crops
and
estimating
total
biomass?
and
(4)
is
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
the
■i v
multi frequency
t r e e -c la s s ific a tio n
model
phenological or biomass differences?
v a r i a b i li t y
dependent
on
Results indicated that inclusion
of C, L, and P band active microwave data from look angles greater
than 35° from nadir with v is ib le and infrared data improved crop dis­
crimination and biomass estimates compared to results using only v i s i ­
ble and infrared data.
The active microwave frequencies were sensi­
tiv e to d iffe re n t biomass levels.
K and C band were sensitive to d i f ­
ferences at low biomass lev e ls , while P band was sensitive to d i f f e r ­
ences at high biomass lev e ls .
In addition, two indices,
one using
only active microwave data and the other using data from the middle
and
near
infrared
bands,
were
well
correlated
to
total
biomass.
Results from the study implied that inclusion of active microwave sen­
sors with v is ib le and infrared sensors on future s a t e llit e s could aid
in crop discrimination and biomass estimation.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
V
ACKNOWLEDGEMENTS
The author would lik e to acknowledge Dr. Bruce Blanchard for his
valuable help and guidance throughout the study period.
Also, I would
lik e to acknowledge Dr. John Nieber, Dr. Joe McFarland, and Dr. C l i f f
Harlan for t h e ir help and suggestions in guiding me through my course
of study.
I would lik e to especially thank Ms. Cheryl Jones, for preparing
many of the fig u re s, and Ms. Linda Kocman for typing the te x t .
ful appreciation goes out
s t it u t e
of
the
University
Grate­
to the personnel at the Remote Sensing
of
California
at Santa
Remote Sensing Center at Texas ASM University
Barbara
Spectral
crew aboard the C-130 is
appreciated. Gratitude
also
and the
for collecting
samples at Guymon and Dalhart.
In­
ground
data collected by the NASA
is
also ex­
pressed to the commercialfarmers in the Guymon and Dalhart
area for
granting permission to take samples in th e ir fie ld s during the experi­
ment.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
vi
TABLE OF CONTENTS
C hapter
Page
I
INTRODUCTION....................................................................................................1
Objectives and Research................................................................ 7
II
REVIEW OF LITERATURE.................................................................................10
Spectral Theory...............................................................................10
C la s s ific a tio n Models.................................................................. 23
Biomass Models ............................................................................ 25
Final L ite ra tu re Overview.......................................................... 27
III
DATA COLLECTION
....................................................................30
Guymon A ir c r a f t and Ground Data..............................................30
Dalhart A ir c r a ft and Ground Data ...................................... 42
Scatterometer Processing ....................................................... 53
NS001/M2S Processing ............................................................... 55
Passive Mic.'owave Processing..................................................56
IV
ANALYSIS...........................................................
57
Techniques....................................................................................... 57
V
RESULTS......................................................................................................... 60
Guymon Crop Condition.................................................................. 60
Dalhart Biomass and Crop Y i e l d ............................................ 60
Problem 1 . ......................................................................................62
Problem 2.......................................................................................... 89
Problem 3.......................................................................................... 95
Problem 4........................................................................................ 134
VI
SUMMARY AND CONCLUSIONS......................................................................149
Problem 1.....................
149
Problem 2......................................................................................... 150
Problem 3...................................... . ...............................................150
Problem 4 , ................................... ' ................................................ 152
O v e rv ie w ....................
153
REFERENCES.......................................... ............................................. ........................ 156
APPENDIX A
DATA QUALITY, CALIBRATION AND OMISSIONS .......................... 161
APPENDIX B DALHART DATA SET.............................................................................180
APPENDIX C GUYMON DATA S E T .............................................................................191
VITA.............................................. ...............................................................................214
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
v ii
LIST OF FIGURES
F ig u re
1
2
Page
Reflectance of 2 and 8 stacked mature cotton
leaves. Standard deviation between observed
and calculated points is about 1%. From Allen
et a l . , 1970..........................................................................................
17
Averaged normalized differences (IR-red/IR+red)
values plotted against soybean wet biomass. From
Tucker et a l . , 1979 ..........................................................................
18
3
Diagram i llu s t r a t in g the principle of the perpen­
dicular vegetation index (PVI) model. A perpendicuar from candidate plant coordinates (Rp5, Rp7)
intersects the soil background lin e at coordinates
(Rg5, Rg7). A PVI=0 indicates s o i l , and a PVI>0
indicates vegetation. From Richardson and Wiegand,
1977........................................................................................................... 28
4a
Legend fo r the Guymon, Oklahoma fie ld s m aps.............................31
4b
Area map of Guymon showing the r e la tiv e
locations of each f ie l d m ap.............................................................32
4c
Locations
of the sample fie ld s at
4d
Locations
of the sample fie ld s at Guymon.................................... 34
4e
Locations
of the sample fie ld s at Guymon.................................... 36
4f
Legend fo r the Clayton, New Mexico f i e l d maps.........................37
4g
Area map of Clayton showing the re la tiv e
location of the f ie l d m ap.................................................................38
4h
Location of the sample fie ld s at Clayton..................................... 39
5
Sampling pattern for field s at Guymon and
Dalhart. Points 1, 2, 7 and 8 were moved
outside the c ir c le for rectangular field s ............................. 43
6
Soil moisture sampling depths at Dalhart
and Guymon. The 15-30 and 30-45 cm core
samples were also taken in addition to the
above. Samples were collected from 5-9 cm
and 9-15 cm at Guymon and 5-15 cm at Dalhart............................44
Guymon.....................................33
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
vi i i
F ig u re
Page
7a
Legend fo r the Dalhart, Texas f i e l d maps......................................45
7b
Area map of Dalhart showing the r e la tiv e
locations of each f i e l d m ap ............................................................. 46
7c
Locations of the sample fie ld s at D a l h a r t ................................. 47
7d
Locations of the sample fie ld s at D a l h a r t ................................. 48
7e
Locations of the sample f ie ld s at D a l h a r t ................................. 49
8
Scatterometer data processing procedure ................................... 54
9
Spectra fo r m ille t and corn fie ld s at Dalhart.
[H = C band horizontal (MFMR), V = C band v e rtic a l
pole (MFMR), L = L band horizontal (MFMR), H = lik e
pole 40° look angle (SCATTS), V = cross pole 40°
look angle (SCATTS), A = 0-2 cm soil moisture (SM),
B = 2-5 cm soil moisture (SM)]......................................................... 64
10
Spectra fo r bare s o i l , pasture and wheat stubble at
Dalhart. [H = C band horizontal (MFMR), V = C band
v e rtic a l pole (MFMR), L = L band horizontal (MFMR),
H = l i k e pole 40° look angle (SCATTS), V = cross
pole 40° look angle (SCATTS), A = 0-2 cm soil
moisture (SM), B = 2-5 cm soil moisture ( S M ) ] ........................ 66
11
Spectra comparing vegetated and non-vegetated fie ld s
at Dalhart. [H = C band horizontal (MFMR), V = C band
v e rtic a l pole (MFMR), L = L band horizontal (MFMR),
H = l i k e pole 40° look angle (SCATTS), V = cross pole
40°look angle (SCATTS), A = 0-2 cm soil moisture (SM),
B = 2-5 cm soil moisture (SM) ] ......................................................... 68
12
Black and white infrared a erial ph'oto (scale 1:45,000)
of stressed corn fie ld s ( fie ld s 1 and 2) at Dalhart.
The healthy areas are dark shaded and the stressed
areas are lig h t shaded.......................................................................... 71
13
Spectra comparing healthy and stressed corn at
Dalhart. No microwave comparisons could be made................. 72
14
Spectra comparing a l f a l f a , sorghum, and bare soil
f ie ld s at Guymon. [H = C band horizontal (MFMR), V
= C band v e rtic a l pole (MFMR), L = L band horizontal
(MFMR), H = l i k e pole 40° look angle (SCATTS), V =
cross pole 40°look angle (SCATTS), A = 0-2 cm soil
moisture (SM), B = 2-5 cm soil moisture (SM) ] ........................ 74
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ix
F ig u re
Page
15
Spectra comparing sorghum f ie ld s with rows perpendic­
u la r and p a ra lle l to the f l i g h t lin e . [H = C band hori­
zontal (MFMR), V = C band v e rtic a l pole (MFMR), L =
L band horizontal (MFMR), H = lik e pole 40° look angle
(SCATTS), V = cross pole 40°look angle (SCATTS), A =
0-2 cm soil moisture (SM), B = 2-5 cm soil moisture
(SM) ] ...........................................................................................................77
16
Spectra comparing wet bare s o i l , and a dry sorghum
f i e l d at Guymon. [H = C band horizontal (MFMR), V =
C band v e r tic a l pole (MFMR), L = L band horizontal
(MFMR), H = l ik e pole 40° look angle (SCATTS), V =
cross pole 40°look angle (SCATTS), A = 0-2 cm soil
moisture (SM), B = 2-5 cm soil moisture (SM) ] ....................... 79
17
Spectra comparing corn and sorghum at Clayton.
No
passive microwave or v is ib le /in fr a r e d data was a v a il ­
able.
[H = l ik e pole 40° look angle (SCATTS), V =
cross pole 40° look angle (SCATTS)]........................................... 80
18
Spectra comparing corn and sorghum at Clayton.
No
passive microwave or v is i b le /i n f r a r e d data was a v a il­
able.
[H = l i k e pole 40° look angle (SCATTS), V =
cross pole 40° look angle (SCATTS)]........................................... 81
19
Line plots (o° vs time) for a l l lik e polarized scatterometer data at 10° and 40° o ff n a d i r ....................................87
20
Line plots (o° vs time) for a ll cross polarized scatterometer data at 10° and 40° o ff nadir ................. . . . .
88
21
Dendrogram ( tr e e - c la s s if ic a tio n ) model using
NS001 bands 2, 3, and 4 , C and L band cross pole
Dalhart data (accuracy 7 8 % ) .............................................................91
22
Dendrogram ( t r e e - c la s s if ic a t io n ) model using
NS001 bands C and L band cross pole Dalhart data
(accuracy 84%)
93
23
Dendrogram ( t r e e - c la s s if ic a tio n ) modeling using
M2S bands 4, 7, 8 and 9 , C and L band cross pole
Guymon data (accuracy 70%)................................................................. 94
24
Dendrogram ( t r e e - c la s s if ic a tio n ) model using
a l l NS001 bands Dalhart (accuracy 78%)........................................ 96
25
Dendrogram ( t r e e - c la s s if ic a tio n ) model using
M2S bands 4, 7, 8 and 9 data at Guymon (65% accuracy) . . 97
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
X
F ig u re
Page
26
The relationship between total biomass (g/m2) ,
and TVI and PVIat Dalhart................................................................ 105
27
The relationship between fin a l crop y ie ld (Kg/Ha),
and TVI and PVIat Dalhart................................................................106
28
Field radiance reflectance values of NS001 bands
1 and 2 versus band 3 at Dalhart in 10-4 watts
cm- 2 ster- 1 .......................................................................................... 108
29
Field radiance reflectance values of NS001 bands
4 and 5 versus baud 3 at Dalhart in 10-4 watts
cm- 2 s te r- 1 ...........................................................................................109
30
Field radiance reflectance values of NS001 bands
6 and 7 versus band 3 at Dalhart in 10-4 watts
cm- 2 ster- 1 ...........................................................................................110
31
Field radiance reflectance values of NS001 bands
1 and 2 versus band 4 at Dalhart in 10-4 watts
cm- 2 ster...........................................................................................I l l
32
Field radiance reflectance values of NS001 bands
3 and 5 versus band 4 at Dalhart in 10-4 watts
cm- 2 ster- 1 ...........................................................................................112
33
Field radiance reflectance values of NS001 bands
6 and 7 versus band 4 at Dalhart in 10-4 watts
cm-2 s ter- 1 ...........................................................................................113
34
Field radiance reflectance values of NS001 bands
1 and 2 versus band 5 at Dalhart in 10-4 watts
cm-2 s ter- 1 ........................................................................................... 114
35
Field radiance reflectance values of NS001 bands
3 and 4 versus band 5 at Dalhart in 10-4 watts
cm-2 s te r- 1 .......................................................................................... 115
36
Field radiance reflectance values of NS001 bands
6 and 7 versus band 5 at Dalhart in 10-4 watts
cm- 2 ster- 1 ........................................................................................... 116
37
The relationshp between to ta l (wet) biomass (g/m2 )
and PVI64 at Dalhart............................................................................. 117
38
The relationshp between PVI64, and PVI and TVI
at D alhart............................................................................................... 118
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
xi
F ig u re
Page
39
A color photo indicating d iff e r e n t PVI64 levels
within a stressed corn f i e l d (1 and 2) at Dalhart . . . .119
40
A color photo indicating d iff e r e n t PVI64 levels
within a sorghum f i e l d (V2) at D alhart...................................... 120
41
A color photo indicating d iff e r e n t PVI64 levels
within a l f a l f a f ie ld s ( V l l , V12, V13) at D alhart..................121
42
The relationship between L band cross pole o°
and look angle fo r a corn f i e l d ( f i e l d 9) and
bare f i e l d ( f i e l d 1 5 ) ........................................................................123
43
The relationship between L band cross pole a0
and look angle for a m ille t f i e l d ( f i e l d 3) under
d iff e r e n t soil moisture conditions...............................................124
44
The L band cross pole a0 response as a function of
look angle for the same sorghum f i e l d ( f i e l d IX)
from two d iffe r e n t d irec tio n s , the f l i g h t lin e paral­
le l and perpendicular to the t i l l a g e d ir e c tio n ...................... 126
45
The relationship between to ta l biomass and the
scatterometer vegetation index, SVI.
(4.75 HV 40°
look angle - 4.75 HV 5° look angle) (R2 = 0.88) .................. 128
46
The relationship between SVI (d b ), and TVI and PVI
at Dalhart................................................................................................ 129
47
The relationship between SVI (db), and TVI and PVI
at Guymon.....................................................
130
48
The relationship between SVI (db), and 0-2 cm so il
moisture (%) for selected f ie ld s at Guymon and
D a l h a r t ....................................................' . ............................................ 131
49
The relationship between soil moisture corrected
SVI (db) and TVI and PVI at D a l h a r t ...........................................132
50
The relationship between the soil moisture cor­
rected SVI (db), and TVI and PVI at Guymon.............................. 133
51
The red/near-in frared relationship for fie ld s at
Guymon and Dalhart................................................................................ 136
52
The K band l ik e pole a0 response as a function of
look angle for bare soil ( f i e l d 2 2), sorghum ( f i e l d
V2 and V6), and corn ( f i e l d 2) at D a l h a r t .............................. 137
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
xi i
F ig u re
Page
53
The C band cross pole o° response as a function of
look angle fo r bare soil ( f i e l d 2 2 ), sorghum ( f i e l d
V2 and V6), and corn ( f i e l d 2) at D a l h a r t ..............................138
54
The L band cross pole o° response as a function of
look angle fo r bare soil ( f i e l d 22), sorghum ( f i e l d
V2 and V6), and corn ( f i e l d 2) at D a l h a r t ................................ 140
55
The P band cross pole a0 response as a function of
look angle fo r bare soil ( f i e l d 22), sorghum ( f i e l d
V2 and V6), and corn ( f i e l d 2) at D a l h a r t ................................141
56
The K band lik e pole a0 response as a function of
look angle fo r bare soil ( f i e l d 14), a l f a l f a ( f i e l d
4 ) , emerging sorghum ( f i e l d 15) and headed sorghum
( f i e l d I X ) ......................................
' . ..................... 142
57
The C band cross pole a0 response as a function of
look angle fo r bare soil ( f i e l d 14), a l f a l f a ( f i e l d
4 ) , emerging sorghum ( f i e l d 15) and headed sorghum
( f i e l d I X ) .................................................................................................143
58
The L band cross pole a0 response as a function of
look angle fo r bare soil ( f i e l d 14), a l f a l f a ( f i e l d
4 ) , emerging sorghum ( f i e l d 15) and headed sorghum
( f i e l d I X ) ......................................................................." . ..................... 144
59
The P band cross pole a0 response as a function of
look angle fo r bare soil ( f i e l d 14), a l f a l f a ( f i e l d
4 ) , emerging sorghum ( f i e l d 15) and headed sorghum
( f i e l d I X ) .....................................
145
60
The relationship between to ta l biomass at Dalhart
and the modified scatterometer vegetation index,
SVIM [(C band cross pole 40° - C b’and cross pole
5°) t (P band cross pole 40° - P band cross pole 5 ° ) ] . .147
61
The relationship between the modified SVI (SIVM)
and TVI and PVI at Guymon............................................................... 148
Al
Field IX (sorghum) P band lik e and cross pole
response with rows perpendicular to the f l i g h t lin e . . .168
A2a
Scatterometer response from the P band l i k e pole
system over f i e l d 25 (sorghum) with rows perpen­
dic u la r to the f l i g h t l i n e ............................................................... 169
A2b
Scatterometer response from the P band cross pole
system over f i e l d 25 (sorghum) with rows perpen­
d ic u la r to the f l i g h t l i n e .....................................................
. .170
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
x ii i
F ig u re
Page
A3
Scatterometer response (C and L l ik e and cross pole)
from f i e l d 25 at Dalhart on 8 / 1 6 / 8 0 ............................................172
A4
Scatterometer response (K band lik e pole) from f ie l d
19 at Dalhart on 8/16/80 and f ie l d 14 at Guymon on
8 /5 /7 8 . Soil moisture conditions were approximately
90% of f ie l d c a p a c i t y ......................................................................173
A5
Scatterometer response (C band lik e and cross pole)
from f ie l d 19 at Dalhart on 8/16/80 and f ie l d 14 at
Guymon on 8/5/78. Soil moisture conditions were
approximately 90% of f i e l d capacity ............................................174
A6
Scatterometer response (L band l ik e and cross pole)
from f i e l d 19 at Dalhart on 8/16/80 and f i e l d 14 at
Guymon on 8/5/78. Soil moisture conditions were
approximately 90% of f ie l d c a p a c i t y ...........................................175
A7
Scatterometer response (P band lik e and cross pole)
from f i e l d 19 at Dalhart on 8/16/80 and f ie l d 14 at
Guymon on 8 /5 /7 8 . Soil moisture conditions were
approximately 90% of f i e l d c a p a c i t y ...........................................176
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
xi v
L IS T Oi- TABLES
T a b le
Paqe
1
Operating sensors for the Guymon, Oklahoma study..................... 41
2
Operating sensors for the D alhart, Texas study..........................52
3
Dalhart biomass and crop y i e l d . . . .
4
Results of Duncan's M ultiple Range Test fo r
Dalhart active microwave data ....................................................... 83
5
Results of Duncan's Multiple Range Test fo r
Guymon active microwave data.............................................................. 85
6
Dalhart discriminant analysis results using (a)
a l l NS001 channels and (b) a l l NSOOl channels
plus K band l i k e pole and L band cross pole (40°
look angle) data from August 14 and 18 as a t r a i n ­
ing c l a s s i f i e r . The results are from August 16
testin g of the model....................
.......................................
61
98
7
Dalhart discriminant analysis using (a) NSOOl
channels 2, 3 and 4 and (b) NSOOl channels 2, 3, and
4 and K band l ik e pole and L band cross pole data.
Contingency ta b le results from the model tested on
August 16 spectral d a t a ...................................................................... 99
8
Discriminant analysis of Guymon v is i b le /i n f r a r e d
data using August 2 and 17 data as the tra in in g
c la s s ifie r.
Results from c la s s if ic a tio n of August
5, 8, 11, and 14 d a t a ........................................................................ 101
9
Dalhart stepwise c la s s ific a tio n regression equations
using (a) a l l NSOOl band (Ch) data'and (b) a l l NSOOl
data plus scatterometer data (40° look angle) [Crop
Type: 10 = corn, 8 = sorghum, 6 = weeds, 4 = bare s o il
and weeds, 3 = pasture, 2 = wheat stubble, 1 = bare
soil ] .........................................
102
10
Guymon stepwise c la s s ific a tio n regression equations
using (a) only v is ib le /in f r a r e d data and (b) scat­
terometer (40° look angle) and v i s i b le /in f r a r e d data
[Crop Type: 8 =sorghum, 4 = a l f a l f a , 0 = bare s o i l ] . . . 103
Al
Equations used to convert raw NS001/M2S d i g it a l
counts (DC) to radiance values, R, (10- i *
watts cm-2
s te r- 1 ) fo r Guymon (a) and Dalhart ( b ) ......................................162
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
XV
Table
Page
A2
Questionable scatterometer data for Dalhart ...........................165
A3
Questionable scatterometer data for Guymon...............................166
A4
Guymon and Dalhart questionable MFMR data .
...................... 178
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1
CHAPTER I
INTRODUCTION
With world population increasing to a point where food supplies
w ill
become scarce, the need to improve global
tion systems becomes c r i t i c a l l y important.
agricultural
informa­
Such emphasis is needed to
avert potential world disasters of starvation and malnutrition due to
inadequate food supplies.
The delicate imbalance is demonstrated by
the fact that since 1948 the amount of exported grain from developed
countries
result,
to
developing
countries
has
risen
dramatically.
the less developed countries are more dependent
production
in a few developed countries
World Food and Nutrition Study (National
(Wortman,
As
a
on surplus
1976).
A recent
Acadeny of Sciences,
1977)
emphasized the need for improved systems by recommending high p r io r it y
research on
1
information needs of producers,
2
crop monitoring systems,
3
inte.national data bases for land and n u tr itio n , and
4
a total information system.
Perhaps the major p r io r it y
is
developing crop monitoring
sys­
tems. This world-wide need was emphasized when the United States lost
millions of dollars by se llin g wheat to the Soviet Union, who la te r
sold the wheat at much higher prices.
An adequate crop monitoring
system v/ould possibly have averted the deal.
The benefits of improved
The citations and style of this and the following pages follow
the style of the Transactions of the ASAE.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2
agricultu ra l
monitoring systems used for
predicting food production
would include
1
commodity prices would be more stable,
2
governments w ill be able to plan foreign policy, and
3
storage, transportation and processing f a c i l i t i e s w ill be
more e f f i c i e n t ly used.
The f i r s t benefit would prevent rapid and drastic seasonal
price
fluctuations
United
States
due to
government,
would be able to deal
needs.
deal
of
large
with
and
small
supplies.
an estimate
of
commodity
Second,
foreign
the
production,
accordingly to the foreign government's tru e
This would prevent events such as the U. S./Soviet Union wheat
1974.
Third, more e f f ic ie n t
use of transport
and storage
f a c i l i t i e s would help achieve the f i r s t two benefits.
The major problem o f monitoring production systems w ithin foreign
countries is
variables.
the inadequate source of data on acreages and climate
Several
countries
do not
presently
have any means for
estimating acreage or production within the country.
Other countries
have production monitoring systems which are highly inaccurate.
age and y ie ld estimates
Acre­
by the government are often inaccurate.
In
addition, several countries do not permit other countries to use the
production information.
Consequently, a universal technique is needed
soon.
One technique developed within the past twenty years uses remote­
ly
sensed data— sensors
production.
aboard s a t e llit e s
or
a i r c r a f t — to
estimate
From remotely-sensed data much information can be ob­
tained with a minimum of ground sampling (Bauer, 1975).
Such in fo r ­
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3
mation would d ra s tic a lly
reduce the cost of monitoring a g ricu ltu ra l
systems.
is
The technique
based prim arily
on the
relationship
of
reflectance in the v is ib le and infrared region of the electromagnetic
spectrum to vegetation type, cover, and crop condition.
ly ,
Id e a lis tic a l­
each healthy species has a c h a ra c te ris tic electromagnetic signa­
ture at a given growth stage.
cates
the
physiological
actual
Any departure from the signature i n d i ­
stress which could impact crop y ie ld .
spectrum
varies
to
an
extent
that
id e n t if ic a t io n is impossible using ava ila b le data.
crop
However,
and
stress
The v a r i a b i l i t y of
a crop spectrum due to stress is much larger than v a r i a b i l i t y due to
differences
between
s ig n ific a n tly
upon
the
crops.
vegetation
spectrum
from the non-vegetated spectrum.
difference
discriminated
The
to a
within
the
good degree
spectrum,
of
accuracy.
also
d iffe r s
Consequently,
crop
types
Also,
based
have
based
been
on
the
spectra, models have been developed which estimate biomass, le a f area
index, or percent cover (Richardson and Wiegand, 1977; Rouse et a l . ,
1973).
y ie ld
Biomass estimates can then be correlated to
(H olliday ,
1960a,
b;
Donald,
1963).
As a
fin a l
economic
r e s u lt ,
v is ib le /
infrared s a t e l l i t e and a i r c r a f t data have been used in (1) estimating
the percentage of area planted in a given crop, and
crop condition and biomass.
duction
estimate
fo r
the
(2)
evaluating
The combination of the two gives a pro­
area
(MacDonald,
1979).
Consequently,
through the use of s a t e l l i t e and a i r c r a f t data, a g ric u ltu ra l c l a s s i f i ­
cation and biomass estimation became important as a means of obtaining
reasonable estimates
addition,
ag ricu ltu ra l
of planted acreage and u ltim a te ly ,
y ie ld .
data can be collected by s a t e l li t e s
c r a f t from isolated areas of the world where a g ric u ltu ra l
In
and a i r ­
information
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4
had been d i f f i c u l t to obtain.
The major experiment during the 1970s which c la s s ifie d wheat and
estimated wheat acreage using only v is ib le and near infrared data from
Landsat
was the Large Area Crop
Donald,
1979).
U. S.
Inventory
Experiment
(LACIE)
LACIE was developed prim arily at the request
government to help monitor foreign production.
was to estimate
foreign wheat production
such as the Soviet Union and Argentina.
in
several
(Mac­
of the
The objective
key countries,
Success of the program would
prevent another U. S./S oviet Union grain trade incident.
Results were
well documented and the experiment was successful in some geographical
areas (Heydorn et a l . , 1979a; Potter et a l . , 1979).
From that experi­
ment and other studies, many crops were discriminated from bare soil
and water, but acreage estimates were s t i l l
inaccurate as a result of
sim ila r
spectral responses from other crops grown during the same time
o f year
(Heydorn et a l . ,
1979a).
To improve estimates, ground a n c il­
la ry data, such as crop growth stage or spectral
wavelength
regions,
are
needed.
With
the
data from d iff e r e n t
proposed
launch
of
the
Thematic Mapper, with fin e r spatial resolution and d iffe r e n t spectral
bands than Landsat, land-use and vegetation c la s s ific a tio n w i l l again
be the primary objective of further research (National Research Coun­
c il,
1976).
The Thematic Mapper w il l have spatial resolution of 30 m
x 30 m while Landsat has a resolution of 80 m x 80 m.
mapper w ill
have spectral
The Thematic
bands of (1) 0.45 to 0.52 ym, (2) 0.52 to
0.60 ym, (3) 0.63 to 0.69 ym, (4) 0.76 to 0.90 ym, (5)
1.00 to 1.30
ym; (5) 1.55 to 1.75 ym and (7) 2.08 to 2.35 ym.
Landsat has spectral
bands of (1) 0.50 to 0.60 ym (2) 0.60 to 0.70
(3) 0.70 to 0.80 and
(4) 0.80 to 1.1 ym.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
5
D ifferen t
supervised and unsupervised c la s s ific a tio n techniques
emerged from LACIE.
In the f i r s t method, "samples" of spectral data
were compared to a "training" sample of known land use.
I f the two
samples were s im ila r , the sample was clas s ifie d as the same land use
or vegetation cover that was present in the training area.
In this
technique, the analyst input the training information in a c l a s s i f ie r
algorithm (Bauer et a l . ,
1977).
In the unsupervised method, sim ilar
responses are grouped together into clusters and these clusters
then compared to
From this
actual
species clusters
technique a tre e -c la s s ific a tio n
based on spectral
are
(Cooley and Lohnes, 1971).
diagram can be developed
differences between the clusters.
Both techniques
are widely used in analyzing visible/near infrared spectral data with
supervised techniques being more widely used with s a t e l l i t e data.
The
major
problems
v is ib le /in fr a r e d
data
in
classifying
agricultural
has been the dependence for
crops
re lia b le
with
data on
clear weather and the v a r ia b ilit y of the c la s s ific a tio n estimate due
to
phenological
or
biomass differences.
B illin g sley
et
a l.
(1976)
proposed to eliminate these problems by including data from additional
bands, such as microwave data, which are independent of cloud cover.
Spectral
data
from many countries
excessive cloud cover.
are
predominantly
influenced
by
In many countries, agricultural Landsat data
was obtained only once during the growing season.
Consequently, more
frequent passes or additional bands were needed to improve s a t e l l i t e
coverage.
Also, with additional bands more accurate biomass estimates
may be possible.
During the LACIE experiment i t was also found that
climate data, primarily precipitation, was necessary before good e s t i ­
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6
mates of y ie ld could be obtained.
was used to estimate the soil
In the LACIE study, precipitation
moisture available to the crop.
The
microwave sensors have been recognized as a possible source of mois­
ture estimates.
In addition to this purpose they could also be used
to aid in discriminating crops.
Sensors can detect from two modes of ra d ia tio n --a c tiv e and pas­
sive.
Active
sensors re fe r to sensing reflected surface radiation
which originated from a known man-made energy source.
re fe r
to detection
tio n .
surface emitted and reflected
radia­
In this case, the surface is the source of rad iatio n .
Con­
siderable
effo rt
of natural
has
been made to
take
effects in active sensors while l i t t l e
e ffe cts
in
differences;
spatial
Passive sensors
passive
systems.
however,
resolution
to
passive
Both
advantage
of
polarization
has been done in polarization
have
microwave
be used e ffe c tiv e ly
s ig n ifican t
systems
in
Microwave data can be e ith e r active or passive.
have
crop
polarization
too
coarse
discrimination.
Active microwave res­
ponses are expressed as o °, the scattering c o e ffic ie n t, while passive
microwave responses are expressed as brightness temperature.
t ra s t
to the microwave data,
systems.
In con­
v is ib le studies are prim arily passive
Active v is ib le /in fr a re d data have been analyzed, but are too
complicated to be widely used.
Active microwave responses are prim arily dependent on two surface
c h a ra cteristics— surface roughness and soil
moisture.
Consequently,
crops having d iffe re n t roughnesses or morphologies would respond d i f ­
fe re n tly
in d iffe re n t
radar bands (Simonett et a l . ,
1967).
Higher
frequencies and the consequent shorter wavelength should be more sen­
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
7
s itiv e
than
vegetation.
lower
frequencies
to
the
roughness
chara cteristics
of
D ifferen t microwave frequencies should also have d i f f e r ­
ent c a p a b ilitie s of penetrating crop canopies and d iff e r e n t sen s itiv ­
ity
to soil
moisture.
Active microwave
responses
in the
8-18 GHz
range at high incidence angles of HH (h o rizo n ta lly polarized transmit
and received) and VV ( v e r t i c a ll y polarized transmit and received) have
been related to vegetative c h a ra cteristics (Ulaby et a l . , 1975).
High
emissivity in the passive microwave have also been related to vegeta­
t i v e biomass (S ib le y, 1973; Peake et a l . , 1966; Newton, 1977).
In
region,
spite
of
the
few studies
extensive
have related
research
in
the
active
microwave
combinations of v i s i b le ,
in fra re d ,
and microwave data to vegetation c h a ra cteristics (Brakke et a l . , 1981;
Ulaby et a l . ,
1981).
Consequently, i t
and biomass estimation
thermal
in fra re d ,
study
using
is f e l t that a c la s s ific a tio n
v i s i b le ,
near
in fr a r e d ,
far
or
and microwave data collected over an a g ric u ltu ra l
area may produce a multi frequency system that w ill
provide improved
estimates of crop acreage and crop conditions.
Objectives and Research
The purpose of this study was to (1) develop and test an ag ricu l­
tural c la s s ific a tio n model using two or more spectral regions ( v is ib le
through microwave),
with
vis ib le
and (2)
estimate biomass by including microwave
and infrared data.
The hypothesis was that
microwave
data can improve c la s s ific a tio n and biomass estimation accuracy over
present c la s s ific a tio n and estimation techniques that use v is ib le and
infrared data.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
8
The study was divided into four problems which were intended to
answer the previously mentioned goals.
The f i r s t two deal prim arily
with crop c la s s ific a tio n and the last two with biomass and crop clas­
s ific a tio n :
1
Are differences in individual
spectral
band responses related to
crop type differences and what is the relationship of each i n d iv i ­
dual multispectral band response to crop type?
2
What is the most accurate multi frequency dendrogram (tre e -c la s s i fic a tio n diagram) of agricultural crops in the Dal h a rt, Texas and
Guymon, Oklahoma areas?
3
What is the u t i l i t y of microwave data alone or in combination with
other spectral bands for classifying agricultural crops and e s t i ­
mating biomass?
4
Is the multifrequency crop tre e -c la s s ific a tio n model influenced by '
phenological
or biomass differences and can the model be adusted
to apply for a ll biophases?
Data used in this study were collected from the Guymon, Oklahoma
area in 1978 and the Dalhart, Texas area in 1980.
A irc ra ft
collected using the NASA C—130 a ir c r a f t with
f u ll
its
sensors and crew from the Johnson Space Center in
data were
complement of
Houston,
Texas.
Ground measurements were collected and processed with extensive sup­
port from graduate students and technical
A&M University
and the
University
personnel
of California at
from both Texas
Santa
Barbara.
Further discussion of the collection and processing of these
data w ill
be found in a following chapter.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
9
A valid
hypothesis implies that more accurate production e s t i ­
mates are possible by including microwave with
visib le and infrared
data.
dimension— vegetative
Microwave
data
roughness— to the analysis
could
add
of v is ib le
another
and infrared
highly correlated to the amount of biomass.
data
which
are
In addition, the inde­
pendence of microwave data to weather conditions allows analysis
of
many other areas of the world which were d i f f i c u l t to monitor using
visib le and infrared data.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
10
CHAPTER 11
REVIEW OF LITERATURE
C lass ific atio n and biomass models are based on spectral response
differences
regions.
between
and
within
crop
types
in
given
wavelength
Consequently, to better understand c la s s ific a tio n models, an
understanding of the spectral response at a ll wavelengths is required.
Spectral Theory
The reflec tio n of electromagnetic radiation from a given surface
as given by equations 1 and 2 is described by Uanza (1975):
- ( e 2c o s 0 1-) + /
Rv =
sin^e-
e2 -
---------------------- :--------- ; -;y—
(1)
(e2C0S6^) + / e2- s in z0 ^
and
(cose^) - / e2- sin ^9
Rh =
------------------------_ _ _ _ _
(cose^)
where Rv and R^ are
horizontal
the
+ /
e 2-
reflection
polarizations, respectively;
(2 )
sinze
coefficients
e2
for
v ertic al
and
is the d ie le c t r ic constant
of the re fle c tin g medium, and 9-j is the incidence angle of the plane
wave source.
Consequently, the d ie le c tric constat plays an important
role in determining reflectance at a ll
constant
varies
wavelengths.
The d ie le c t r ic
with wavelength, moisture content, and temperature.
For example, variations of the d ie le c tric with wavelength is
demon­
strated by water—the d ie le c tric at high microwave frequencies is 81,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
n
and
in
the
v is ib le ,
between wavelength
1.77
(Janza,
and roughness
1975).
affects
Also,
the
reflectance.
relationship
If
surface
roughness is greater than one-eighth of the wavelength, the r e f le c t ­
ance is diffuse;
otherwise, reflectance is prim arily specular.
This
explains why some surfaces look rough at one frequency and smooth in
another.
Equations 1 and 2 apply for conditions involving an external
source.
In the v is ib le and near-infrared spectral
regions, solar rad ia­
tio n is the primary source for reflected radiation at the earth sur­
face.
In this spectral
region, different materials possess d iffe r e n t
r e f le c tiv e properties.
These spectral differences can be analyzed and
used in
many materials
discriminating
radiation
is
r e la tiv e ly
constant at
on earth.
Given that
a given zenith
solar
angle--assuming
constant atmospheric absorption and transmission— reflectance is ana­
lyzed through radiance.
per unit
1975).
Radiance (L) can be defined as radiant flux
of projected source area in a specified direction
Radiance is calculated for a wavelength channel,
L = ^ / 2 |^E( A)R( A) (Tb( X)Tz ( A)p(T)sin B + p'
(Janza,
by
(x )J d X
(3)
where E(x) is the specular solar irradiance at the top of the atmos­
phere at normal incidence, R(x) the spectral response function of the
wavelength channel, Tb ( x ) the monochromatic one-way tranm issivity of
the atmosphere at elevation angle B, Tz(x) the monochromatic trans­
m issivity of the atmosphere in the zenith direction for solar radia-
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
12
tion reflected by the surface to the nadir-viewing sensor, p(A) the
reflectance
of the surface,
and
p' bU
)
the atmospheric reflectances
as dependent on solar elevation, B.
Microwave emissions can be measured in two modes—active
(sur­
face reflec tio n of energy from a source) or passive (emitted from the
surface).
This is in contrast with visib le and infrared data which is
generally sensed in a passive mode.
conducted using l i d a r ,
Active visib le research has been
but measurements are quite complicated.
The
active microwave (radar) responses from many surfaces have been exten­
sively analyzed primarily due to the application of active systems by
the m ilita r y ; however, passive microwave research has been less devel­
oped due to lim itations in spectral resolution or antenna size.
Since
active and passive microwave data are two d iffe re n t sensing modes, the
responses are expressed d if f e r e n t ly — radar returns are expressed in a°
and passive microwave returns are expressed as brightness temperature.
The microwave region has more complex relationships which define
reflected radiation.
t e r is t ic s
With active microwave systems, surface charac­
have been analyzed by comparing the
power returned to
a
radar receiver with the transmitted power as calculated from the radar
equation
r
where Wr
4-nR2
is
the
4ttR2
received
r
power,
Wt the transmitted
power,
&t the
gain of the transmitting antenna in the direction of the ta rg e t, R the
distance between the antenna and ta rg e t,
a the radar cross section,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
13
and Ar the e ffe c tiv e area of the receiving antenna aperture (Janza,
1975).
Most applications involve targets which are larger than a re­
solution cell of radar.. Consequently, i t is more convenient to consi­
der the average return
d iff e r e n t ia l
a0.
power over an irradiated area.
cross-section
The above equation
is
known as the
implies that
radar
scattering
returns
The average
c o e ffic ie n t,
from a target
depend upon the strength of the transmitted energy and the re fle c tin g
cap ability of the ta rg e t.
The target roughness and d ie le c tric charac­
t e r is t i c s produce varying proportions of the return described by
backscatter.
the
In addition to determining the return power, scattering
properties of targets can also depolarize the return causing cross­
polarized (HV or VH) radar data to be useful in geological and ag ri­
cultural applications.
Such depolarization leaves the cross-polarized
data sensitive to d ie le c t r ic properties.
The effe c t
of roughness and the
d ie le c tric
constant
on active
and
passive microwave returns d i f f e r .
The roughness effect
dominates
the
active microwave returns, while the d ie le c tric influence
dominates
the
passive microwave return. The effects also depend on look angle.
At high look angles, roughness becomes even more predominant.
According to Planck's equation, emitted radiation from the earth
surface peaks in the thermal infrared region.
Total
emitted surface
radiation is described by the Stephan-Boltzmann Law (Planck's Equation
applied over a ll wavelengths):
R = es aT*
(5)
where R is emitted radiation, es is the emissivity of the surface, a
is
the
Stephan-Boltzmann
constant
(5.7xlO~ 8Wm“ 2oK“ lt) ,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
and
14
T is the absolute temperature.
Most natural objects have emissivities
between 0.8 and 1.0 in the thermal region.
the
microwave
region.
Several
This w ill be d iffe re n t in
facto rs,
such
as
topography
and
weather, have made i t d i f f i c u l t to c la s s ify crops using thermal i n f r a ­
red data.
Thermal
data,
however,
have often been used to evaluate
soil moisture conditions.
Emissions in the passive microwave region are much smaller than
thermal
infrared
emission.
Emitted
responses
Rayleigh-Jean's approximation to Plank's equation
are
based
upon
(Wolfe and Z issis,
1978)
where R5 is
radiation
temperature,
k Plank's constant and x the wavelength.
radiation
from a blackbody, T the absolute
in the microwave region is
temperature.
pheric
(brightness)
It
emissivity
often
The emitted
expressed as brightness
can be expressed as a function of ground and atmos­
(eg
and
ea ) ,
ground
reflectance
(Pg),
and sky,
ground, and atmospheric- (clouds, water vapor, particulates) tempera­
tures (Ts ,Tg,Ta):
(7)
Effects of the atmosphere are often n e g lig ib le , especially with cloud­
less sky.
Consequently, Ta is often neglected giving
tk
b =
g g + (1 " Og T *s
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(8)
15
Since
Ts and
(1
-
eg)
are both small,
the
reflection
term,
(1
-
£g) Ts> is often omitted leaving only
Tb -
The variation
=gTg
in ground em issivity,
d ie le c t r ic constant and roughness.
<9 >
eg provides much information on
Since healthy crops contain over
50% water and appear rough in certain microwave wavelengths,
emissivity
w ill
vary
under
d iffe re n t
vegetation
conditions
ground
(Peake,
1966; Sibley, 1973).
Given
the
spectral
theory,
which
is
applicable
at
a ll
wave­
lengths, one must turn to the factors which primarily influence spec­
tra l
responses of agricultural
crops.
To simplify the description,
the electromagnetic spectrum w ill be divided into the v is ib le /in fr a re d
and the microwave regions.
V is ib le /In fra re d Responses
Water and chlorophyll
influence
vegetation
are the most
and soil
reflectance
important
in
the
substances which
v is ib le /in fr a r e d .
At high solar elevation angles, water strongly absorbs solar radiation
in both the v is ib le and infrared .
Consequently, v is ib le and infrared
reflectance from a soil would often decrease under high moisture con­
ditio n s.
The moisture e ffe c t is highly dependent on conditions within
the top thin layer of the surface being observed.
ture can be d ire c tly
No subsurface mois­
determined using wavelengths
shorter than
one
centimeter (Davis et a l . , 1965).
Leaves, however, have a completely d ifferen t spectrum.
Fresnel
reflectance at air/w a te r
Due to
interfaces within the leaves, near
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
16
and middle infrared radiation is strongly reflected (Figure 1) (Gates,
1980).
Figures 2
demonstrates the relationship between biomass and
reflectance is dependent upon crop type and maturity (Tucker et a l . ,
1979, Park and Deering,
1981).
Reflectance
increases
to ta l biomass in the near- and middle-infrared region
ated reflectance is reached.
At that
s itiv e to increases in biomass.
reflectance in this
quently
cover,
for
corn
region
and
reflectance is
given period of time.
estimates
in th is
chlorophyll
blue regions,
reflectance
(1969)
until
a satur­
point reflectance becomes insen­
Then at a point near maturity, the
begins to decrease with biomass.
soybeans,
crops
with
insensitive to to ta l
Conse­
a near-complete
biomass
canopy
increases
for
a
Other techniques are needed to quantify biomass
region.
content.
rapidly with
Reflectance is
Chlorophyll
and has a slight
also
a
absorbs radiation
reflectance
in the near infrared.
in
the
function
of the
in the red and
green
and high
Studies by Hoffer and Johannsen
indicated changes in chlorophyll
contentallowed other caro­
tenes and xanthophylls to become evident, thus affecting primarily the
v is ib le /in fr a r e d reflectance.
dependent
on the
Since infrared reflectance is strongly
air/w a te r interface
and chlorophyll
content,
any
environmental e ffe c t which changes the area of air/w ater interface or
the
number of
disease
leaves w il l
and stress
influence the reflectance. Consequently,
(moisture,
infrared reflectance.
n u trie n t,
etc .)
d ra s tic a lly
decrease
In spite of these e ffe c ts , differences between
the v is ib le and near infrared data have been the basis for classifying
vegetation
and estimating biomass.
given phenological
The main premise is that
at a
period for a crop, spectral characteristics in the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction
100
80
8 le a v e s
60
O
2 lea ves
c
H
O
40
20
prohibited without perm ission.
0
0.5
1.0
1.5
2.0
2.5
WAVELENGTH(^m)
FIG. 1
Reflectance of 2 and 8 stacked mature cotton leaves. Standard deviation between
observed and calculated points is about 1%. From Allen et a l . , 1970.
18
1.0
8
VEGETATION INDEX
.6
.4
,2
Me tu ri ty
0
-.2
0
10
20
30
40
50
S 0Y8EA N WET BIOMASS (in thousands)
FIG. 2.
Averaged normalized difference (IR -red /IR + red) values
plotted against soybean wet biomass. From Tucker et a l . ,
1979.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
19
crop allow for crop discrimination— assuming that spectral differences
within the crop attributed to stress or disease are less than the d i f ­
ferences between crops.
Also,
i f two crops have the same phenology
and spectral characteristics, they w ill
Given difference in chlorophyll
not be spectrally separable.
content and le a f succulence between
plant species, classifica tion and biomass estimation models have been
developed.
The detection is consequently based on v is ib le /in fr a re d
differences between crop types.
Different biomass models w ill be dis­
cussed la te r.
Integrating the soil and vegetation reflectance has been a prob­
lem.
Many
have
trie d
to
model
canopy
(integrated)
reflectance
(Kubelka and Munk, 1931; Chance and LeMaster, 1977; Richardson et a l . ,
1975).
Chance and LeMaster (1977)
used the Suits model to estimate
reflected and non-reflected radiation from a boundary layer.
the model
function
showed l i t t l e
of
solar
However,
agreement with wheat reflectance data as a
angle.
Richardson
et
a l.,
(1975)
used
the
Kubelka-Munk and a regression model, using biophysical parameters for
extracting plant, s o i l , and shadow reflectance components of cropped
fie ld s .
The model did correlate well to actual scene reflectance.
Microwave Responses
Three
factors
primarily
affe ct
reflectance
and emission
from
agricultural surfaces in the microwave region--surface roughness, soil
moisture,
and vegetation.
agricultural
To f u lly
understand the
return
scene, one must account for each fa c to r.
from an
Each factor
w ill be discussed in greater d e ta il.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
20
Roughness - As mentioned before, for active microwave systems o°
is
governed
(1966)
by
the
geometric properties
of the
surface.
Beckman
found the backscatter to be related to the variance and mean
slope of the surface.
Ulaby et a l .
utable
to
soil roughness
nad ir,
which is
(1981)
and Ulaby and Bare (1979)
the
decrease with
least
tant in the radar return.
(1978) found a0 variations a t t r i b ­
look
sen sitive to
angle out to 10 ° from
roughness.
Fenner
et
a l.
found row direction was very impor­
Rows perpendicular to the emitted beam have
much higher returns compared to rows p a ra lle l to the emitted beam.
certain
At
look angles and frequencies the surface roughness e ffe c t may
dominate the terms that are due to changes in the d ie le c t r ic constant
brought about by changes in soil moisture.
Wang et a l.
(1980)
noted t i l l e d
row direction
is also a major
factor in passive microwave emission, especially when the antenna is
directed o ff nadir to the ground.
The difference between vertic al and
horizontal polarized returns in passive microwave returns can be r e l a ­
ted
to
1979).
the
soil
surface
roughness
(Newton
1977,
Choudary et
a l.,
The e ffe c t appears to decrease at look angles larger than 35
degrees o f f nadir.
ativ e height of
The roughness e ffe c t is also dependent on the r e l ­
the roughness in
re la tio n to the wavelength of the
sensor.
Soil
moisture -
The e ffe c t
response is
of
the
d ie l e c t r i c
constant
active
microwave
t u re .
In the high frequency microwave regions, soil has a d ie l e c t r i c
constant of 3, and water, 81.
demonstrated by changes in soil
on the
mois­
Consequently, any s ig n ific a n t change in
soil moisture should be detectable.
The relationship has been studied
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
21
in
great
detail
using
active
systems.
Laboratory
Lundien (1971) showed L band (21 cm wavelength)
sensitive
to
soil
moisture
differences
experiments
by
data should be more
than K band
(1.55
cm wave­
length) due to differences in the d ie le c t r ic constant of water at the
two frequencies.
However,
Ulaby et
a l.
(1978)
microwave data to be most sensitive to soil
the surface two centimeters.
found C band active
moisture differences in
The severe e f fe c t of roughness that is
inherent in active microwave returns was minimum in Ulaby's experiment
which was carried out over t i l l a g e
common to Kansas using C band at
10° o f f nadir.
F ield experiments by Newton (1977) and analysis of s a t e l l i t e data
by McFarland (1976) had shown L band passive microwave data was sensi­
t iv e to soil moisture changes within approximately the
layer.
surface 5 cm
Other s im ila r work had been done in using active and passive
microwave data.
An excellent review of studies concerning soil mois­
ture estimates using microwave systems was given by Schmugge (1978).
Vegetation
-
The effe c t
vegetation
on the active microwave
return has been studied since the mid-1960s.
Early work concentrated
on analyzing effects
1967,
in the K band (l-2'"cm)
Ellermeier et a l . ,
potential
tool
1969).
decreases.
to
the
At
time
look
that
angles
indicated
radar was a
The response is based on
As a crop matures, the crop moisture
the
crop
begins
of greater than
strongly correlated to plant water content
and Bush, 1976a and 1976b).
region (Simonett et a l . ,
The studies
for discriminating crops.
both moisture and roughness.
increases
of
to
40°
senesce
and then
from n a d ir,
o° is
in corn and wheat (Ulaby
Consequently, biomass could be estimated
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
22
fo r the growing period.
Also, crops have d iffe r e n t morphologies which
can be applied to crop discrim ination.
influence the scatterometer return,
However, other factors may
de Loor et a l .
(1974) found o° to
vary as much as 4 to 5 db under d iff e r e n t wind speeds.
Brakke et a l .
(1981), however, found no influence of wind speed on o° over wheat and
sorghum in the K band region.
Ulaby et a l .
(1975) found crops can be
discriminated with multi frequency v e r t ic a ll y polarized data (between 8
to 18 GHz (2 .5 -3 .5 cm)).
Look angles at 30° to 65° from nadir removed
the soil moisture e ffe cts leaving only the vegetative e ffe c ts .
Com­
parisons between l i k e - and cross-polarized active microwave data (1.25
GHz— 25 cm) also provided valuable information on vegetation.
Classi­
fic a tio n accuracies improved from 65% to 71% by including cross with
lik e -p o la r iz e d data (Ulaby et a l . , 1980).
Comparisons of d iffe r e n t polarizations of passive microwave data
also
indicated
1979).
crop
morphological
differences
(Kirdyashev
et
a l.,
Relationships between biomass, height, plant moisture content
and brightness temperature at m ultiple frequencies were found.
parameters
can
be
related
to
crop
type
differences.
Such
The passive
microwave data, however, is less practical for crop discrimination due
to the poor resolution associated with a i r c r a f t and spacecraft passive
systems.
To summarize, active microwave data at look angles greater than
30°
from nadir
appear
to
be related
which can imply crop type differences.
to
vegetative
ch a ra cteristics
Active microwave systems are
more sensitive to roughness, while passive systems are more sensitive
to soil
been
moisture.
related
to
Multi frequency passive microwave data
s im ila r
vegetative
ch a ra cteristics
but
also
have
are
less
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
23
sensitive
to
roughness
and
vegetation,
and
have
resolution c a p a b ilitie s than the active systems.
a ll
less
acceptable
The s e n s itiv ity to
three factors is dependent on wavelength (frequency) as well
as
polarization and look angle for both active and passive systems.
C lass ific atio n Models
Supervised Models
From the previously mentioned v is ib le and near-infrared r e la tio n ­
ships of vegetation,
oped.
several
Heydorn et a l .
(1979b)
c la s s ific a tio n
models have been devel­
gave a general
description of several
supervised and unsupervised techniques which emerged from studies with
LACIE.
Supervised c la s s ific a tio n techniques became one of the key c la s ­
s if i c a t i o n
techniques.
The
methods
required
information
classes--means, standard deviations, or vectors of data.
mation was termed the tra in in g c la s s i f ie r .
on
the
This i n f o r ­
Using various comparison
techniques, sampled data was compared to the tra in in g c l a s s i f i e r and
placed into the proper class.
To separate classes, discriminant func­
tions as determined from class s t a t is t ic s were calculated.
which f e l l
Any sample
on e ith e r side of the function was placed into one of the
classes (Swain and Davis,
1979).
Several
of the widely used super­
vised techniques were maximum likelihood per point, maximum likelihood
per homogeneous group, ECHO--Extraction and Classification of Homogen­
eous Objects—minimum distance to the class means, and standard devia­
tions to calculate the p ro b a b ility of including the sample in a given
class.
The only difference between the ECHO c l a s s i f ie r and the maxi­
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
24
mum likelihood
c l a s s i f ie r
was the
sample;
ECHO uses a homogeneous
group of sample points, while the maximum likelihood per point method
analyzes
only
one sample point at a time.
In the minimum distance
c l a s s s if ie r , a Euclidean distance was calculated between the data vec­
tor at one point and the mean vector.
given threshold,
layered
I f the distance was less than a
the point was placed
c la s s i f ie r
differed
into
the given
class.
from the maximum likelihood
per
The
point
c l a s s if ie r in that multiple decisions, rather than one decision were
made at each point .
to be used.
This allowed for d iffe r e n t subsets of channels
Bauer et a l.
(1977)
found no sig n ific a n t difference
accuracy using each of these techniques.
in
However, the minimum dis­
tance c l a s s i f ie r had the lowest computer cost.
Unsupervised Models
Unsupervised
c la s s if ic a tio n ,
information on classes.
ages.
or
clu sterin g,
models
require
no
The techniques grouped sim ilar spectral aver­
The most widely used technique involved the minimum distance
between observations
(Johnson,
1967).
Another s im ila rity
c r ite r io n
technique involved minimizing variance or the sum of squares.
techniques were described by Orloci
Other
(1978) and Hartigan (1974).
Such
techniques had been used in combination with other supervised tech­
niques
to classify
agricultural
scenes and estimate areal
from Landsat data (Heydorn et a l . ,
1979a).
s ifie r
which
was
the
"tree
structure"
coverage
A major part of the clas­
defined decision
points
as
determined by variable differences between spectral classes involved.
C la s s ific a tio n accuracies using these techniques had varied from
one location to another.
The areas
having the
lowest accuracy had
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
25
"confusion" crops growing in the same area—crops which have the same
spectra at a given period.
Accuracies ranged from 60% to over 90% in
some areas.
In the microwave region,
been equally as accurate.
to classify
data.
Simonett et a l .
an agricultural
Ulaby et a l.
success in classifying
(1980)
(1967) was one of the f i r s t
scene using l i k e -
and cross-
Other work was done
by Morain and Simonett (1967), Schwarz and Caspell
MacDonald (1971), and Ulaby et a l.
pasture,
timber
polarized
also c la s s ifie d correctly 71% of an area
using l ik e - and cross-polarized microwave data.
c la s s ifie d
vegetation has
(1975).
and bare
using airborne scatterometer data.
soil
(1968), Waite and
Blanchard et a l .
with
reasonable
(1979)
accuracy
Land use was correctly determined
in greater than 80% of the cases by analyzing the differences in the
10° and 35° look angle a0 values for lik e -p o la rize d data, differences
in
the
lik e -
and cross-polarized data at 10°
cross polarized data at 10° look angle.
combined
active
and
near-infrared data.
Landsat
data
passive
microwave
Ulaby et a l.
collected
over
(1931)
an
look angle,
and the
Few studies, however,
data
with
have
v is ib le
and
analyzed scatterometer and
agricultu ra l
area
in
1978.
C lass ific atio n accuracy increased 10% by including scatterometer data
with Landsat data.
Further work needs to be done relatin g vegetation
type to v is ib le , in fra re d , and passive and active microwave data.
Biomass Models
V is ib le /In fra r e d Region
Because infrared le a f reflectance is strongly influenced by the
number of leaves, which in turn
is
related to
plant
biomass,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
many
26
models have been developed using a combination of
reflectance data.
v is ib le /in fr a re d
Only a few significant models are mentioned here.
The transformed vegetation index (TVI) has been used primarily as
an estimator of rangeland biomass (Rouse et a l . , 1973; Deering et a l . ,
1975).
The model was expressed as
TVI =
(MSS7 - MSS5)
(MSS7 + MSS5)
n ,
, .
( 10)
0,5
where MSS7 and 5 are radiances from Landsat bands 7 (0 .8 -1 .1 pm) and 5
(0 .6 -0 .7 pm), respectively.
The ratio was used as a normalizing term
to remove temporal index variations, such as illum ination differences
due to aerosols and solar angle, and 0.5 was added to keep the term
under the square root
from going negative.
A modification of the
index involved replacing band 6 ( 0 .7 -0 .8 pm) data for band 7.
modified index was TVI 6 .
The
Both were well correlated to green biomass.
Kauth and Thomas (1976) developed transformation matrices which
convert Landsat data for cultivated agricultu ral areas to data which
enhanced greenness, brightness, and yellowness.
formed data from temporal
8y comparing trans­
scenes, the progression of phenology f o l ­
lowed the shape of.a "tasseled cap."
Converting the matrices to index
GVI = -0.290 MSS4 - 0.562 MSS5 + 0.600 MSS6 + 0.491 MSS7
(11)
and the brightness index was
SBI = 0.433 MSS4 + 0.632 MSS5 + 0.586 MSS6 + 0.264
MSS7
(12)
where MSS4, 5, 6 and 7 refer to Landsat bands 4, 5, 6 and 7 d ig ita l
counts.
GVI had been found to be highly correlated to le a f area index
(Richardson and Wiegand, 1977).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
27
Another vegetation
index model used to estimate biomass is the
perpendicular vegetation index (PVI), developed by Richardson and Wiegand (1977).
PVI was calculated by the equation
PVI = —y/(Rgg5 - Rp5) 2 + (Rgg7 - Rp7) 2
(13)
where Rp is the reflectance fo r a candidate vegetation point for Land­
sat bands MSS5 and MSS7 and Rgg is the reflectance of soil background
corresponding to the same candidate vegetation point.
Figure 3 des­
cribes the principle of the perpendicular vegetation index.
Simply,
PVI is the perpendicular distance from a given radiance in bands 5 and
7 to the soil background lin e .
Wiegand
(1977)
that
PVI6
I t was demonstrated by Richardson and
and TVI 6
(where
Landsat
band 6 is
used
instead o f band 7) are both highly correlated io leaf area index.
Microwave Models
Work is just beginning in relating microwave data to vegetation
characteristics.
Brakke et a l.
(1981) related corn, wheat, and sor­
ghum ch a ra c te ris tic s , such as plant moisture content, crop height, and
le a f area index, to microwave,
v is ib le and near-infrared data.
The
authors determined dry matter was highly correlated with <j° at look
angles
of
70°
o ff
nadir.
Jackson et
al.
(1981)
compared biomass
estimates to changes in the slope of regression lines relatin g soil
moisture and normalized passive microwave brightness temperature.
As
biomass increased, the s e n s itiv ity of normalized brightness tempera­
ture related to soil moisture decreased.
Final L ite ra tu re Overview
From the research reported, i t is evident that simultaneous data
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
28
2.3
2.0
1.8
1.5
P V I >o
VEG ETATIO N
1.2
.9
,6
3
0
1.30
1.95
2.60
3.35
LANOSAT MSS7 (mw; cm'^ ster '*)
FIG. 3.
Diagram i ll u s t r a t i n g the principle of the perpendicular vege­
tation index (PVI) model. A perpendicular from candidate
plant coordinates (Rp5, Rp7) intersects the soil background
lin e at coordinates (Rg5, Rg7). A PVI=0 indicates s o i l , and
a PVI>0 indicates vegetation. From Richardson and Wiegand
(1977).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
29
using v is ib le ,
lected.
in frared ,
and microwave bands
have rarely been col­
More data sets of v is ib le , infrared, and microwave data are
needed to compare against vegetation type and c h a ra c te ris tic s , such as
biomass.
According to theory, microwave frequencies should be sensi­
tiv e to d iffe re n t vegetation characteristics (prim arily geometric and
d ie le c tric properties) than characteristics seen by v is ib le and i n f r a ­
red data.
As a re s u lt,
c la s s ific a tio n accuracies and biomass e s t i ­
mates should improve by including microwave (active or passive) bands
with visib le and infrared.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
30
CHAPTER I I I
DATA COLLECTION
A ir c r a ft
data were collected near Guymon,
Oklahoma in August,
1978, and near Dal ha rt, Texas in August 1980.
Data collection and
processing w ill be described for each s ite .
Guymon A irc ra ft and Ground Data
In August, 1978, a i r c r a f t and ground data were collected in com­
mercial agricultural
fields located from 3 to 20 km southwest of Guy­
mon, Oklahoma and near Clayton, New Mexico (Figures 4a through 4h).
vegetative cover in the area included bare s o i l , corn,
a lfa lfa .
Soil
type was generally a s i l t y
sorghum, and
clay (averaging 35% clay,
35% s i l t , and 30% sand) with many areas having a caliche (CaC03) layer
near the surface.
D ifferen t t i l l a g e practices allowed spectral data
from sorghum and bare fie ld s having rows perpendicular and p arallel to
the f lig h t
line to be analyzed.
A ir c r a ft and ground data were col­
lected in field s along four f li g h t lines covering 38.4 km2 area (1.6 x
24 km).
A irc ra ft data collected by the NASA C-130 on August 2, 5, 8 , 11,
14, and 17 consisted of (1) seven scatterometer frequencies and polar­
izatio n s ,
(3)
fiv e
( 2) three passive microwave frequencies and p o larizatio n s,
visible/near-infrared/therm al
channels,
(4)
radiometer thermal data, and (5) black and white aerial
The a ir c r a f t
Barnes
photography.
flew at least twice at 500 m over each f lig h t
each f li g h t day.
PRT-5
line on
Also, on August 5, the C-130 collected only sca tter­
ometer data over fie ld s near Clayton, New Mexico.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction
GUYMON, OKLAHOMA 1978
LEGEND FOR FIELD M APS
CROP
Corn
1 ,2 & 3
Milo
A lfalfa
Consult field notes for row crop orientation to aircraft flight lines.
prohibited without perm ission.
P re p a re d b y th e T e x a s A & M U n iv e r s it y R e m o te S e n s in g C e n te r.
B a s e d a ta c o m p ile d fro m U SG S to p o g r a p h ic m a p s ,R .S .C . te a m
f ie l d n o te s a n d N A S A c o n tr a c te d a e r ia l p h o to g ra p h y c o lle c t e d
A u g u s t 2 -1 7 , 1978.
A P P R O X IM A T E
I
200 0
0
4a
1 : 49000
0
200 0
2____________
FIG.
SC ALE
40 0 0
6000
2 M IL E
8000
10000
12000
14000 F E E T
1_________________ 0___________________________________ 2 K IL O M E T E R
Legend f o r the Guymon, Oklahoma f i e l d maps.
C/)
<f)
035
-,-n «. 0 V°
-»AvVo
i
o
o
■O
j. Q At
^
\.rr
ight owner. Further reproduction
.
1
°
irVAHv
'
3/iOU
T .2 .N .
MAP 2
f we,lso ■)
l> ^ P
s
1^0
s wells
prohibited without perm ission.
I
> 1 - 0
7 -"VI \
.»
/
O
Gas ^ W e llil R" T3 r
n S •
~
_
I
_
"'
36 30
rAT
, ..
_..
JwndraiU
GUYM ON AREA M AP
INDEX TO FIELD MAPS
O KLAH O M A
Approximate scale is 1 : 250,000
FIG. 4b
Area map of Guymon showing the r e la tiv e locations of each f i e l d map.
OJ
r\3
33
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
FIG.
4d
Locations
of
the
sample
fields
i.v
at Guymon.
«/>V
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
FIG.
4e
Locations of the sample f ie ld s at Guymon.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
36
Guym on,O kla.
Hap 3 o f 3 South end
27
101 40 10
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
I
CLAYTON, NEW MEXICO 1978
CROP
LEGEND FOR FIELD M A P
Com
Milo
Sample fields in Clayton had no ground truth measurements.
P re p a re d b y H ie T e x a s A & M U n iv e r s it y R e m o te S e n s in g C e n te r.
B a s e d a ta c o m p ile d fro m USGS to p o g r a p h ic m a p s , R .S .C . team
f ie l d n o te s a n d v is ic o r d e r d a ta fro m a N A S A a ir c r a f t f lig h t on
A u g u s t 5 , 1978.
A P P R O X IM A T E
3
2
1
300 0
2
FIG.
4f
—
6000
SC ALE
1 :7 1 0 0 0
0
3 M IL E S
-----
9000
I___________ 0_______
120 0 0
160 00
18000
21000 F E E T
3 K IL O M E T E R S
Legend for the Clayton, New Mexico f i e l d maps.
CO
103 ° 0 0 /
U6O0
iavel Pit
14 7 / 6
,4 9 7 0
3 6°
F IE L D MAP
CLAYTON AREA M AP
Approximate sc ale is 1 : 2 5 0 ,0 0 0
FIG.
4g
NEW
MEXICO
Area map of Clayton showing the r e la t i v e location of the
f i e l d map.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
39
FIG.
4h
Location
of the
sample
fields
at
Clayton.
.X,n.K
.or.njc
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
40
The scatterometer frequencies and polarizations included ( I )
GHz VV (K band)
v e r t ic a lly polarized transmitted and received),
13.3
(2)
4.75 GHz HH (C band horizontally polarized transmitted and received),
(3)
4.75
GHz HV (horizontally
polarized transmitted
and v e r t ic a lly
polarized received), (4) 1.6 GHz HH (L band), (5) 1.6 GHz HV, ( 6 ) 0.4
GHz HH (P band),
referred
to
as
and
(7)
K band,
0.4
GHz HV.
C band,
remainder of this report.
These frequencies
L band or
w il l
P band throughout
The polarizations w ill
be
the
be referred to as
lik e pole or cross pole instead of HH or HV, respectively.
Data from
eight look angles from nadir were processed for each frequency:
5°,
10°, 15°, 20°, 25°, 35°, 4 0 °, 45°.
Passive microwave data were collected in 1.6 GHz (L band) h o r i­
zontal
p o larizatio n ,
polarizations.
and 4.75 GHz (C band)
vertic al
and horizontal
These data w ill be referred to as L band ho rizo n tal, C
band vertic al and C band horizontal, respectively.
Five channels from the modular multispectral
available:
ym,
(1)
(3) channel
channel 4: 0.548-0.583 urn, (2)
scanner (M2S) were
channel 7: 0.662-0.701
8 : 0.703-0.747 ym , (4) channel9: 0.770-0.863
(5) channel 11: 8.000-12.080
ym ,
and
ym.
Barnes PRT-5 measurements were also included to c a lib r a te the M2S
thermal
band (channel 8 ) and normalize the passive microwave bright­
ness temperature.
The
sensors
were operating
at
d iffe re n t
times
throughout
the
stucjy because the active microwave data would in te rfe re with the pas­
sive microwave
data.
Windy conditions
th ir d run over each f li g h t lin e .
on August 14 also
forced a
Table 1 l is t s the operating sensors
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
41
TABLE 1.
Operating Sensors for the Guymon, Oklahoma Study
Date
Line
Run
8/2/78
1-4
1
8/5/78
Operating Sensors
a ll scatterometer; M2 S; PRT-5; C-band
passive microwave; photos;
8/8/78
8/11/78
1-4
2
8/17/78
8/14/78
K-band, C-band, P-band scatterometer; and
L-band passive microwave; PRT-5; photos
1-4
1
a l l scatterometer; M2 S; C-band passive
microwave; PRT-5; photos
1-4
2
K-band, C-band, P-band scatterometer; and
L-band passive microwave; PRT-5; photos
1-4
3
a l l scatterometer; M2S; C-band passive
microwave; PRT-5, photos
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
42
fo r each f l i g h t line and run. Field averages were determined for each
sensor.
Because of the uncertainty of the target
and look angle,
f i e l d averages were deleted from the data set when the NASA C-130 had
excessive ro ll (greater than 3.5°) and/or d r i f t (greater than 9 ° ) .
Soil moisture samples were collected at eight points approximate­
ly 200 m apart within each 32 hectare f i e l d (Figure 5 ).
Samples co l­
lected at each s ite were 0-2 cm, 2-5 cm, 5-9 cm, 9-15 cm, 0-15 cm,
15-30 cm, and 30-45 cm (Figure 6 ) .
each depth.
Field averages were calculated for
Data included in calculating the average were from sites
within the maximum sensor swath width.
data from a ll
In the majority of the cases,
eight sample points were included.
Approximately one-
th ird of the fie ld s were sampled on f li g h t days.
As a r e s u lt, mois­
ture averages fo r field s not sampled on f li g h t days were interpolated
from time series plots of measurements taken the day before or the day
a fte r
f li g h t s .
wet/dry
lected
Field
notes of t i l l a g e ,
areas were also tabulated.
at
experiment
center
pivot
location
and
No biomass information was co l­
Guymon; however, photographs of crops at the time of the
were collected which provided a rough estimate
of
crop
cover.
Dal hart A ircraft and Ground Data
During August, 1980, a ir c r a f t and ground data were collected in
commercial agricultural fields 20 km northwest of Dalhart, Texas (F ig ­
ures 7a through
7e).
Figure 7a represents the general
area showing the re la tiv e locations of 7b, c and d.
legend which describes the crop types.
view of the
Figure 7e is the
Crop types within the area
included bare s o i l , pasture, corn, a l f a l f a and sorghum.
The soil type
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
43
All odd numbered fields.
a
C=3
04
ro
All even numbered fields.
§
o
04
CO
CO
CO
o
330
660
660
660
330
2640 FEET —&*
' These points were moved outside the pivot boundary for Fields 5 & 6.
FIG. 5
Sampling pattern fo r fie ld s at Guymon and Dalhart.
Points 1, 2 , 7 and 8 were moved outside the c ir c le fo r
rectangular f i e l d s .
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
44
SOIL SURFACE
0 - 2 cm
%*»**• •
• •«,i * * •
, • • <•'* *
1 * *•■ o .t
• .#v°*' ■• * © ! •
# .•••« «
*
, ..*•
m
, • O «, * • f
v• *• *v. . °; ••' • • • *
.•
•
* •
■V
. ‘ * • * * ft
*o • .• •
••*
.
O
••••.*. .v : ■»/.;
'* . • • O • • • • •
m
.v
o •
;»•
■•
•
• .O * .
®
•
•. * •
.* 0.*•** *o ' “
•*
• «'
• .
LU
-J
Q_
<c
OO
LU
» . . *• * o ;
* ’ * .® V
•©• •
.
.
•
•
-,v.
.
*
4 • •*
, •o * . .
* - * • \ -7_fl
-/-V ' H
.*•
•
o
' o ‘ . o'. •'
• • . •o
• f .
••
.*« ,/•
* * .•
• • o *.. •o * • «
9 .•
p .
2 - 5 cm
■••••* * ; •
• ••
• * •*
. • * '
«. «
' * #\*r .' ** *’
*- -I0.’ ‘ **.®#*■•*•
■
*>•
. • • . ••
• • .
.0
C£
•
5 - 9 cm
O
O
•
, V*
o
/ * *.©. •.
* •: • :
# *°. V
C
G
LO
o
. •o
s'
•
* !'»
* • *\-
? ; ; > .V
t
o. * •
• • .* 1 #0 \
9 - 1 5 cm
‘ S \
o '
» • *o• * •
o
»
\
' • ft. ‘
.■ ft. * I * ■
r■: ^
* ' f t . ' . *■
* f *0
’'
•*
o . ^ - o . •
\ +
,
• O.
\V*.. ; •
• • 4
.‘ o ';«
»• ■.* * • ~
V:»i
>*
.
'./a
°"o ■ >
0
‘'.'J’ S ’.V
•
1 .•» . ,
•
•ft*,•••’« «' '
•■.* V° '*: •
v*«.. *
. •
. o • .
’. b• •*'
*. o* •' •
6, •
FIG.
6
Soil moisture sampling depths at Dalhart and Guymon. The
15-30 and 30-45 cm core samples were also taken in addition
to the above. Samples were collected from 5-9 cm and 9-15 cm
at Guymon and 5-15 cm at Dalhart.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
45
DALHART, TEXAS 1980
LEGEND FOR FIELD MAPS 1,2 & 3
CRO P
I"; ! i I j I 11
Bare: wheat stubble
1111
' ' I'
sy / / /
'///,
Corn
disked wheat stubble
Alfalfa
mulched wheat stubble
Pasture
Grazed
Millet
I*1* *'1-1
*.
Milo
—
Flight line markers
▲
Corner reflectors
*
Rain gauges
•
Vegetation sample sites
now Direction was east-west tor an sample n eias witn row crops.
A P P R O X IM A T E
SC ALE
1 :4 9 0 0 0
G
2000
400 0
\
6000
0
2 M IL E S
3000
10000
12000
140 0 0 F E E T
2 K IL O M E T E R S
P re p a re d b y th e T e x a s A & M U n iv e r s it y R e m o te S e n s in g C e n te r. B a s e d a ta c o m p ile d fro m US D S t o p o g r a p h ic m o p s ,
R .S .C . te a m f ie l d n o re s a n d N A S A c o n tr a c te d a e r ia l p h o to g r a p h y c o lle c t e d A u g u s t 14*18, 1980.
FIG.
7a
Legend fo r the D a lh a rt, Texas f i e l d maps.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
■o
CD
iSsr-i
—L\ —1- J
N -^ T
;
l\
I--J v
/ x
'A
C
<f)/)
^
V
CD
O
O
■o
ight owner. Further reproduction prohibited without permission.
- t
'— 1
«15 / I
I I
— t —
i! . i M A P
warMiMx
dmilll
_ .
I
4040
— I
3
MAP 2
MAP 1
V-070
1
Windmil:
1“
“ - tevsoi-----[--- i
74065 T"
4050
4055
Dajha
I ^
.
DALLAM COUNTY
HARTLEY COUNTY
I
I
"
„--i
/l
/
llliiat
windm
/
I 5 .'A: NT“ f Wmdm,ll^e7--"A T
\
-»«5
i" k
k
Oathirt
^
V I
I ■. ^
K4DM
3 6 °0 0 /
103
00
DALHART AREA M A P
IN D E X TO F IE L D M APS
ST” 1 T E X A S
A p p r o x im a te s c a l e
FIG.
7b
1 : 250,00 0
Area map of Dalhart showing the r e la tiv e locations of each f i e l d map.
V
\
-P*
CT)
Dalhart.
at
fields
sample
of the
Locations
7c
FIG.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
48
7d
Locations
of the
sample
fields
at
Dalhart.
+
3 S K ^\:$t&»sTG
- -•-J_:-....rT ' / jt f-
FIG.
Y w
.■/
.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
49
:; • ■
a il
FIG.
7e
Locations
of the
sample
fields
at
Dalhart.
iSwii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
50
of the surface 30 cm was a sandy loam (75% sand,
c la y ).
The commercial
fie ld s
were located
10% s i l t
and 15%
along two f l i g h t
lines
covering a 36 km2 area (1.6 x 22.5 km).
A irc ra ft data, which was collected by the NASA C-130 on August
14, 16, and 18, consisted of (1) seven scatterometer frequencies and
polarizations,
( 2 ) three passive microwave radiometer frequencies and
po larizatio n s,
bands,
(3)
eight
v is ib le ,
near-
middle-
and
fa r-in fra re d
(4) Barnes PRT-5 radiometer thermal data, and (5) color in f r a ­
red aerial
photography.
The a ir c r a f t
flew twice at 500 m over each
f li g h t line and once at 1500 m over the general area.
The scatterometer frequencies and polarizations are the same as
the scatterometer sensors at Dalhart.
For each scatterometer,
data
were processed at the same look angles analyzed at Guymon: 5°, 10°,
15°, 20°, 25°, 35°, 40°, 45°.
The passive microwave radiometer frequencies
and polarizations
operating over Dalhart were the same channels operating over Guymon:
L band horizontal and C-band horizontal and v e rtic a l polarizations.
The L band passive microwave radiometer used at Dalhart was not the
same instrument used at Guymon.
The eight
mapper bands)
pm,
channel
1.00-1.30
channels
of
NS001 scanner data
included channel
3:
0.63-0.69
pm, channel
pm,
(simulated
thematic
1: 0.45-0.52 pm, channel 2: 0.52-0.60
channel
4:
0.76-0.90
6 : 1.55-1.75 pm, channel
7:
pm,
channel
5:
2.08-2.35 pm, and
channel 8 : 10.40-12.50 pm. The channels are sim ilar to the proposed
data channels
of the thematic mapper aboard Landsat D.
Channel
7
(M2S) matches well with channel 3(NS001); channel 9 (M2S) matches with
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
51
channel
4
(NS001);
and
channel
11
(M2S)
matches
with
channel
8
(NS001).
The sensors were operating at d iffe re n t times compared to the
Guymon study.
For example,
during the f i r s t
at
Dalhart a ll
scatterometers
were on
run, while at Guymon selected scatterometer sensors
operated at a ll times.
f li g h t lin e and run.
Table 2 l is t s the operating sensors fo r each
Field averages were determined for each f ie l d .
Again, f i e l d averages, of the sensor data were deleted from the data
set when the a ir c r a f t has excessive ro ll
(greater than 3.5°)
and/or
d r i f t (greater than 9 ° ).
The ground data consisted only of soil moisture samples, biomass
data, and photographs of crops.
The soil moisture sampling scheme was
sim ilar to Guymon except for minor modification of the depth intervals
and time of sampling.
F i r s t , the 5-9 and 9-15 cm sampling depths were
combined into a 5-15 sampling depth.
intensively on each f l i g h t day.
every other
day,
on the same
day (8 /1 6 /8 0 ).
Second, fie ld s were sampled less
And f i n a l l y , each f i e l d was sampled
rather thanevery th ird day.
Two flig h ts were flown
The rest of the soil
moisture sampling
scheme was sim ilar to the Guymon study.
Biomass samples were collected within each soil moisture sampling
f i e l d along
the f l i g h t lines in addition to several a l f a l f a and sor­
ghum fields
just
south of the f li g h t
are shown in Figure 7c, d and e.
lines. The sampling locations
Samples were collected from a i m 2
area representative of biomass conditions in the f i e l d .
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
52
TABLE 2.
Operating Sensors for the Dalhart, Texas Study
Date
Line
Run
8/14/80
11
1
scatterometers, NS001, PRT-5, color IR
photos
12
1
scatterometers, NS001, PRT-5, color IR
photos
11
2
passive microwave, NS001, PRT-5, color
photos
IR
12
2
passive microwave, NS001, PRT-5, color
photos
IR
13
1
NS001, PRT-5, and color IR photos
8/16/80
(2
flig h ts )
and
8/18/80
11
1
12
1
Operating Sensors
passive microwave, NS001, PRT-5, color IR
photos
passive microwave, NS001, PRT-5, color
IR photos
11
2
scatterometers, NS001, PRT-5, color
photos
IR
12
2
scatterometers, NS001, PRT-5, color
photos
IR
13
1
NS001, PRT-5, and color IR photos
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
53
Scatterometer Processing
Scatterometer data were collected aboard the NASA C-130 in analog
form on a 14-track tape.
Copies of the tape were la te r sent to Texas
A5M University/Remote Sensing Center for processing, which consisted
of two phases (Figure 8 ) .
log data to d ig ita l
magnetic tapes.
software which
look angle at
The i n i t i a l
processing converted the ana­
values and copied the d ig ita l
data onto 9-track
The second phase processed the d ig ita l
calculated the scattering
given time in te rv a ls .
co e ffic ien t
data using
(<j°)
for each
Data were processed so that a
cell size roughly had a length of 25 m for K band, 38 m for C band, 50
m fo r L band, and 75 m for P band.
cribed by Claussen et a l .
over effe c ts
from the
The processing software was des­
(1979) and Clark and Newton (1979).
lik e -p o la rize d
Cross­
data to the cross-polarized L
band data were removed using a technique described by Blanchard and
The is (1981).
The
cross-over
e ffe ct
is
due
to
the
in a b ility
to
construct
receivers which detect microwave energy in a single p o larizatio n .
In
a c t u a lit y , a single polarized transmitter emits energy in one p o la r i­
zation when upon interacting with the surface is further modified and
is received in two polarizations, thus influencing the cross- as well
as the lik e -p o la riz e d data.
Blanchard and Theis (1981) modeled the
effect of the signal impurity on the cross-polarized data and e ffe c ­
t iv e ly calculated a correction factor for the small look angles.
A fter processing scatterometer data, f i e l d
start and stop times
were determined for each frequency and polarization from lin e plots of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
54
PHASE I
Analog Radar
Data
14 TRK'
PCM TAPE
FORMAT AND
OUTPUT
A/D
LISTING
9 TRK
CCT
S '
PHASE I I
9 TRK
CCT
D igital
Radar
Data
PROGRAM
SCATTER
LISTING
Output
Products
CARD
9 TRK
CCT
FIG.
8
Scatterometer data processing procedure.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
55
a 0 versus
time,
and aerial
photographs.
Times
were
adjusted
by
shifting the start/stop times at least 0.5 seconds toward the f i e l d
center to insure f u ll
scatterometer coverage within the f ie l d .
The
fin a l s tart and stop times defined the f ie l d boundary and were used in
determining f ie ld averages for each frequency, polarization, and look
angle.
Time frames during excessive a i r c r a f t
greater than 3 .5 °;
d rift
greater than 9°)
ro ll
and d r i f t
(r o ll
were noted and data from
affected look angles were deleted from further analysis.
No known technique or mechanism was available to calib ra te a ll of
the scatterometers.
Consequently, any temporal variation
assumed to
eith er
indicate
soil
moisture,
roughness,
in
a° was
or vegetation
changes.
NS001/MZS Processing
The data was processed onto 9-track tapes at NASA/Johnson Space
Center.
Included
with
consisting of d ig ita l
within
d ig ita l
the sensor.
the
surface
data
were
calibration
data
counts from looks at constant radiance targets
The calibration
counts to radiance.
data were then used to convert
To minimize processing costs, only data
from the f i r s t runs were processed.
Since radiance is a function of the solar angle, a correction
factor was needed before comparing crop radiance differences.
All the
Dalhart data were normalized to August 18—the day with the smallest
solar zenith
angle;
angle conditions.
Guymon data were adjusted to August 11 zenith
The correction factor used was
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
56
R
Ri
— 2—
=
COS
where
R-j
and
Rc
are
the
(14)
0
non-normal ized
and
normalized
radiance
values, respectively, and 6 is the solar zenith angle.
Passive Microwave Processing
The raw analog data collected aboard the a ir c a r f t were converted
to d ig ita l
Fight
uncorrected brightness temperatures at NASA/Goddard Space
Center
(GSFC).
Corrected
brightness
temperatures
(Tb )
were
calculated from an equation developed at NASA/JSC ( O 'N e ill, 1981):
, r
r2(T ) ( L )
L
TB ■ t | V ^ >
-|
- - 7 7 7 7 -
-
- * Tr ]
<15>
where t is the transmittance of the radome, e is the emissivity of the
radome,
Tu is
d ig ita l
counts,
temperature
internal
the
uncorrected
L is
fa c to r,
Tr
parameter for
of the receiver.
brightness temperature based on raw
antenna cable
is
the
each
loss
factor,
Tl
is
radome temperature fa c to r,
an antenna
r 2 is
an
frequency, and TCT is the self-emission
For the Dalhart L band horizontal data, the radome
terms are omitted since the sensor used on these flig h ts was operating
in the open rear door of the a i r c r a f t .
the equation
The various constants used in
were determined from flig h ts
over
Once brightness temperatures were calculated,
homogeneous
areas.
line plots of Tb ver­
sus time were produced and f i e l d s ta rt and stop times were determined
from the plots.
The times defining f ie l d boundaries used for sca tter­
ometer data were also used in calculating
fields
averages for each
frequency and polarization.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
57
CHAPTER IV
ANALYSIS
Techniques
Once f i e l d averages had been calculated for each sensor and soil
moisture depth, the ground and a ir c r a f t data sets were merged.
problem
mentioned
in
the
objectives
and
research
subsection
Each
was
analyzed.
In the f i r s t problem, the major task was to note sensor variables
which responded well to differences in crop type.
Analysis techniques
included a Duncan's multiple range technique, and graphical ana lys is-spectrums and response changes as a function of time
Lohnes, 1971).
lyzed.
Both Dal hart and Guymon spectral
(Cooley and
data sets were ana­
The results consisted of a l i s t of sensor variables which are
sensitive to crop type differences.
From this set,
linear combina­
tions were developed which should enhance crop discrimination sensi­
tiv ity .
The
procedure
to
solve
the
second
problem
used
unsupervised
(based on a minimized distance c r ite r io n ) c la s s ific a tio n techniques to
discriminate crops.
A hierarchical
developed using separation c r ite r io n
techniques.
(tree)
c la s s ific a tio n
emerging from the
system was
unsupervised
Individual spectral bands and combinations, such as TVI,
PVI, and other v is ib le /in fr a r e d and scatterometer combinations, were
analyzed.
The supervised c la s s ific a tio n technique was developed using
August 2 and 17, 1978 and August 14 and 18, 1980, data.
then
tested
on August
5,
8,
11 and 14,
The model was
1978 and August
16,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1980
58
spectral
data.
Guymon and
The unsupervised c la s s ific a tio n
Dalhart
data
sets.
From the
technique used a ll
unsupervised
technique,
t r e e -c la s s ific a tio n models (dendrograms) were developed for the Guymon
and Dalhart
data sets.
same separation
example,
if
The dendrograms were constructed using the
c rite rio n
used in the unsupervised technique.
the separation
c rite rio n
between two clusters
were
For
o°
differences in the L band cross pole data, then th is variable was used
in the dendrogram model to separate groups.
locations were compared and s im ila r itie s
The dendrograms at both
noted, which may be appli­
cable in developing a multi frequency dendrogram c la s s ific a tio n model.
The th ir d problem was solved by developing lin e a r step-wise re­
gression, supervised and unsupervised crop c la s s ific a tio n and biomass
estimation models to see i f microwave data could improve c la s s ific a ­
tion
and biomass estimation
accuracy.
Models
using
only
v is ib le /
infrared data were compared to models which included v is ib le /in fr a r e d
and microwave data.
Any microwave sensor or combination which was
more strongly related to crop type differences or biomass estimation
than
other v is ib le /in fr a r e d
variables
or
combinations
improvement over present techniques using only v is ib le
data.
suggested an
and infrared
The lin e a r step-wise models used spectral data from Guymon and
Dalhart.
The supervised and unsupervised c la s s ific a tio n models were
developed and tested on the same spectral
data set as mentioned for
problem 2.
The fourth problem analyzed the v a r i a b i li t y of the c la s s ific a tio n
and biomass
associated
estimation
the
models
v a r i a b i li t y
developed
with
biomass
in
problems
differences
2 and 3,
and
(phenological
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
59
differences)
or soil moisture differences.
nique was graphical
infrared
analysis
of
a0 versus
responses due to d iffe re n t
moisture regimes.
The basic analysis tech­
look angle and v i s i b le /
growth stages or d iffe re n t soil
The results gave an indication of the model u t i l i t y
under d iffe re n t phenological and moisture regimes.
I f the model out­
put v a r i a b i li t y was too large, the model was adjusted to remove i n f l u ­
encing e ffe c ts .
This physically involves reducing the component v a ri­
ances of soil moisture and roughness, leaving vegetation variance as
the major component of the to ta l
variance.
Care was taken to not
remove variance created by d iffe re n t biophases or stress conditions.
The results from each problem were merged to give an overall view
of c la s s ific a tio n improvements that are possible with combinations of
v is ib le ,
infrared and microwave data,
and sim ilar improvements that
can be made in biomass estimation.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
60
CHAPTER V
RESULTS
With the analysis divided into four problems, the results from
each problem w ill
be discussed separately.
But preceding each prob­
lem, a discussion of biomass and fina l yield conditions is in order.
Guymon Crop Condition
A wide range of growing conditions was evident at Guymon.
Irri­
gated sorghum fie ld s ranged in height from 20 cm to 1 m, and in growth
stage from just emerging (fie ld s 7 and 8 ) to anthesis ( f i e l d IX ).
Two
irrig a te d a l f a lf a fie ld s ( fie ld s 22 and 27) were cut on August 17, the
last measurement day.
of the
bare field s
A lfa lfa height ranged from 15 cm to 60 cm.
(fie ld
2X)
was t i l l e d
extensively
f l i g h t day where furrows were as deep as 30 cm.
on the
One
la s t
Two bare fie ld s were
irrig ated during the experiment (fie ld s 6 and 14).
Most of the other
vegetated field s were also irrig a te d .
Since no biomass or y ie ld
data was collected from Guymon, a l l
biomass data was inferred using present v is ib le /in fra re d combinations,
such as PVI and TVI.
Dalhart Biomass and Crop Yield
The 1980 crop year proved to be a below normal year in crop bio­
mass and y ie ld
due to extremely high temperatures
moisture during c r i t i c a l growth stages (Table 3).
and shortage of
Corn fie ld s were in
the tasseling stage and the m ille t f ie l d was just beginning to enter
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
61
TABLE 3.
Dalhart biomass and crop y ie ld
Field_______
Crop
Type
Wet
Biomass
(g/m3)
Dry
Biomass
(g/m3)
Corn
Yield Height
Popul.
(Kg/Ha)
(m)
(plants/m)
1/2 (Healthy)
Corn
6915.1
1259.8
1/2 (Stressed)
Corn
2005.7
411.1
3/4
Mi 11 et
797.5
120.6
5/6
Pasture
125.3
16.2
7/8
Corn
7891.1
1340.6
5676
2 .1 -2 .4
10
9/10
Corn
7665.3
1280.4
5499
2 .1 -2 .4
7
11/12
Corn
5892.7
1148.6
9245
2 . 1- 2 .4
7
17/18(Wheat)
Stubble
365.2
340.5
-
0.3
VI
Sorghum
642.0
139.8
-
0 .9 - 1 .2
V2
Sorghum
1268.2
305.0
3500
0 .9 - 1 .2
V3
Sorghum
2117.0
387.4
-
1.2
V4
Sorghum
4804.3
844.2
-
2.1
V5
A lf a lf a
945.3
108.7
-
0 . 3 - 0 .6
V6
Sorghum
801.6
173.9
-
0 . 6 - 0 .9
V7
A l f a lf a
218.2
62.8
-
0.15
V8
' A lf a lf a
1202.7
128.3
-
0.9
V9
A lf a lf a
897.7
95.0
-
0.8
V10
A lf a lf a
524.7
54.1
-
0.6
Vll
A lf a lf a
946.5
113.1
-
0 .8
V12
A lf a lf a
556.0
66.7
-
0.6
V13
A1 f a l f a
814.9
115.4
4287
2 .1 -2 .4
6
0
1.8
6
1500
0.3
-
0.05
0.8
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
62
the heading stage
during the experiment
temperatures
40°
near
C,
the yields
period.
were
With maximum a i r
reduced
as
much as
50%
compared to 1979 y ie ld s .
The biomass samples were generally related to final crop y i e l d s - higher biomass indicated higher y ie ld s .
The exception was f i e l d 11/12
where corn y ie ld was the highest, but biomass was th ird highest.
The
discrepancy is lik e ly in the unrepresentative biomass sample.
Problem 1
The easiest method of graphical analysis of crop type differences
was through spectral analysis.
Returns from each spectral channel fo r
each f ie l d were compared and differences attrib u ted to soil moisture,
roughness or vegetation.
Figures
9 through
11.
Several
examples of spectra are
The range of
radiance fo r
the v is ib le
infrared region (bands 1-7) is 0 to 3.0 mw cm-2 st eradi anperature range for the thermal
given in
and
the tem­
(band 8 or 5) and microwave brightness
temperature (BT) is 220° to 325°K. The normalized brightness tempera­
ture
(E)
ranged from 0.70 to
1.0 and the scatterometer response (K
band to P band) for lik e (H) and cross (V) pole data ranges from -60
to 0 db.
The soil
moisture f ie l d averages (SM) ranged from 0 to 25%
by volume for each sampling depth (0-2 cm = A, 2-5 cm = B).
terometer
strong
40°
look
angle
relationship
with
was
a rb itra rily
vegetation
as
selected
The scat­
because of
determined
through
the
other
studies reported in the lit e r a t u r e .
Examples of mature corn
with
sim ilar
surface soil
(fie ld
2) and m ille t
moisture conditions
fie ld s
(fie ld
(approximately
3)
9% by
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
FIG.
9
Spectra for m ille t and corn fie ld s at Dalhart.
[H = C band
horizontal (MFMR),V = C band v e rtic a l pole (MFMR), L = L band
horizontal (MFMR), H = l ik e pole 40°look angle (SCATTS), V =
cross pole 40° look angle (SCATTS), A = 0-2 cm soil moisture
(SM), B = 2-5 cm soil moisture (SM)].
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
64
(V
o
w
C4
SCATTS
CS
w
MFMR
r>
> -i > >X
CD
NS001
C**> O—W
CO
GO
CX>
3 SN 0dS3d Q 3 Z llV n d O N
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
FIG. 10
Spectra fo r bare s o i l , pasture and wheat stubble at D alhart.
[H = C band horizontal (MFMR), V = C band v e rtic a l (MFMR), L
= L band horizontal (MFMR), H = l ik e pole 40° look angle
(SCATTS), V = cross pole 40° look angle (SCATTS), A = 0-2 cm
soil moisture (SM), 8 = 2-5 cm soil moisture (SM)].
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
> >
o
rt —
-i>
cscs
SCATTS
66
MFMR
cs
n
00
NS001
-*>
o
oo
3S NOd S3 a Q 3 Z llV W y O N
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
FIG.
11
Spectra comparing vegetated and non-vegetated f ie ld s at
D alhart.
[H = C band horizontal (MFMR), V = C band v e rtic a l
pole (MFMR), L = L band horizontal (MFMR), H = l i k e pole 40°
look angle (SCATTS), V = cross pole 40° look angle (SCATTS),
A = 0-2 cm soil moisture (SM), B = 2-5 cm soil moisture
(SM)].
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
68
CS
o
MFMR
i>
SCATTS
o.
NS001
CO
QO
o
o
CO
3SN0dS3a a3ZllVWaON
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
69
volume) are i llu s t r a t e d in Figure 9.
The largest difference was in
the C, L, andP band active microwave
L band cross
pole data.
mw cm- 2 steradian- 1 .
6 and 7.
data— as large as 6 db
in the
Band 4 data also showed a difference of 0.3
No NS001 data was collected in the corn in bands
Under wetter conditions in the corn (Field 8) the d i f f e r ­
ence was enhanced in several frequencies and the maximum difference in
return was 15
db in the P band cross pole data.
The difference in
L band cross pole and bands 4 and 5 (NS001) remained
quently,
the major variation
the
the same. Conse­
in o° at the 40° look angle in L band
cross pole data appeared to be caused by vegetation.
Responses from
lik e -p o la riz e d microwave data were not very sensitive to the crop type
differences.
Examples of bare s o il, pasture, and wheat stubble having s im ilar
surface moisture
are
shown
in
Figure
10.
Only minor
differences
occurred in the v is ib le and infrared bands, especially in bands 4 and
6.
Band 6 and 7 data was unavailable for f ie l d 15.
Other bands which
had differences were L band lik e and cross pole and C band cross pole
scatteromater
data.
These
differences
roughness differences between the f ie ld s .
ture fie ld s
are
lik e ly
surface
The wheat stubble and pas­
were smoother than the other t i l l e d
smoother fields
due to
consequently acted as a spectral
bare
f ie l d s .
The
r e fle c to r giving a
lower o° at the 40° look angle.
Comparing the
vegetation f i e l d s ,
response
differences
between
vegetated and non­
several
spectral
regions were s ig n ific a n t
11).
Obvious differences
were in
bands 4,
data.
Possible combinations using these bands may prove to be helpful
5,
(Figure
and 6 of the NS001
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
70
in discriminating vegetation from non-vegetation.
In addition, a l l of
the active microwave channels were able to distinguish vegetative d i f ­
ferences to some degree of success.
The most s ig n ific a n t differences
occurred in the C band and L band o° values—as much as 12 db in the L
band cross pole data.
An interesting
anomaly demonstrating stressed and non-stressed
conditions was evident in corn field s
1 and 2.
Parts of the f ie l d
were stressed as a result of a fau lty irr ig a tio n system which did not
apply adequate amounts of water in several areas through the growing
season.
A black and white aerial
ure 12.
stress.
to ta l
Approximately 30-50% of the f i e l d
was undergoing moisture
The stressed areas essentially had no grain y ie ld ; thus the
y ie ld
infrared
photo of the f i e l d is shown in Fig­
represented y ie ld
of the healthy
spectra showed significant
unhealthy corn
in several
especially s ig n ifican t
bands
differences
(Figure 13).
areas.
between
The v is ib le /
healthy and
The differences were
(0.3 mw cm-2 ster- 1 ) in NS001 channels 4, 5,
and 7, suggesting possible combinations using these bands may indicate
biomass differences or stress conditions.
At Guymon, the crop types were d iff e r e n t — a l f a l f a ,
bare s o i l .
Examples of bare soil
I X ) , and a l f a lf a
(fie ld
sorghum, and
10), mature sorghum ( f i e l d
( f i e l d 4) spectra having sim ilar surface soil mois­
ture conditions are shown in Figure 14.
Reflectance in the v is ib le
and infrared
d iffered sig n ifican tly
between vegetated and non-vege-
tated fie ld s
(as much as 6-10 irw cm-2 ster- 1 ) .
Differences
in the
active microwave, especially L, C and P band were also indicative of
crop types differences.
For example, a difference of 9 db in the L
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
71
FIG.
12
Black and white infrared a erial photo (scale 1:45,000) of
stressed corn fie ld s (fie ld s 1 and 2) at Dalhart. The
healthy areas are dark shaded and the stressed areas are
lig h t shaded.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
DALHART
F ie ld I B
S tressed com
F ie ld 1C H e a lth y com
L
U
co
Z
o
Qu
CO
LU
a:
Q
UJ
N |
s
a:
o
z
.4-
12
3
4
5
6
7
8
NS001
FIG.
13
Spectra comparing healthy and stressed corn a t Dalhart.
microwave comparisons could be made.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
FIG. 14
Spectra comparing a l f a l f a , sorghum and bare soil f ie l d s at
Guymon [H = C band horizontal (MFMR), V = C band v e r t ic a l
(MFMR), L = L band horizontal (MFMR), H = l i k e pole 40° look
angle (SCATTS), V = cross pole 40° look angle (SCATTS), A =
0-2 cm soil moisture (SM), B = 2 -5 cm so il moisture (SM)].
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
74
_ i
o
S C A TTS
> >
H-
M
00
MMS
CN
MFMR
n
O
°o
CNJ
CD
3SN 0dS3d a s z n v w a o N
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
75
and P band lik e pole data was common between sorghum and bare soil or
sorghum and a l f a l f a .
Part of the difference may be due to roughness
v a r i a b i l i t y in the soil surface.
be penetrating
through
Also some microwave frequencies may
the canopy and detecting t i l l a g e
The sorghum responses in f ie l d
IX figure
rows perpendicular to the f li g h t lin e .
a sorghum f ie l d with rows parallel
given in Figure 15.
d irection .
14 were from a f i e l d with
An example o f a response from
to the f l i g h t lin e
( f i e l d 2A) is
The most sig n ifican t differences were in the C
band l ik e pole and L band data— a 5db difference.
band indicated f ie l d 2A had less canopy cover.
affected the return.
The near infrared
Wetter conditions also
For example, the spectra from a wet bare s o il,
f ie l d 14 (Figure 16) was sim ilar to spectra for a dry sorghum f ie l d
(fie ld
2A), especially
quently,
in the scatterometer lik e pole data.
responses which include roughness and soil
Conse­
moisture d i f f e r ­
ences are masking the crop type differences.
Soil moisture differences were removed from the analysis of data
from Clayton, New Mexico since the entire area had been saturated with
a uniform r a in f a ll on a large area of uniform s o ils .
As a result o f
the rains, every f ie l d had approximately the same high soil moisture
content,
thus
leaving
only
active microwave return.
roughness and
vegetation
to
a ffe ct
the
Assuming t i l l a g e practices were s im ilar be­
tween crop types (corn and sorghum), the roughness e ffe c t is also min­
imized, leaving only vegetation effe c ts .
Analysis of the spectra from
four co rn (Cl through C4) and two sorghum f i e l d s , Ml and M2 (Figures
17 and 18)
indicated that scatterometer L and P band lik e and cross
pole data discriminated between corn and sorghum w ell.
Corn tended to
Reproduced with permission of the copyright owner. Further reproduction prohibited w ithout permission.
FIG.
15
Spectra comparing sorghum f ie ld s with rows perpendicular
and p a r a lle l to the f l i g h t l i n e . [H = C band horizontal
(MFMR), V = C band v e r t ic a l (MFMR), L = L band horizontal
(MFMR), H = l i k e pole 40° look angle (SCATTS), V = cross
pole 40° look angle (SCATTS), A = 0-2 cm soil moisture
(SM), B = 2-5 cm soil moisture (SM)].
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Field
IX
Sorghum
77
TOO
BSNOdSBU Q3znvwaow
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
FIG. 16
Spectra comparing wet bare s o i l , and a dry sorghum f i e l d at
Guymon [H = C band horizontal (MFMR), V = C band v e r t ic a l
(MFMR), L = L band horizontal (MFMR), H = l i k e pole 40° look
angle (SCATTS), V = cross pole 40° look angle (SCATTS), A =
0-2 cm soil moisture (SM), B = 2-5 cm soil moisture (SM)].
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
79
m<
CM
o
hf—
O
(N
CN
s
u.
s
CD
oo
CM
9SN0dS9a aaznvwaoM
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
80
1H
1. 0 -
1V23H
23V
IV
3V
2V
.8 -
12 H
2V
IV
NORMALIZED RESPONSE
3H
3V
12H
3H
12V
3V
.4 -
1
.2 -
F i e ld C 2 Com
2
F i e ld C l C o m
3
F i e ld M l Sorghum
I
K
"i- - - - - - - r
c
L
SCATTS
FIG.
17
Spectra comparing corn and sorghum at Clayton. No passive
microwave or v is ib le /in fr a r e d data was a v a ila b le .
[H = l i k e
pole 40° look angle (SCATTS), V = cross pole 40° look angle
(SCATTS)]
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
81
2H
1H
1.0 “ 1
3H
ft
3V
2V
IV
3V
2H
lH
IV
2V
NORMALIZED RESPONSE
3H
3V
.6 -
1H
IV
3H
23 V
.4 GUYMON
.2 -
1
F i e id C 4
C om
2
F ie ld C 3
Com
3
F ie ld M2
Sorghum
l - - - - - - - - 1- - - - - - r
K
FIG.
18
C
L
SCATTS
Spectra comparing corn and sorghum at Clayton. N'o passive
microwave or v is ib le /in fr a r e d data was a v a ila b le .
[H = l i k e
pole 40° look angle (SCATTS), V = cross pole 40° look angle
(SCATTS)]
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
82
have higher
returns
in the
returns from sorghum fie ld s .
L and P band data as compared to
the
Other frequencies had smaller or no res­
ponse difference between corn and sorghum.
S ta tis tic a l
analysis of the Dalhart and Guymon data sets, using
Duncan's M ultiple Range Technique confirmed results noted in graphical
analysis.
The channels which discriminated the crops at Dalhart best
were the K, C and L band active microwave data at look angles from 40°
and 45° o f f nadir (Table 4 ) .
The visib le and infrared bands were able
to discriminate between vegetated and non-vegetated fie ld s very w e ll,
but not differences within the vegetated f ie l d s .
At Guymon, the same
active microwave frequencies did the best job of discriminating crops
(Table 5).
Fields and crops with higher biomass had the higher res­
ponse, while field s with l i t t l e or no biomass had the lower response.
However, roughness also played an important role as indicated by d i f ­
ferences
rows.
between
sorghum fie ld s
having
perpendicular
and p a ra lle l
The roughness effect was reduced in the cross-polarized data,
thus suggesting the L band cross pole and C band cross pole active
microwave data as possibly the best microwave frequencies and p o la r i­
zations to use.
Another means of demonstrating the effe ct of vegetation in the
active microwave region was analyzing line plots of the data (o° as a
function of tim e).
An example of three
fields
having roughly the
same surface soil moisture is given in Figures 19 and 20.
near (10°) and far
f ie ld s
V6,
(40°)
look angle were plotted .
1 and 19, on 8/16/80 at Dalhart,
Texas.
Data from a
The area covered
The crop types
represented included sorghum ( f i e l d V6), corn, ( f i e l d 1) and bare soil
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
TABLE 4.
Results of Duncan's M ultiple Range Test fo r Dalhart active
microwave data
40° look angle
Crop
K band l ik e pole
Corn
nl i i c
Weeds and Bare Soil
Bare Soil
Pasture
Wheat Stubble
45° look angle
Mean
“7.1
-9.1
-10.9
-11.3
-14.0
-14.6
Crop
a*
b
c
c
d
d
L band l ik e pole
Corn
Weeds and Bare Soil
M ille t
Bare Soil
Pasture
Wheat Stubble
-22.4
-29.8
-30.6
-30.7
-34.7
-36.2
a
a
b
b
c
c
-28.9
-37.1
-39.5
-39.7
-44.2
-44.2
a
b
c
c
d
d
C band lik e pole
Corn
Mi 1le t
Weeds and Bare Soil
Bare Soil
Pasture
Wheat Stubble
Corn
M ille t
Weeds and Bare Soil
Bare Soil
Pasture
Wheat Stubble
-7.1
- 8 .8
-10.6
-10.9
-13.6
-14.3
a
b
c
c
d
d
-23.1
-30.9
-31.9
-32.9
-36.8
-37.3
a
b
b
b
c
c
-28.6
-37.2
-39.3
-41.2
-44.6
-48.8
a
b
be
c
d
d
L band l i k e pole
L band cross pole________
Corn
Mi 1 le t
Bare Soil
Weeds and Bare Soil
Wheat Stubble
Pasture
Mean
K band l ik e pole
Corn
Weeds and Bare Soil
Mi 11 et
Bare Soil
Pasture
Wheat Stubble
L band cross pole
Corn
M ille t
Weeds and Bare Soil
Bare Soil
Pasture
Wheat Stubble
C band l i k e pole
-2.1
-4.2
-7.0
-7.3
-11.1
-12.4
a
a b
b c
b c
c
c
Corn
M il le t
Weeds and Bare Soil
Bare Soil
Pasture
Wheat Stubble
-5 .8
-7 .5
-10.4
-11.5
-14.8
-17.1
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
a
a b
b c
c
c d
d
TABLE 4.
(Continued)
40° Look Angle
45° Look Angle
_______ C band cross pole
Corn
Mi 11 et
Weeds and Bare Soil
Wheat Stubble
Bare Soil
Pasture
P band l i k e pole
Corn
Weeds and Bare Soil
Wheat Stubble
Mi 11et
Bare Soil
Pasture
C band cross pole
-8.5
-14.4
-17.3
-19.4
-19.7
-21.2
a
b
b c
b c
c
c
Mean
-28.7
-35.1
-35.3
-36.2
-37.3
-37.5
a
b
b
b
b
b
P band cross pole
Corn
Weeds and Bare Soil
Wheat Stubble
Bare Soil
Mi 11 et
Pasture
Corn
Mi 11et
Weeds and Bare Soil
Bare Soil
Wheat Stubble
Pasture
P band l i k e pole
Corn
Weeds and Bare Soil
Wheat Stubble
Mi 11 et
Bare Soil
Pasture
-10.0
-15.6
-18.2
-20.4
-20.8
-21.2
a
b
b
b
b
b
Mean
-28.9
-36.3
-37.3
-37.6
-38.0
-38.5
a
b
b
b
b
b
-43.9
-52.9
-54.2
-54.2
-54.8
-55.1
a
b
b
b
b
b
P band cross pole
-43.9 a
-47.6
-52.7
-52.8
-52.9
-54.9 c
Corn
Weeds and Bare Soil
Bare Soil
Mi 11 et
Wheat Stubble
Pasture
*The treatment mean followed by the same l e t t e r in each column are not
s ig n ific a n tly d iffe re n t at the 5% probability level of Duncan's
M ultiple Range Test.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
85
TABLE
Crop
5.
Results of Duncan's Multiple Range Test for Guymon active
microwave data
40° Look Angle
Mean
K band lik e pole
Sorghum(perp.
rows)
Sorghum(paral. rows)
Bare Soil
A lf a lf a
-7.1
-9 .5
-12.1
-12.1
a
b
c
c
-9 .3
-18.1
-18.2
-20.5
a
b
b
b
-19.1
-21.5
-27.1
-27.7
a
a
b
b
-8 .2
-12.5
-14.2
-15.2
a
b
b
b
a
b
c
c
Sorghum (perp. rows)
Sorghum (paral. rows)
Bare Soil
A lfa lfa
-11.9
-19.2
-21.1
-21.9
a
b
b
b
Sorghum (perp. rows)
Sorghum (paral. rows).
A lf a lf a
Bare Soil
-20.2
-22.4
-27.9
-28.5
a
a
b
b
Sorghum (perp. rows)
Sorghum (paral. rows)
A lf a lf a
Bare Soil
-10.3
-13.7
-15.4
-16.3
a
b
b
b
-19.5
-22.0
-23.7
-28.7
a
a
b
c
-23.7
-30.3
-32.0
-35.1
a
b
b
c
C band cross pole
-17.2
-19.6
-22.6
-26.9
a
a b
b
c
P band l ik e pole
Sorghum (perp. rows)
Bare Soil
Sorghum (paral. rows)
A lf a lf a
-7.7
-9.7
-12.3
-12.5
C band l ik e pole
C band cross pole
Sorghum(perp. rows)
Sorghum(paral. rows)
A l f a lf a
Bare Soil
Sorghum (perp. rows)
Sorghum (paral. rows)
Bare Soil
A lfa lfa
L band cross pole
C band lik e pole
Sorghum(perp.
rows)
Sorghum(paral. rows)
A lf a lf a
Bare Soil
Mean
L band l ik e pole
L band cross pole
Sorghum(perp. rows)
Sorghum(paral. rows)
Bare Soil
A lf a lf a
45° Look Angle
K band l ik e pole
L band lik e pole
Sorghum(perp.
rows)
Sorghum(paral. rows)
Bare Soil
A lf a lf a
Crop
Sorghum (perp. rows)
Sorghum (paral. rows)
A lf a lf a
Bare Soil
P band lik e pole
-27.8
-31.4
-31.5
-35.6
a
b
b
c
Sorghum (perp. rows)
Bare Soil
Sorghum (paral. rows)
A lfa lfa
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
86
TABLE 5.
(Continued)
P band cross pole
Sorghum (perp. rows)
Sorghum (paral. rows)
A lfa lfa
Bare Soil
P band cross pole
-37.2
-38.5
-46.5
-47.4
a
a
b
b
Sorghum (perp. rows)
Sorghum (paral. rows)
Bare Soil
A lfa lfa
-34.3
-37.4
-45.6
-46.9
a
a
b
b
*The treatment means followed by the same l e t t e r in each column are not
s ig n ific a n tly d iffe r e n t at the 5% probability level of Duncan's
Multiple Range Test.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
DALHART
K BAND LIKE POLE DATA
+ 10 DEGREES
o40DEGREES.
I
CD
O
O
£ -1 0
CC
UJ
rvi
CORN
M
BARE
soil!
SORGHUM
i
CD
SORGHUM
-30
CO
CO
-20
14:46:30
14:46:40
TIME (GMT)
14:46:50
14:47:00
| C^ J
I
I
BARE]
SOIL
I
20
14:46:20
14:46:30
14:46:40
14:46:50
14:47:00
TIME (GMT)
DALHART
C BAND LIKE POLE DATA
+ 10 DEGREES
o 40 DEGREES
30
10
DALHART
P BAND LIKE POLE DATA
+ 10 DEGREES
*>40 DEGREES
m
o
o
o
as 10
M
<
0
CD
P5
|
-40
14:46:20
CD
DALHART
L BAND LIKE POLE DATA
+10 DEGREES
o 40 DEGREES
g-10
I
I CORN
SORGHUM
UJ
BARE
SOIL
fs l
<
CD
-2 0
-10
CORN
SORGHUM
in
-30
-20
I
-40
14:46:20
14:46:30
14:46:40
TIME (GMT)
FIG. 19
14:46:50
14:47:00
14:46:20
14:46:30
14:46:40
TIME (GMT)
I
14:46:50
14:47:00
Line plots ( a 0 vs time) fo r a l l l i k e polarized scatterometer data at 10° and 40° o f f nadir.
00
88
30
C BAND CROSS POLE DATA
+ 1 0 DEGREES
o 40 DEGREES
20
i
co
s 10 ------- ------------ j
^
f
o
CE
w
M
<
S
d
C5
l
!
^
|
.
CORN j
n
0
Ibar e
SORGHUM
| SOIL!
-1 0
i
i
i
i
-2 0
i
i
-3 0
14:46:20
i
i
i
14:46:30
!
!
14:46:40
14:46:50
TIM E (GMT)
14:47:00
14:47:10
DALHART
L BAND CROSS POLE DATA
+ 10 DEGREES
° 40 DEG REES,
CORN
SORGHUM
< -3 0
BARE
i
14:46:20
14:46:30
14:46:40
14:46:50
TIM E (GMT)
i
14:47:C
14:47:10
14:47:00
14:47:10
0
DALHART
P BAND CROSS POLE DATA
+ 10 DEGREES
, o 4 0 DEGREES
-1 0
CD
O
-2 0
V
I
. --- ----- I
'T *'*' 1
+
I
< -3 0
d
en
SORGHUM
I
1
1 CORN
j
j
1
1 /u A
I
1
I
j
IBARE
SOIL
-4 0
1
-5 0
14:46:20
FIG 20.
i
t
i
i
i
i
i
i
14:46:30
14:46:40
14:46:50
TIM E(GM T)
Line plots (o° vs time) for a l l cross polarized scatterometer
data at 10° and 40° o f f nadir.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
89
(fie ld
19).
Crop
type
differences were enhanced at
angles, especially in the C, L and P band data.
the
fa r
look
The responses from
the near look angles tended to be f a i r l y stable along the f l i g h t l i n e ,
especially at the lower frequencies.
Summarizing,
in
addition
to several
v is i b le /i n f r a r e d
channels,
active microwave frequencies (C, L and P band) are sensitive to crop
type differences between selected crop pairs.
For instance, L band
and P band discriminated between sorghum and
corn, while C band did
not.
and a l f a l f a while K, L
C band discriminated between bare soil
and P bands did not discriminate between this
criminated between corn and bare s o il.
Soil
p a ir .
All bands dis­
moisture and roughness
had an e ffe c t on the active microwave responses, but the vegetation
e ffe c t
generally
predominated at the fa r look
angles
(greater than
3 5 °).
Problem 2
To develop the proper combination for analyzing crop type d i f f e r ­
ences
in a t r e e -c la s s ific a t io n
clustering
routine was used.
The routine
was based on a cluster
clu s te r.
By going through the same classifying c r i t e r i a
individual
individual
clusters
through several
were
Euclidean
(unsupervised)
of
routine,
minimum
a hierarchical
c r ite r io n
the
a
model,
channels
detected.
distance
or
from
the
combinations
By
following
mean
of
the
used within
which
separated
this
technique
ite r a t io n s , a dendrogram ( tr e e - c la s s if ic a tio n system)
using v is ib le , in fra re d , and microwave data was developed.
Data from
crop discriminating scatterometer frequencies and polarizations at 40°
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
90
look angles were included with the
v is ib le /in fr a r e d
data
(omitting
thermal) at Guymon and Dalhart.
In addition , a dendrogram was devel­
oped from the Dalhart spectral
data set using the scatterometer 40°
look angle and only bands 2, 3, and 4 from the NS001 data.
This ana­
lysis was done to allow unbiased comparisons of c la s s ific a tio n accur­
acy between the Dalhart and Guymon data sets.
Active microwave data
from the 40° look angle was used because the data from th is look angle
was most sensitive to crop type differences (results from the previous
problem).
Results from the Dalhart dendrogram using the active microwave
bands and NS001 bands 2, 3 and 4 indicated that C and L band cross
pole data can classify reasonably well without v is ib le and n e a r-in fra ­
red information (Figure 21).
The largest error was separating wheat
stubble and pasture from bare s o i l .
Allowing these three groups to be
c la s s ifie d the same, the overall accuracy was 78%.
The f i r s t separa­
tion c r ite rio n used differences in the sum of L band and C band cross
pole 40° look angle data to separate corn, sorghum, m ille t and weeds
from pasture,
bare s o i l ,
and wheat
stubble.
The second c rite rio n
again used differences in the sum of L band and C band cross pole 40°
look angle data to separate corn and sorghum (class 1) from m ille t and
weeds (class 2 ) .
The th ird c rite rio n used C band cross pole 40° look
angle data to separate pasture, wheat stubble and bare soil
from other pasture and bare s o il.
"Then the last c rite rio n used was L
band cross pole data to separate bare soil
bare soil
(class 5).
(class 3)
(class 4) from weeds and
The difference between the bare fie ld s in class
4 and 5 was the class 5 bare fields included some weeds while class 4
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
91
L band cross pole + C band cross pole (4 0 ° look a n g le )
>-46 db
<-46 db
C band cross pole (40° look angle)
<-16
> -16 db
db
L band cross poleQ+
C band cross pole (40 look angle)
<-37 db
>-37 db
L band cross pole
(40 look angle)
Class 1
21 Corn
3 Sorghum
Class 2
3 Corn
2 Sorghum
5 M ille t
5 Weeds
1 Bare Soil
Class 3
2 Pasture
5 Wheat
Stubble
8 Bare
Soil
>-37 db
Class 4
11 Bare
Soil
FIG.
21
<-37 db
Class 5
1 M il le t
4 Weeds &
Bare
Soil
3 Bare
Soil
4 Pasture
Dendrogram (tre e -c la s s ific a tio n ) model using NS001 bands 2,
3, and 4, and C and L band cross pole Dalhart data (accuracy
7 8 % ) .
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
92
bare fie ld s did not.
Consequently, responses in class 5 appear to be
sensitive to low biomass levels.
Using a l l
of the NS001 with active microwave data, the accuracy
improved to 84% as more information was gathered in NS001 bands 3, 4,
5 and 6.
The dendrogram was d ifferen t in that the f i r s t and last two
c rite rio n
used L band cross pole and NS001 bands 3,
4,
5,
and 6,
respectively (Figure 22).
In spite of the d iffe re n t crop types and v is ib le /in fr a r e d bands,
a sim ilar dendrogram to the one using a ll NS001 data was developed at
Guymon (Figure 23).
The f i r s t criterio n
data as Dalhart—-L band cross pole.
sorghum from other crops.
level used the same type of
These steps separated corn and
The next c rite rio n used differences in the
sum of C and L band cross pole data.
The last two steps used M2S band
9 data to separate vegetation from bare s o i l .
the model was 70%.
fields
The overall accuracy of
One bare f i e l d , 10, was frequently c la s s ifie d with
having vegetation.
The reason for the m isclassification was
due to the presence of weeds within the f ie l d late in the experiment.
The s im ila r ity between the two models is s trik in g .
biomass were
separated
from other
field s
vegetation was separated from bare soil
data.
Fields with high
using microwave data and
using v is ib le
and infrared
The s im ila r it y w ill be discussed further in the next section.
A problem arose when data sets from both Guymon and Dalhart were
combined.
Due to the fact the visible and infrared regions did not
match and no calibration of the scatterometer data was ava ilab le, no
dendrogram for the combined data set was developed.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
93
L band cross pole (40
look a n g le )
<-33 db
>-33 db
C band cross pole +
L band cross pole (40 look angle)
Class
32 Corn
5 Sorghum
(NS001 band 6 NS001 band 5)
(NS001
NS001
-2
<1.0 mw cm
mw cm
-1
st
mw
cm
st
LI
Class
6 M il l e t
2 Weeds
9 Sorqhum
FIG.
22
Class 3
5 Weeds &
Bare Soil
9 Bare Soil
3 Weeds
Class
Class 5
8 Pasture
Sorghum
8 Wheat
Stubble
16 Bare Soil
1 Weeds & Bare
Soil
Dendrogram (tre e -c la s s ific a tio n ) model using a l l NS001 bands
C and L band cross pole Dalhart data (accuracy 84%).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
94
L band cross pole (4 0 ° look a n g le )
<-21 db
>-21 db
C band cross pole +
L band cross pole (40 look angle)
>-42 db
Class 1
4 Corn
38 Sorghum
.1 A lf a lf a
MMS band 9
<0.45 mw
cm
Class 2
9 Bare
Soi 1
FIG.
23
<-42 db
MMS band 9
>0.45 mw
cm
Class 3
5 A lf a lf a
8 Sorghum
2 Bare
Soil
<0.45 mw
cm
Class 4
17 Bare
Soil
>0.45 mw
cm- ^
Class 5
9 A l f a lf a
11 Sorghum
8 Bare
Soil
Dendrogram (tr e e -c la s s ific a tio n ) model using M2S bands 4,
7, 8 and 9, C and L band cross pole Guymon data (accuracy
70%).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
95
Problem 3
This
problem deals
estimations.
with
both
crop
c la s s ific a tio n
and biomass
One technique used to determine the u t i l i t y of microwave
data in c lassifica tio n was to make a comparison between unsupervised
c la s s ific a tio n result accuracies using v is ib le , infrared and microwave
data and accuracies using only visib le and infrared data.
As men­
tioned in the previous subsection, cluster analysis using microwave,
v is ib le , and infrared data had c la s s ific a tio n accuracies equal
greater than 70%.
to or
Using only v is ib le /in fr a re d data, the c la s s ific a ­
tion accuracies decreased to 65% at Guymon and 78% at Dalhart.
The
t r e e -c la s s ific a tio n system using v is ib le and infrared data at Dalhart
and Guymon are given in Figures 24 and 25,
misclassification
field s
being
using v is ib le
c lassified
as
respectively.
The major
and infrared data were high biomass
one
group.
For
instance,
at
Guymon
twenty-one observations of a l f a l f a and twenty-two observations of sor­
ghum field s at
Consequently,
proved
that
d ifferen t
result
inclusion
biophases were clas s ifie d
comparisons
of
from
microwave
the
data
into one group.
unsupervised
enhanced
technique
c la s s ific a tio n
accuracy.
Supervised
class ific a tio n
(discriminant
analysis)
results
indicated microwave data improved c la s s ific a tio n accuracy.
also
The con­
tingency table results from classifying fie ld s on August 16 using only
NS001 data from August 14 and 18 as the training c l a s s if ie r is given
in Table 6a.
The overall accuracy was 73%.
By including K band l ik e
pole and L band cross pole data the accuracy increased to 92% (Table
6b).
To make unbiased comparisons with the Guymon spectral data sets,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
96
NS001 band 6
>1.3 mw cm
<1.3 mw cm
NS001 band 6 +
NS001 band 4
-2
<3.1 mw cm"*
>3.1 mw cm
Class 1
24 Corn
2 Sorghum
NS001 band 4
>0.9 mw cm
<0.9 mw
cm
Class 2
6 M il le t
3 Weeds
1 Bare Soil
& Weeds
4 Sorghum
FIG.
24
Class 3
5 Pasture
6 Wheat
Stubble
6 Bare
Soi 1
NS001 brnd 6
<2.7 mw
<2.7 mw
cm
Class 4
1 Pasture
12 Bare
Soil
5 Bare Soil
& Weeds
3 Weeds
Dendrogram (tre e -c la s s ific a tio n ) model using a l l
at Dalhart.
(78% accuracy)
cm
Class 5
5 Bare
Soi 1
NS001 data
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
97
MMS band 9 + MMS band 7
-2
<0.6 mw cm
st
-1
-2
>0.6 mw cm
st
-1
MMS band 9
MMS band 9
<0.9 mw cm"
<0.35 mw cm"
s t-'
Class 1
25 Bare
Soil
FIG.
25
>0.35 mw
-2
cm
-1
st
Class 2
23 Bare
Soi 1
41 Sorghum
(28 para)
(13 perp)
3 A lfa lfa
>0.9 mw cm"
L-1
s t-'
MMS band 7 +
MMS band 4
>0.2 mw
9 Bare
Soil
3 Sorghum
(para)
<0.2 mw
Class 5
1 Sorghum
(perp)
Class
21 A lfa lfa
22 Sorghum
(perp)
Dendrogram ( tr e e -c la s s ific a tio n ) model using M2S band 4 7
8 and 9 data at Guymon (65% accuracy).
* *
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
98
TABLE 6.
Dalhart discriminant analysis results using (a) a ll NS001
channels and (b) a ll NS001 channels plus K band lik e pole and
L band cross pole (40° look angle) data from August 14 and 18
as a tra in in g c l a s s i f ie r . The results are from August 16
te s tin g of the model.
(a)
Number of Observations Classified into Crop Types:
From Crop Types:
Corn
Corn
Bare Soil
Wheat Stubble
Weeds and Bare
Soil
Pasture
Mil l e t
Weeds
Bare
Soil
Wheat
Stubble
Weeds and
Bare Soil
Pasture
Mi 11 et Weeds
16
0
0
0
16
4
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
3
4
0
0
0
0
0
0
0
0
0
0
0
0
0
2
0
0
4
2
0
0
0
0
*Accuracy of 73%
(b )
Number of Observations Classified into Crop Types:
From Crop Types:
Corn
Corn
Bare Soil
Weeds and Bare
Soi 1
Pasture
Mil le t
Wheat Stubble
Weeds
Bare
Soil
Weeds and
Bare Soil
Pasture
Mi 1 le t
Wheat
Stubble Weeds
16
0
0
11
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
4
0
0
0
0
0
0
0
0
0
0
4
0
0
0
0
0
4
0
1
0
0
0
4
0
0
0
0
0
0
*Accuracy of 92%
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
99
TABLE 7.
Dalhart discriminant analysis using (a) NS001 channels 2, 3,
and 4 and (b) NSOOl channels 2, 3 and 4 and K band lik e pole
and L band cross pole data.
Contingency table results from
the model tested on August 16 spectral data.
(a)
Number of Observations C lassified into Crop Types:
From Crop Types:
Corn
Corn
Bare Soil
Weeds and Bare
Soil
Pasture
Mi 11 et
Weeds
Wheat Stubble
Bare
Soil
Weeds and
Bare Soi1
Pasture
Mi 11et Weeds
Wheat
Stubble
16
0
0
12
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
4
3
0
0
0
0
0
3
0
0
0
0
0
0
0
0
1
1
4
4
0
0
0
0
0
0
♦Accuracy of 81%
(a)
Number of Observations Classified into Crop Types:
From Crop Types:
Corn
Corn
Bare Soil
Weeds and Bare
Soil
Pasture
M ille t
Weeds
Wheat Stubble
Sorghum
Bare
Soi 1
Weeds and
Bare Soil
Pasture
Mi 11 et Weeds
Wheat
Stubble
15
0
0
12
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
3
3
0
0
0
4.
1
1
0
0
1
0
0
0
4
0
0
2
0
0
0
4
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
♦Accuracy of 84%
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
TOO
NSOOl bands 2, 3 and 4 were analyzed.
Following the same techniques,
the August 16 c l a s s if ie r accuracy was 81% (Table 7a).
and L band cross pole active microwave data,
only s lig h tly to 84% (Table 7b).
By including K
the accuracy improved
No known reason explained the d is ­
crepancy between results using a l l or parts of the NSOOl data.
At Guymon, spectral data from August 2 and 17 were used as inputs
into the tra in in g c l a s s i f i e r , and the c la s s if ie r was tested on August
5, 8, 11 and 14 spectral data.
Using only M2S visib le and infrared
data, the c la s s ific a tio n accuracy was 88% (Table 8a).
By including K
band lik e pole and L band cross pole data the accuracy remained the
same 88% (Table 8b).
Consequently, supervised c la s s ific a tio n results
using the Dalhart and Guymon spectral data sets indicated inclusion of
microwave data with v is ib le /in fr a re d data maintained or improved clas­
s ific a tio n accuracy compared to using only v is ib le and near infrared
data.
Using step-wide regression techniques to determine the u t i l i t y of
microwave data, an increase in the c o e ffic ien t of determination using
microwave data is apparent (Tables 9 and 10).
At Guymon and Dalhart,
the C and L band active microwave data was especially sensitive to
crop types differences.
Biomass estimation was the second portion of the problem and the
results from the previous section have already indicated that combina­
tions of red and near-infrared data may help in estimating biomass.
Two such combinations described previously are the perpendicular vege­
tatio n index (PVI) and the transformed vegetation index (T V I).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
101
TABLE
8.
Discriminant Analysis of Guymon v is ib le /in fra re d data using
August 2 and 17 data as the training c la s s i f ie r .
Results
from c la s s ific a tio n of August 5, 8, 11, and 14 data.
(a)
Number of Observations Classified into Crop Types:
From Crop Types:
A1falfa
Ba re
Parallel Row
Sorghum
Perpendicular
Sorghum
Perp. Sorghum
A lfa lfa
Bare
Paral. Sorghum
12
0
0
32
3
4
1
1
1
1
18
1
1
0
2
21
*Accuracy is 88% (assuming p a ra lle l sorghum and perpendicular sorghum
are one group)
(b)
Number of Observations Classified into Crop Types:
From Crop Types:
A1falfa
Bare
Parallel Row
Sorghum
Perpendicular Row
Sorghum
A1falfa
Bare
Paral. Sorghum
9
0
0
23
2
2
1
2
1
1
8
6
0
0
0
19
Perp. Sorghum
*Accuracy is 88% (assuming p arallel sorghum and perpendicular sorghum
are one group)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
102
TABLE
9.
Dalhart stepwise c la s s ific a tio n regression equations using
(a) a ll NSOOl band (Ch) data and (b) a ll NSOOl data plus
scatterometer data (40° look angle) [Crop Type:
10 = corn,
8 = sorghum, 6 = weeds, 4 = bare soil and weeds, 3 =
pasture, 2 = wheat stubble, 1 = bare s o i l ] .
R2
Crop Type = -(Ch3*1.99)+(Ch4*0.71)+3.03
0.94
Crop Type = (Ch2*1.78)-(Ch3*3.60)+(Ch4*0.60)+3.26
0.95
Crop Type = (Ch2*1.90)-(Ch3*3.66)+(Ch4*0.63)-(Ch5*0.07)
+3.26
0.95
Crop Type = (Ch2*1.87)-(Ch3*3.69)+(Ch4*0.60)-(Ch6*0.05)
+(Ch7*0.11)+3.31
0.95
Crop Type = -(C hl*0.04)+(C h2*1.87)- ( Ch3*3.67)+(Ch4*0.60)
-(Ch6*0.05)+(Ch7*0.12)+3.35
0.95
Crop Type = -(Ch3*2.07)+(Ch4*0.65)+3.85
0.95
Crop Type = (Ch2*2.03)-(Ch3*3.90)+(Ch4*0.54)+3.83
0.96
Crop Type = -(Ch3*2.35)+(Ch4*0.63)-(L band cross pole
*0.13)+(C band lik e pole*0.13)+0.88
0.96
Crop Type = (Ch2*2.38)-(Ch3*4.34)+(Ch4*0.55)+(L band lik e
pole*0.15)-(L band cross pole*0.15)+2.39
0.96
Crop Type = (Ch2*1.73)-(Ch3*3.83)+(Ch4*0.55)+(L band lik e
pole*0. 1 4 ) - (L band cross pole*0.19)+(C band
l i k e pole*0.07)
0.96
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
103
TABLE 10.
Guymon stepwise c la s s ific a tio n regression equations using
(a) only v is ib le /in fr a re d data and (b) scatterometer (40°
look angle) and v is ib le /in fr a re d data [Crop Type: 8=sorghum,
4 = a lfa lfa , 0=bare s o i l ] .
(a) Crop Type = (M2SCh 4*17.350)-(M 2SCh 7*14.76)(M2SCh 8*1.30)+2.85
0.59
(b) Crop Type = (P band cross pole*0.26)+(C band cross
pole*0.49)+26.147
0.67
Crop Type = (P band cross pole*0.27)-(C band lik e
pole*0.57)+(C band cross pole*0.88)+28.07
0.73
Crop Type = (L band cross pole*0.25)+(L band cross pole
* 0 .2 3 )-(C band lik e pole*0.76)+(C band cross
pole*0.80)+28.22
0.74
Crop Type = (K band lik e pole*0.30)+(L band cross pole
*0.29)+(P band cross pole*0.18)-(C band lik e
pole*0.89)+(C band cross pole*0.74)+27.39
0.75
Crop Type = (M2SlCh5*0.27)+(K band lik e pole*0.32)+(L band
cross pole*0.32)+(P band cross pole*0.17)-(C band
lik e pole*0.81)+(C band cross pole*0.60)+24.2
0.76
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
104
In spite of the difference in the sensor wavelength regions, the
soil regression lines for both Guymon and Dalhart data sets were quite
s im ila r.
Consequently,
it
was
comparable at Guymon and Dalhart.
fe lt
PVI
and TVI
were
reasonably
The equations used to calculate PVI
at Guymon and Dalhart were
(16)
PVI
RG5
=
(0.176 * Z15) + (0.381 * Z25)
(17)
RG7
=
(0.381 * Z15) + (0.825 * Z25)
(18)
where Z15 is the scene radiance from band 9 at Guymon or band 3 at
D alhart, and Z25 is the scene radiance from band 8 at Guymon or band 5
at Dalhart.
Both combinations were strongly related to to ta l biomass
at Dalhart
at
(Figure 26) with PVI showing s lig h tly
higher biomass levels.
greater s e n s it iv it y
Due to the higher s e n s itiv ity and strong
relationship to biomass, PVI was used as the basic combination which
other combinations were compared.
However, the "saturated" zone of
PVI and TVI, where s e n s itiv ity decreased for moderate biomass changes,
was at biomass levels above 1000 g/m2.
'*
The relationship between PVI, TVI and cropy ie ld is
less s i g n i f i ­
cant than the relationship to biomass due a dependence on crop type
(Figure
grain
27).
y ie ld
This
dependency is
comprises
a d iffe r e n t
expected because the economic or
proportion
of the biological
or
vegetative y i e l d for each crop type.
With
the
additional
narrow wavelength
study of the in tercorrelations
bands
for the NSOOl,
a
between bands was needed to evaluate
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction
800G-
8000-
“000H
E 6000T
E 6000-
w 5000-1
E
I
G UOOO-j
H
T
3000-^
W 5000
3000
G
/ 2000M
2
1000-^
0-
XXtX' X
7000
/
1^
«o
1COO­
YfiYi'imY
Him
.9
2000-
TS' y
0- ffrm
0
1
1. 1
1.0
2
3
TVI
PVI
prohibited without perm ission.
LEGEND:
FIG. 26
CROP
+
x
+
□
o
BORE
CuRN
M IL LE T
POSTURE
SORGHUM
A NEEDS
* HEEDS/BRRE
Y NHERT STUBBLE
The relationship between total biomass (g/m2) , and TVI and PVI at Dalhart.
o
cn
Reproduced with permission of the copyright owner. Further reproduction
9000-
9000
8000-
8000-
7000-
7000-
r
r
I 6000E
:
L
0 5000-
L
D 50 00-
K 4000g
:
K 4 0 00G
I
/
/
H 3000-
n
6000
E
H 3000-
:
n
2000-
2000-
1000-
1000-
00.9
1. 1
1.0
3
prohibited without perm ission.
TVI
PVI
LEGEND:
FIG. 27
CROP
BORE
CORN
MILLE T
POSTURE
SORGHUM
WEEDS
HEED 5/ 80 RE
H'riERT STUBBLE
The relationship between fin a l crop y ie ld (Kg/Ha), and TVI and PVI at Dalhart.
o
CD
107
other potential
v is ib le /in fr a r e d combinations.
Figures 28 through 36
display inte rc o rre latio n s of each NSOOl band to bands 1, 2 and 3.
The
relationship
pm)
(Figure
33)
between band 4 and 6 (1.0 0-1 .30
was sim ilar
which PVI is based.
to the
le ft
possible
side
PVI
of
in frared
relatio n s h ip ,
the
corn and dense sorghum fie ld s f e l l
lin e .
relationship
absorption band.
vis ib le /n e a r
All of the bare soil and low biomass f ie ld s f e l l
along the lower right lin e ;
the
pm and 1.55-1.75
The
using
a
relationship
near-infrared
suggested
along
another
band and a water
The equations used to calculate the new PVI were
(19)
PVI64
RG4 =
-1.919 + 0.365(Z35) + 0.158(Z20)
(20)
RG6 =
0.831 + 0.842(Z35) + 0.365(Z20)
(21)
where Z20 is the scene radiance in
NSOOl band 4 and Z35 is the scene
radiance in NSOOl band 6.
A plot of the new PVI versus to ta l
is
A d e fin ite
shown
in
Figure
37.
conventional PVI and PVI64.
the new PVI
(PVI64)
s im ila r it y
exists
biomass
between the
A plot of the two combinations revealed
gave more information on corn fie ld s compared to
PVI and TVI— corn gave a higher PVI64 compared to PVI and TVI (Figure
38).
Not
enough
ground
data
was collected
to
explain
th is
PVI
difference.
Figures 39 through 41 demonstrate the v a r i a b i l i t y of PVI64 within
corn, a l f a l f a and sorghum field s at Dalhart.
The most s tr ik in g exam­
ple was the detection of moisture stressed areas in corn fie ld s 1 and
2.
The
severely
stressed
ring-shaped
areas
w ithin
the
fie ld
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
is
Reproduced with permission of the copyright owner. Further reproduction
T M 3 (6 3 0 - 6 9 0 N M ) VS TM 4 ( 7 6 0 -9 0 0 N M )
TM 3 ( 6 3 0 - 6 9 0 N M ) VS TM 5 ( 1 0 0 0 - 1 3 0 0 N M )
5-1
5-1
+.+
4-
++
8A
T
T
M
M
3
3
2-
1-
0-
prohibited without perm ission.
5
6
7
8
9
10
11
12
13
14
2
+ +
X X
* *
□ □
0 o
4
+
ESRr i E
X
CORN
MILLET
PRSTURE
SORGHUM
HEEDS
HEEQS/ BRRE
WHt RT STUBBLE
*
□
o
A A A
8 8 8
Y Y Y
FIG. 29
3
5
6
7
8
108
F i e l d radiance r e f le c t a n c e values o f NSOOl bands 4 and 5 versus band 3 a t D a lh a r t
in 10 “ k w atts cm- 2 s t e r - 1 .
Reproduced with permission of the copyright owner. Further reproduction
TM 3 (6 3 0 -6 9 0 N M ) VS TM 1 (4 5 0 -5 2 0 N M )
5H
T M 3 ( 6 3 0 —6 9 0 N M ) V S T M 2 ( 5 2 0 - 6 0 0 N M )
5H
+
#+
V
'i'f +oA
&
+ +
+ +
+ ++
+
++ Tj+“ n
+: , v +
^
41+
T
M
3
2-
prohibited without perm ission.
0-
T
0. i
1.2
1.0
2.4
3.0
]
Tin
L EGEND:
CROP
+ + + BRRE
X X CORN
* * * MILLET
□ □ □ PRSTURE
0 0 o SORGHUM
A A A HEEDS
n tt # WEEDS/ BRRE
Y Y Y WHERT STUBBLE
2
3
5
TH2
X
F IG. 28
F i e l d radiance r e f le c t a n c e values o f NSOOl bands 1 and 2 versus band 3 a t D a lh a rt
in 10“ 4 w atts cm- 2 s t e r - 1 .
Reproduced with permission of the copyright owner. Further reproduction
T M 3 ( 6 3 0 - 6 9 0 N M ) VS T M 7 ( 2 3 0 0 - 2 5 0 0 N M )
TM 3 (6 3 0 -6 9 0 N M ) VS TM 6 (1 5 5 0 -1 7 5 0 N M )
5-j
5H
++
+
+
-H3-
T
If A
3-
AO
Y V +An
M
3
A
+
*
$*
XX&
o-
prohibited without perm ission.
2
6
8
10
12
14
■i ■■■1■■■1■l ■■
7.5
12.5
TH7
17.5
22.5
27.5
TM6
L EGEND:
CROP
-r
+
x
*
a
O
A
ft
x
*
□
0
A
ft
+
QHHC
CORN
MILLET
PRSTURE
50RGHUM
HEEOS
M E EDS / BRn E
Y Y Y l-IHERT STUBBLE
FIG. 30
x
*
q
O
A
ft
Field radiance reflectance values of NSOOl bands 6 and 7 versus band 3 at Dalhart
in 10“ 4 watts cm-2 s t e r " 1.
z!
O
Reproduced with permission of the copyright owner. Further reproduction
TM4 (7 6 0 - 9 0 0 N M ) VS TLil ( 4 5 0 - 5 2 0 N M )
TM4 ( 7 6 0 —900NM) VS TM2 ( 5 2 0 - 6 0 0 N M )
15-
15X
14-
14-
XX
<x
$
13-
13-
*
0
o
o
12-
o
*
A
*
12-
x*
1 1-
T
M
X
10-j
A
A
*
#
A
9-
+
9-
+
4+
10-
tt
? *
*
8-
4
+
A
A
11-
4
8-
X
Y +
7-
prohibited without perm ission.
6-
a
7-
+
' r1 tI
T~r ' i * • 1 1
0 .6
1.2
yv
i 1 • i i i i ^
1.8
G-
■
1
2.4
2
TM1
4
5
TM2
LEGEND:
CROP
+
X
*
□
o
A
#
Y
FIG. 31
3
+ + BARE
X X CORN
* * MILLET
□ □ PRSTURE
0- o SORGHUM
A A NEEDS
n n NEEDS/ BARE
Y Y WHEAT STUBBLE
F i e l d radiance r e f l e c t a n c e values o f NSOOl bands 1 and 2 versus band 4 at Dalhart
in 10"4 watts cm- 2 s t e r - 1 .
Reproduced with permission of the copyright owner. Further reproduction
T M 4 (7 G 0 - 9 0 0 N M ) VS TM 3 (S 3 0 -6 9 0 W M )
TM4 (7S0-900N M ) VS TM5 (1000-1300N M )
15-
Is
14
14-1
XX
XX
.X
13
*
0
12
*
*
11
T
M
4 10-
12-
o
.
X X
*
$$
1 * *
T
X
M
A
8
4+
« 8 u ++,
++
+
4 104
8
9H
8
8
++
a ^ + +
7
prohibited without perm ission.
o. \
X
-X
X
11-
x v
rnrrrrr
$
X* *
8
*
* *
9-
s
A
.
A
*
o
*
.
14:
o
o
11"
1 .5
I I
! : I ■I I .
2 .5
aA +
A
^+v
a*
U
7-]
T rT
3 .5
S
1I '
^Y+
11 i r i 1 1 ■ i | i
4 .5
vr - r pTT
7
TM3
10
TM5
L EGEND:
CROP
+
x
*
□
o
+
x
+
□
o
+ BRRE
x
CORN
I- M I L L E T
□ PASTURE
o SCRGPUM
A
A A
NEE05
8 * 8
UEEDS/ BRRE
v v Y U H t RT STUBBLE
FIG. 32
Field radiance reflectance values of NSOOl bands 3 and 5 versus band 4 at Dalhart
in 10" 4 watts cm-2 ster- 1 .
-j
PO
TM4 (760-900N M ) VS TM7 (2300-2500N M )
15-j
15X
x<
14-
XX
X
X
13-
13-
12-
12 -
o
A
o
Reproduced with permission of the copyright owner. Further reproduction
TM4 (7 6 0 —900NM) VS TM6 (1550-1750N M )
8
11-
11-
T
X
xx
H
4
10-
10“
x
X
A
A
8
9-
A
A
A
A
8
9T
+
« ,
8+
^
+V V
8-
B~
7-
X
7-
a
■
6-
prohibited without perm ission.
7.5
12.5
17.5
22.5
6 - 11 11
111.......... 1...............
2
4
6
27.5
+
THS
D+
x
+
□nn +
Q+
.
+
V
+
V rTTT|Trl-|,!-| . 1, | . I 1■. , , •
B
10
1
TTTT
14
TM7
LEGEND:
CnOP
+
+
+
BrifiE
X X X
* ■+ *
MILLET
□
PRSTURE
□
d
CORN
c> o o SCRGHUH
A A A MEEDS
8 « 8 HE E D5 / B RRE
Y y
FIG.
33
v
i.'HERT
STUBBLE
F i e l d radiance r e f le c t a n c e values o f NSOOl bands 6 and 7 versus band 4 a t D a lh a rt
in 10“ ** watts cm- 2 s t e r - 1 .
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
TM 5 ( 1 0 0 0 —1 3 0 0 N M ) VS TM 1 ( 4 5 0 - 5 2 0 N M )
TM5 (1 0 0 0 —1300NM) VS TM2 (520-600N M )
o
8-
*
*
%
x
x'
*
+
X
X
%
X
r
£
+
A +*
+±
%+
* +
+ +
++ Q
*
o
7-
+
+
°
x
m
-
%>><
T
M
5
y
□
CD
4.
5-
I
'
0.6
~rT ■'
' I -r'
' i ' ' ■1 ' 1
1.2
1.8
2 .4
1
3 .0
Till
2
3
4
5
7 M2
LEGEND:
CROP
+
+
+
X X X
BFl RE
CORN
* + * MILLET
□
o
D □
o o
PASTURE
SCi RGhUM
A A A HEEDS
# » » HEED5/6ARE
Y y Y NHEAT STUBBLE
FIG. 34
F i e l d radiance r e f le c t a n c e values o f NSOOl bands 1 and 2 versus band 5 a t D alh art
in 10“ 4 w atts cm- 2 s t e r - 1 .
Reproduced with permission of the copyright owner. Further reproduction
TM5 ( 1 0 0 0 - 1300NM) VS TM3 (630-690N .vl)
TM5 (1000-1300N M ) VS TM4 (760-900N M )
9-
9-
xx
7T
M
T
M
5
cn
5
6-
prohibited without perm ission.
0.5
1.5
2.5
5
3.5
6
TH3
8
9
10
11
12
13
14
TM4
LEGEND;
CROP
+
X
*
□
+
X
*
□
o
o
A A
« «
Y v
FIG . 35
7
+
X
*
□
Bf l nE
CORN
MILLeT
PRSTUnE
o SORGHUM
A NEEDS
» NEEDS/BP. RE
Y WHEAT STUBBL E
F i e l d radiance r e fl e c t a n c e values o f NSOOl bands 3 and 4 versus band 3 a t D a lh a rt
in 10-*+ watts cm- 2 s t e r - 1 .
Reproduced with permission of the copyright owner. Further reproduction
TM5 (1000-1300NM ) VS TMS (1550-1750NM )
TM5 (1000—1300NM) VS TM7 (2300-2500NM )
9-
8-
++
T
T
M
M
5
5
□Y
6-
+ -P
5-
prohibited without perm ission.
4-
47.5
12.5
17.5
22.5
2
27. 5
TM5
36
6
8
10
12
14
TH7
L EGEND:
FIG.
4
CROP
+ + + Br!RE
X X X CORN
MILLET
* *
□ □ □ PRSTUnE
o O 0 SORGHUM
A A A HEEDS
tt n « I-IEEDS/BORE
Y Y Y HHERT S T U 3 3 L E
F i e l d radiance r e f le c t a n c e values o f NSOOl bands 6 and 7 versus band 5 a t D a lh a rt
in 10_l* watts cm- 2 s t e r - 1 .
a\
117
booch
sac >0<
XXXX X
7000-
W
6000-
XtOO'.X
E
T
W
5000-
E
1
G 4QC0H
T
3000-
G
/
M
2 20001000WY
0-,
Y
rrm
1
'
2
I
3
I
■• ■ ’ I 1 ’ 1 1 , •
4
5
I
6
7
1.
3
1
,
^
I
1
■i ■
9 1 0 1 1
PV ! 64
LEGEND:
FIG.
37
CROP
4X
*
□
o
+
X
Br i RE
CuRN
* MILLET
□ PnSTUBE
0 SQnGriUM
A A A WEEDS
X it x W E E D S / S n R E
Y Y Y WHESi S T U B B L E
X
*
□
0
The relationship between to ta l
PVI64 at Dalhart.
(wet) biomass (g/m2) and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction
1.20-1
5-
4T
v
i
oo
1 .0 5 -
31 . 00-
0 .9 5 -
A
0 .9 0 -
0
2
4
6
8
10
PVI 64
I....
6
’ T'
TTT
10
12
PV164
prohibited without perm ission.
LEGEND:
FIG. 38
CROP
+
X
*
□
o
A
#
Y
+
X
+
□
o
A
#
Y
+
X
*
□
0
A
41
Y
DARE
CORN
M IL LE T
PRSTURE
SORGHUM
WEEDS
WEEDS/BARE
WHEAT STUBBLE
The relationship between PVI64, and PVI and TVI at Dalhart.
co
119
FIG. 39
A color photo indicating d iffe r e n t PVI64 levels w ithin a
stressed corn f i e l d (1 and 2) at Dalhart.
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
FIG.
40
A color photo indicating d if f e r e n t PVI64 levels within
sorghum f i e l d (V2) at Dalhart.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
121
FIG.
41
A color photo in d ic atin g d iffe r e n t PVI64 levels within
a l f a l f a fie ld s ( V l l , V12, V13) at Dalhart.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
122
demonstrated by the red color which corresponded to PVI64 values of 4
or less.
Dark green areas represent healthy areas within the f i e l d
with PVI64 values of 6 or greater.
Biomass differences are also e v i­
dent in several a l f a l f a and sorghum f ie l d s .
Summarizing, spectral
data from Dalhart suggested the additional
proposed thematic mapper wavelength
information
on
crop
regions
characteristics
than
provided
present
s lig h tly
techniques
more
using
v i s i b le /in f r a r e d data.
As mentioned,
a normalization
technique
applied to
the active
microwave data was needed to help remove roughness and soil
effe cts in the Guymon and Dalhart data sets.
moisture
Based on the o° response
with look angle, as biomass increases, the vegetative response at high
look angles should also increase compared to the o° response from the
lower
look
(Figures
angles.
This
19 and 20).
was especially
noted
in
the
lin e
plots
Figure 42 demonstrates this e ffe c t for L band
cross pole data from corn (high biomass) and bare soil
(low biomass).
Biomass differences were strongly evident at the larg er look angles,
especially greater than 15° o f f nadir.
Figure 43 represents changes
in the L band cross pole o° due to soil moisture differences within a
m ille t
fie ld
at
Dalhart.
Any
s ig n ific a n t
soil
caused a s im ila r response as the biomass increased.
moisture
increase
However, by c a l­
culating the difference between the response from a large and small
look angle, the soil moisture e ffe c t was diminished while maintaining
a high degree of s e n s itiv ity to biomass differences.
difference
For example, the
between the 40° and 10° look angles was roughly the same
under d iff e r e n t
surface
(0-2 cm) moisture conditions,
12.5 dB.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The
123
DALHART
1.6 H V
F i e l d No.
% of F i e l d C ap acit y
J00%
_
100
-10
-20
-30
-40
5
15
25
35
45
LOOK ANGLE IN DEGREES
FIG.
42
The relationship between L band cross pole a0 and look
angle for a corn f ie l d ( f i e l d 9) and bare f i e l d ( f i e l d 15).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
124
DALHART
1.6 H V
F ie ld No.
% o f F ie ld Capacity
—
60%
20
SIGMA
-10
-2 0
-30
-40
5
15
25
35
45
LOOK ANGLE IN DEGREES
FIG.
43
The relatio n s h ip of L band cross pole o° and look angle for
a m il l e t f i e l d ( f i e l d 3) under d iffe r e n t soil moisture con­
d itio n s .
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
125
la s t e f f e c t , surface roughness was minimized by analyzing cross rather
than lik e polarized data.
Figure 44 demonstrates active
microwave returns
from the same
sorghum f i e l d at two d iffe r e n t look directions--rows p a ra lle l and per­
pendicular to the f li g h t lin e .
the a0 return from rows p a ra lle l
A general s h ift higher was evident for
to the look d ire c tio n .
ence between the near and fa r look angles also
constant under d iffe r e n t surface roughnesses.
The d i f f e r ­
remained r e la tiv e ly
Consequently, most of
the information in the return differences between a near and f a r look
angle in cross-polarized data was related to crop biomass.
Since o°
is expressed in terms of logarithms, a difference between <j° is the
same as an arithmetic r a tio
(a normalization technique).
Also, i t was
anticipated that comparisons of differences in several frequencies and
polarizations
indicated biomass differences.
Comparison of
several
differences
( i . e . 40° L band
cross pole o° - 10° L band cross pole a0;
40° C band
cross pole o°-
5° C band cross pole o°)
band cross
pole 40°
independent
of roughness and soil moisture and most sensitive to bio­
and C
band cross
pole
5°
indicated the C
difference
was most
mass differences.
Other differences, such as the L band cross pole difference be­
tween the 40° and 10° look angle, were sensitive to surface roughness
by penetrating through
several
a lfa lfa
and sorghum canopies.
example, a l f a l f a gave the s im ila r index values as bare s o i l .
For
Conse­
quently, the C band relationship was analyzed and is defined as the
scatterometer vegetation index (SVI).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
126
CU YMON
1.6 H V
F i e l d No.
IX
% of F ie l d Capacity
—
33 %
IX
-10
SIGMA
perpendicular
para llel
-20
-30
-40
5
15
25
35
45
LOOK ANGLE IN DEGREES
FIG.
44
The L band cross pole a0 response as a function of look
angle fo r the same sorghum f i e l d ( f i e l d IX) from two d i f f e r ­
ent d ire c tio n s , the f l i g h t lin e p a ra lle l and perpendicular
to the t i l l a g e d ire c tio n .
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
127
The relationship between SVI and to ta l biomass was s im ila r to the
PVI/total
biomass relationship (Figure 45).
ship between SVI and to ta l
The quadratic
biomass (R2 = 0.88)
relationship between PVI and to ta l
biomass
re la tio n ­
was better than the
(R2 = 0 .7 4 ),
or TVI and
total biomass (R2 = 0 .6 9 ).
The relationship between PVI, TVI, and SVI
was generally
bare fields
lin e ar with
having low SVI
fie ld s with higher index values (Figures 46 and 47).
tended to
fie ld s .
have lower
index values compared to
and vegetated
A l f a l f a fields
the other
vegetated
The lower value indicated the scatterometer signal was either
penetrating through the vegetation and responding to the soil surface,
or the signal was responding to the canopy surface only.
Changes of
SVI within individual fie ld s a ttrib u ta b le to soil moisture differences
were negligible (Figure 48).
At Dalhart, the soil moisture correction
factor for bare field s was 2 db/ 10% change in soil
moisture
(0% to
100% of f ie l d capacity); at Guymon, the factor was 4.5 db/15% change
in soil moisture (a change of 80% of f ie l d capacity).
The e ffe c t was
also dependent on crop type as SVI values from fields
biomass were less dependent on surface soil moisture.
for soil
tures
having higher
Correcting SVI
moisture using C band passive microwave brightness tempera­
improved the relationship only s lig h tly
(Figures
49 and 50).
Part of the variance of SVI within each crop type can be explained by
roughness differences.
For example, at Guymon, SVI values from fie ld s
having rows p arallel to the f l i g h t line were s lig h tly higher, 2-3 db,
then values from fie ld s with rows pHpendicular__to- the f l i g h t l in e .
Attempts to remove the roughness effects were f r u itle s s as the vegeta­
tion e ffe c t was also lo s t.
Analysis of Figures 49 and 50 indicated
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
128
8000-
x x yx. x
70006000-
5003-
3000-
2 200
1000YY Y
0-
□□
m
-35
-30
-25
SVI
L EGE ND:
CROP
+ +
+ BARE
CORN
* MILLET
□ P A 3 l URE
0 SORGHUM
A NEEDS
** N E E D S / B A R E
Y WHEAT S T U B B L E
X X X
* *
□ □
o o
A
A
*t
Y Y
FIG.
15
The relationship between to ta l biomass and the scatterometer
vegetation index, SVI.
(4.75 HV 40° look angle - 4.75 HV 5°
look angle) (R2 = 0 .8 8 ).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction
1. 20-)
■is, yS-:-,
T
V
1 .0 5 -4
p
v
\
I
1.00
0. S5
«
Y
a w : ,
+ iR + +
+
'
m
o.soH
-36
-3 0
tH
-3 7 .5
prohibited without perm ission.
+
X
*
□
+
X
+
□
+
X
+
□
o o o
A
#
Y
FIG. 46
A A
4* #
Y Y
-3 2 .5
-2 7 .5
BrtRE
CORN
MIL LE T
PRSTURE
SGRGMUM
NEEDS
NEED5/BRRE
UHERT STU65LE
The relationship between SVI(db), and TVI and PVI at Dalhart.
ro
10
130
12 X
O
o
E
>
3>
CO
Li_ CC
L
T<x
—J
I- X
CJ
— J —
UJ O
-J X
_1LJ
CL drca:
CLUJ
a.
TD
c
<C
Cl. Q-
tnin
oo
cccc
II
c
c zczz
u_
-J
“U
c
X X
LEll) o
o
u. cc cc iz
jc ro o
cr cn oi tn
+x *□
+x *□
+ X * □
_c
■o
>
00
c;
cu
+->
<u
-Q
o
LU
CD
jC
to
-P
<T>
X X
<U
JC
o »—
fu >
I 01
r-".
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
AA
4 0 -I
0
2
2
o
on
Reproduced with permission of the copyright owner. Further reproduction
40-t
C
□
C
M
M
□
-r
+
+
+
S
0
I
L
S
0 20I
L
20
03
M
0
M
0
I 10S
T
U
R
E
0-
30
□r
+
»
x
-j—
i—
r
~
i—
t—
t•
prohibited without perm ission.
-27.5
o
X -£>
:+
6
o>
□ O
+
I
S
T 10
U
2
R
7
E
Zt
^ ::«
** Y it Y
0-
.-n -
T
-17.5
■12.5
*
FIE LD
+
*
+
+
o
X
t ''r
*I
45
-40
rn
>
i i 1i i i i i i
-35
.
-30
i r_t i ' ' 1
-25
SVI
IX
* 19
0 0 0 21
tt tt 4
z z z 7
FIG. 48
XX
z
SVI
L EGEND:
+
+
Y
-22.5
O HO
X X X 14
□ □ O 20
A A A 22
Y Y Y‘ G
8
L EGE ND:
FIELD
+
*
+ + 01
* * 05
o o 09
n
n
15
Z
z
z
19
X X 03
□ □ 07
A A A 11
Y Y Y 17
21
X
D
The r e l a t i o n s h i p between S V I( d b ) , and 0 -2 cin s o il moisture (%) f o r s e le c te d f i e l d s at
Guymon and D a lh a r t .
Reproduced with permission of the copyright owner. Further reproduction
r
v
i
P
V
I
l.oo.
2-3
0. 95J
Y+
□□
0 . 9 0 —t
-3 9
-3 3
-2 7
0-3
-21
SVI
-110
LEGEND:
CROP
-3 5
-3 0
-2 5
+ + + BARE
prohibited without perm ission.
x x x CORN
+ + * M IL LE T
O d d PASTURE
« o o SORGHUM
A A A HEEDS
# # tt HEEDS/BRAE
v Y y HHERT STUBBLE
FIG. 49
The relationship between soil moisture corrected SVI(db), and TVI and PVI at Dalhart.
CO
ro
+
+
++,
nO
+o
T
v
2-
i
x
9
* +
1.00
X
0 .9 5
0-
*
X>#
TJTTT,
24
- 22
o*
Reproduced with permission of the copyright owner. Further reproduction
4-
**
**
x
x l Xx!feXx^ X
"JTTTT "I"......
.
-20
-1 8
-1 6
-1 4
-12
-10
SVI
0 .9 0 -2 5
.
15
-20
10
prohibited without perm ission.
S VI
LEGEND:
CROP
+ + + RI.FRLFH
w ^
^^^^
+ + + S0RGHUM--R0WS PARALLEL TO FL IG HT LIN E
a □ o SORGHUM— ROWS PERPENDICULAR TO FL IGHT LIN E
FIG.
50
The r e l a t i o n s h i p between the s o il moisture c o rre c te d S V I( d b ) , and TVI and PVI a t Guymon.
CO
CO
134
th at SVI was insensitive to low PVI or TVI changes; however, at higher
PVI and TVI
SVI
(PVI greater than 1.5 and TVI greater than
became sensitive to changes in biomass.
1.06)
Indications
levels
also show
that SVI was s lig h tly more sensitive to biomass changes at high bio­
mass levels than PVI or TVI.
Other
attempts
scatterometer
data
to
determine
proved
combinations
f r u it le s s .
that
Consequently,
normalized
each
data
the
set
could only be analyzed separately.
Problem 4
Considering the results from the previous three problems, biomass
was
a strong
indicator
of
crop type differences
within the
active
microwave region— crops with greater biomass had higher active micro­
wave responses and were c la s s ifie d separately from other low biomass
groups.
I f the tree classifcation model were applied to an a g ric u l­
tu r a l region which has a crop with d iffe re n t biomass or biophase, misc la s s ific a tio n with other crops .is l i k e l y .
For example, the unsuper­
vised c la s s ific a tio n technique tended to confuse immature sorghum with
'
a lfa lfa .
model
To f u ll y
understand the u t i l i t y of the tre e -c la s s ific a tio n
under d iffe re n t
for applications under
active
microwave
biophases and adjust
the c la s s ific a tio n model
d iffe re n t biomass le v e ls , v is ib le /in fr a r e d and
responses
needed to
be
considered.
The
sorghum
fie ld s at Dalhart and Guymon covered a wide range of biomass and bio­
phases
ranging from crops that were just
emerging to f u ll y
headed.
Analysis the response difference within a given crop type due to bio­
mass differences
indicated possible
errors
of m isclassification
gave physical explanation for the tree c la s s ific a tio n model.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
and
135
The v is ib le /in fr a r e d response showed a d e f in it e trend as biomass
increased and crops matured.
red responses at
data from
Dal hart
Figure 51 represents the red/near i n f r a ­
and Guymon,
respectively.
In both cases,
bare soil and low biomass fie ld s were lin e a rly relate d .
the crop matured, the distance from the soil
increased.
reproductive
Data
from fie ld s
biophase
with
had the
the
largest
As
lin e to the data point
highest
biomass
and at
distance from the
soil
the
lin e .
The perpendicular distance had been defined as the perpendicular vege­
ta tio n
index (PVI).
As the crop matured from heading, leaves began
to senesce and PVI decreases.
No fie ld s at Guymon or Dalhart were in
the last biophase.
The active microwave response from several
f ie ld s
at D a lh a rt--
22, V2 and V6 , and 12--indicated differences at far look angles which
appeared to represent d iffe re n t biomass le v e ls .
fie ld
F ield 22 was a bare
at Dalhart; V2 was an ir r ig a t e d sorghum f ie l d at Dalhart that
had reached the heading stage; V6 was a dryland immature sorghum f i e l d
only 60 cm t a l l
at Dalhart; and 2 was a corn f i e l d with a high biomass
at
The K band data indicated no s ig n ific a n t
Dalhart.
between the
d iffe r e n t
biomass levels
(Figure
52)
while the C band
cross pole data indicated some differences (Figure 53).
sorghum f i e l d ,
22.
ghum,
angle.
V2, had s lig h tly
higher
The immature
returns than the bare f i e l d ,
The largest difference was between the vegetation
corn)and the
The
bare
s o il- - a s
much
as 10 db in
(mature sor­
the 40°
L band cross pole data also indicated some
between d iff e r e n t biomass levels.
differences
look
differences
Again, the corn and mature sorghum
fie ld s had higher returns at high look angles compared to the bare and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction
TM 3 (6 3 0 -6 9 0 N M ) VS TM 4 (7 6 0 -9 0 0 N M )
M M S 7 ( 6 6 2 - 7 0 1 M M) VS M M 3 9 ( 7 7 0 - B S 3 N M )
2 . 7-
5-
2.4
+
4-
^ +. + «
+++ a
3-
xx
x x
2. 1
f
K
#A
1.8
* *
2-
"
M
**
M
1 .5 4
rA
>
°
XX
XX
>sk
X
□
□
S n
cb
a
x**
0.9-
0-
prohibited without perm ission.
11
, dD 2
Eip
12
13
14
0 . 6-
TM4
L EGE ND:
+ + + BRRE
X X X CORN
* * * MILLET
□ D □ FRSTUFiE
0 0 o SORGHUM
A A A MEEDS
# # » HEE DS/ 6 RRE
Y Y Y WHEAT STUBBLE
CROP
FIG.
51
-
_
a#
X
X
T n ’TTr
10
X
1.2
xx
0.3-
□
□
D+
r
111111111111111111
2
4
"T1
rr' i '
10
12
MHS9
L EGEND:
CROP
+
x
*
□
+
x
*
□
x
+ RL FRL F R
BRRE
* CORN
□ SORGHUM
The r e d / n e a r - i n f r a r e d r e la t io n s h i p f o r f i e l d s at Guymon and D a lh a r t .
CO
O)
137
10
0
SIGMA
s•
DALHART
13. 3 V V
F i e l d iV o.
% oI F i e l d C a p a c i t y
60%
V2
40
V6
10
40
5
15
25
35
45
LOOKANGLE INDEGREES
FIG.
The K band l i k e pole o° response as a function of look angle
fo r bare soil ( f i e l d 22), sorghum ( f i e l d V2 and V6 ) , and
corn ( f i e l d 2) at Dalhart.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
138
1- - - - - - - - - - - - - - - - - 1- - - - - - - - :- - - - - - - 1- - - - - - - - - - - - - - 1- - - - - - - - - - - - - - - -
r
DALHART
4.75 H V
F ie l d No.
% of F i e l d Capacity
60%
V2
_
40
_
40
10
V6
10
SIGMA
0
-1 0
-2 0
-30 -
5
15
25
35
45
LOOKANGLE INDEGREES
FIG.
The C band cross pole a0 response as a function of look
angle for bare soil ( f i e l d 2 2 ), sorghum ( f i e l d V2 and V6 ) ,
and corn ( f i e l d 2) at Dalhart.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
139
low biomass fie ld s --a s much as 7 db (Figure 54).
However, the respon­
ses at the high look angles in the P band cross pole data were sensi­
tiv e
to
fields
therefore
only with
high biomass
(Figure
55).
implied high frequency active microwave
ated" at r e la tiv e ly
"saturated"
at
The analysis
responses
"satur­
low biomass levels while low frequency responses
very high biomass levels.
C band would then
best
separate lower biomass crops, L band would separate moderate biomass
crops and P band would separate high biomass crops.
The Guymon results
also
tended to
indicate the same situation
(Figures 56 through 59).
However, roughness from row direction played
an important factor also.
The best example indicating biomass d i f f e r ­
ence was L band cross pole from f ie ld I X , —headed, dense sorghum, 15—
emerging sorghum, 4 - - a l f a l f a ,
and 14--bare soil
(Figure 58).
the far look angles were responding to high biomass levels.
other
look
angles
indicated
that
surface
roughness
return by masking the vegetative differences.
Again
Data from
influenced
the
Attempts to elim inate
roughness effects proved to be unsuccessful, as removal of roughness
also reduced the vegetation e ffe c t.
From the analysis of both spectral
data sets,
a multi frequency
active microwave system using a low and high frequency could improve
c la s s ific a tio n
and biomass estimation accuracy.
meter vegetation index (SVI), which was strongly
Given the scatterorelated to biomass
and PVI, a sim ilar combination using 40° P band cross pole a 0 - P band
cross pole a0 was included with SVI.
The resulting modified index
(SVIM) is defined as
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
140
DALHART
1.6 H V
10
F i e l d No.
% of F ie l d Capacity
22
60%
V2
40
10
0
40
SIGMA
-1 0
-2 0
-30
-40
5
15
25
35
45
LOOK ANGLE IN DEGREES
FIG.
54
The L band cross pole a0 response as a function of look
angle for bare soil ( f i e l d 22), sorghum ( f i e l d V2 and V6 ) ,
and corn ( f i e l d 2) at Dalhart.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
141
DALHART
0.4 H V
0
F i e l d No.
% o l F ie l d Capacity
_
60%
V2
V6
10
SIGMA
40
5
15
25
35
45
LOOK ANGLE INDEGREES
FIG.
The P band cross pole o response as a function of look
angle for bare soil ( f i e l d 22), sorghum ( f i e l d V2 and V6 ) ,
and corn ( f i e l d 2) at Dalhart.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
142
10 _
\
\
0 -
-10
-
-2 0
-
o
CO
GUYMON
13.3 V V
-30 F ie l d No.
4
15
100”o
14
100
15
100
I X
-40 -
% ol F ie l d Capacity
_
________________ 100
25
35
45
LOOKANGLE INDEGREES
FI(
56
The K band l i k e pole ct response as a function of look angle
for bare soil ( f i e l d 14), a l f a l f a ( f i e l d 4 ) , emerging
sorghum ( f i e l d 15) and headed sorghum ( f i e l d I X ) .
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
143
10
GUYMON
4 .7 S H V
-
%. of F i e l d C a p a c ity
F i e ld N o .
4
14
__
—
________
15
0 -
-1 0
-
-2 0
-
-3 0
-
-4 0
-
IX
1007c
100
1 00
____
100
CJ
I
I
15
25
35
45
LOOK ANGLE IN DEGREES
FIG.
57
The C band cross pole o response as a function of look
angle for bare soil ( f i e l d 14), a l f a l f a ( f i e l d 4 ) , emerging
sorghum ( f i e l d 15) and headed sorghum ( f i e l d I X ) .
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
144
GUYMON
1.6.H V
F ie l d No.
% o( F i e l d Capacity
;oo%
too
100
IX
t oo
SIGMA
-1 0
-2 0
-
-30
-40
5
15
25
35
45
LOOKANGLE INDEGREES
FIG. 58
The L band cross pole a0 response as a function of look angle
fo r bare soil ( f i e l d 14), a l f a l f a ( f i e l d 4 ) , emerging sorghum
( f i e l d 15) and headed sorghum ( f i e l d IX ).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
145
F i e l d No.
% of F ie l d Capacity
_
iOO%
_
100
100
-10
IX
100
SIGMA
-2 0
-30
-40
-50
5
15
25
35
45
LOOKANGLE INDEGREES
FIG.
The P band cross pole a0 response as a function of look
angle for bare soil ( f i e l d 1 4), a l f a lf a ( f i e l d 4 ) , emerging
sorghum ( f i e l d 15) and headed sorghum ( f i e l d IX ).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
146
SVIM
=
(40° C band cross pole - 5° C band cross pole) + (40° P
band cross pole
- 5° P band cross pole)
(22)
The modified SVI was also strongly related to to ta l biomass at Dalhart
(R2 = 0.73) (Figure 60).
In comparison, the relationship of SVIM to
biomass at Dalhart was not as strongly related to PVI or TVI at Guymon
(Figure 61).
active
Again, a l f a l f a did not have high SVI values indicating
microwave
penetration
through
the
canopy
for
P band data.
Higher frequency scatterometer data may indicate the presence of dense
a l f a l f a f ie ld s .
The SVIM responses from sorghum fie ld s were, however,
greater than low biomass or bare f ie ld s .
With the s e n s itiv ity of the P band cross pole data to differences
in high biomass, the only change needed in the c la s s ific a tio n model
was to use P band cross pole differences as a f i r s t step to separate
the
high
biomass
fie ld s
from fields
with
medium and
low biomass.
Higher frequency L or C band cross pole data was then used as c r i t e r i ­
on to separate fie ld s with medium and low biomass levels.
c rite rio n ,
the
corn
and
dense
sorghum
fie ld s
at
Using these
Guymon
were
separated— anything having a return of -47 db or higher would be clas­
s ifie d as corn at Dalhart and -36 db or higher at Guymon.
c r ite r io n , the accuracy of the tree c l a s s if ie r
Using these
improved s lig h tly at
Dalhart and Guymon--81% at Dalhart and 76% at Guymon.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
147
6000-1
X >XX X X X
x xxx
:k x
70QQ
6000
500
30QQ-
2 20GD-
o oo o
1000-
0-
-60
-75
SVIM
L EGEND:
Fiti.
60
CROP
+
x
*
□
o
+
x
*
□
o
+ BARE
CORN
* M 'LLcT
□ PA ST URE
o SORGHUM
a a a
HEEDS
« « » MEED3/2AREv r y WHEAT 5 T U 5 3 L E
x
The relationship between to ta l biomass and the modified
scatterometer vegetation index. SVIM [ (C band cross pole
40° - C band cross pole 5°) + (P band cross Dole 40° - P
band cross pole 5 ° ) ] .
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction
+; ++
c H -.
+
1 .1 5 -
♦ a ~
mg a o '-30
*
„
A **
* □
+ +
1.10
%
+
.cP
+a
T
q, □ D °
+ 4v
1 .0 5
V
I
2-
.+
Q A
0
X
*
*
-2 5
-2 0
□
1. 00
*+
* *%
X
V
0 .9 5 -
****
0-;■•r'""’™!...... 11111111
-5 0
0 .9 0 -
111
[n
i'M
m
i
-5 0
-1 5
.
.
.
.
.
.
.
“ 40
-3 5
-2 5
-0 0
-35
T
-30
-1 5
SVIM
1
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
-3 0
-'1 5
Xx
TTJTTTT
-20
prohibited without perm ission.
SVIM
LEGEND:
CROP
+ + + RLFRLFA
x x x BARE
+ * ♦ 50RGHUM — RONS PARALLEL TO FLIGHT LINE
P a □ S0RGHUM--ROHS PERPENDICULflR TO FL IGHT LIN E
FIG. 61
The relationship between the modified SVI (SVIM) and TVI and PVI at Guymon.
-F*
00
149
CHAPTER VI
SUMMARY AND CONCLUSIONS
Since the study was divided into four problems, results from each
w ill be discussed in d e t a i l .
Also, an overview summarizing the study
and its implications w ill be follow the dicussions of the results.
Problem 1
The f i r s t problem determined spectral bands which were sensitive
to crop type differences.
frequencies
Results
implied s e v e r a l
were sensitive to crop type
look angles greater than 35° o f f nadir.
to
active microwave
differences,
especially
at
The response differences due
vegetation dominated the effects of roughness and soil
moisture.
The most sensitive frequencies and polarizations included C band cross
pole,
L band lik e
Depending
on
the
discriminated crops.
and cross
crop
pole
type,
and P band l ik e
responses
from
and cross pole.
certain
frequencies
For example, L and P band discriminated between
sorghum and corn, and C band was able to discriminated between a l f a l f a
and
bare
s o il.
Other
active
microwave
sensitive to roughness or so il moisture.
sensors
were
prim arily
The v is ib le /in fr a r e d sensors
were not as sensitive while the passive microwave data were sensitive
to soil moisture differences.
especially well
The biomass differences were detected
in the v is i b le /i n f r a r e d bands.
were noted using NS001 band 6
data
(water
Also,
absorption
stressed areas
band).
The
v is ib le and infrared data were sensitive to the presence or absence of
vegetation, but not necessarily certain crop type p a irs .
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
150
Problem 2
The second problem determined the most accurate crop classifying
dendrogram for the Guymon and Dalhart spectral data.
In th is problem,
a r e la t iv e ly accurate dendrogram using active microwave, v is ib le , and
infrared data was developed for both Guymon and Dalhart spectral data
sets.
The dendrogram was based f i r s t
on separating
"rough"
from
"smooth" fie ld s using active microwave data, and second, on separating
each class between the bare and low biomass fie ld s from heayily vege­
tated f ie ld s .
The preferred active microwave frequencies and p o la ri­
zation were L and C band cross pole which were most sensitive to bio­
mass differences
between crop types.
Response differences
frequencies c la s s ifie d different scales of roughness.
infrared
data were
non-vegetated
then
f ie l d s .
used to distinguish
Classification
Red and near-
between vegetated and
accuracies
using the
dendrograms were 77% for Dalhart and 70% for Guymon.
individual
in both
bands did not improve the accuracy.
sim ilar
Data from other
The implication was
that one model requiring data from four bands (v is ib le through active
microwave)
could discriminate
d iffe re n t
crop types
with
reasonable
accuracy.
More data sets are needed, however, to thoroughly test the
tree c la s s ific a tio n model.
Problem 3
Problem three determined the u t i l i t y
of estimating biomass and
discriminating crops using visible/infrared/microwave data compared to
v is ib le /in fr a re d
data.
The primary
result
in
problem 3 was the
indication that microwave data improved or maintained c la s s ific a tio n
and
biomass
estimation
accuracy
in
comparison
to
conventional
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
151
c la s s ific a tio n .
The conventional
classifica tio n technique used only
v is ib le /in fr a re d data to classify and estimate biomass.
tis tic a l
techniques
such
as
discriminant
Various sta­
analysis
and
step-wise
regression indicated the inclusion of active microwave aided in clas­
sifying
agricultu ral
crops.
visible/infrared/m icrowave
With
higher
s a te llite
accuracy,
o r a ir c r a ft
less
passes
frequent
would
be
required fo r an adequate estimate of crop acreage or biomass.
In addition, the proposed thematic mapper wavelength bands pro­
vided more information on vegetation than the Landsat v is ib le /in fr a re d
combinations.
vegetation
(0.76 -
For example, a combination sim ilar to the perpendicular
index (P V I), but using input data from the near infrared
0.90 pm) and water absorption
additional
information
on corn
band
(1.55-17.5
compared to the
results
pm) provided
from broad
band MSS red and near infrared wavelengths.
Not enough ground data
was collected to determine what physiological
parameter within f ie ld
differences of the the new combination was detecting.
The new combi­
nation, PVI64, was s lig h tly more related to biomass than the original
combination of red and near-infrared data that had been used to calcu­
la te PVI.
Further studies using these bands are needed.
F in a lly , an active microwave vegetation index (SVI) was developed
using C band cross pole data from the 5° and 40° look angles.
The
combination, which was developed to normalize the two data sets, was
highly correlated to PVI.
The major implication was that use of this
combination would allow a c la s s ific a tio n and biomass estimation that
would be possible regardless of cloud conditions.
nized that
the
sensor
combination
required
to
I t is fu lly recog­
c o llect
5°
and 40°
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
152
imagery over the same areas with active microwave is highly impracti­
cal and most lik e ly
not economically feasib le.
The result i s ,
how­
ever, sig n ific a n t from an academic standpoint and may help in under­
standing
cover.
the
It
scattering
phenomena
that
takes
place
in
vegetative
is s ig n ific a n t to note that L band differences between 5°
and 40° did not
respond to
vegetation
other than corn and sorghum
since the L band energy was penetrating through the canopy more than C
band.
However, further tests of the model are needed in agric u ltu ra l
regions having d iffe r e n t management practices.
In spite of the success in discriminating crops and estimating
biomass within each data set—Guymon and Dalhart— the sets could not
be combined due to the absence of active microwave c a lib ra tio n .
V a ri­
ous attempts to normalize the data sets using combinations, such as
the SVI, were unsuccessful.
lyzed
separately.
active
microwave
Consequently, both data sets were ana­
Any further
data
must
experiment
include
requiring
some means of
collection
calib ra tin g
of
the
microwave sensors.
Problem 4
The fourth problem determined the e ffe c t of biomass differences
on the crop classifying dendrogram developed in problem 2.
from problem four
s ig n ific a n tly
indicated
dependent
that
the t r e e -c la s s ific a tio n
upon biomass.
Implications
Results
model
are that
was
crops
which have sim ilar responses at the same time of year, such as wheat
and barley
may be
physiological
able.
indiscriminant.
However,
at
certain
biophases
differences, such as plant water content may be detect­
Consequently, multi-temporal data is s t i l l needed to accurately
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
153
separate two "confusion" crops.
To make the model even more sensi­
t i v e , multifrequency microwave data are needed to separate even higher
biomass levels.
Results proved that the P band cross pole scattero-
meter returns are sensitive at high biomass levels at Dalhart.
Inclu­
sion of the P band cross pole data improved crop c la s s ific a tio n accur­
acy over the use of L band and C band data.
Overview
Having
answered
the
questions
posed
by
each
problem,
the
hypothesis—can microwave data help improve c la s s ific a tio n and biomass
estimation
compared to
present
infrared data—can be validated.
techniques
using
only
v is ib le
and
Given the results from Guymon, Okla­
homa, and Dalhart, Texas, active microwave data does aid in improving
c la s s ific a tio n and biomass estimation.
Results indicated that m ulti­
frequency active microwave data would be needed to classify m u ltip le cropped agricultural areas accurately.
L and P band data can discrim­
inate between sorghum and corn; C band can discriminate between bare
soil and a l f a l f a but not between corn and sorghum.
In addition, NS001
data indicated combinations of the water absorption
band (1.55-1.75
pm) and the near-infrared band ( 1 . 0 - 1.3 pm) gave more crop information
than the red/near infrared combinations.
Accurate multispectral clas­
s ific a tio n and biomass estimation models were developed from both data
sets.
However, two major factors pose problems in using active micro­
wave data--soil
moisture and surface
roughness.
With many of the
vegetated crops being irrig a te d and the non-vegetated f ie l d remaining
fallow, a bias entered into this analysis due to soil moisture d i f f e r ­
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
154
ences.
The most accurate technique to remove the soil moisture e ffe c t
would be to develop a correction factor using passive microwave data
which is primarily sensitive to soil
the model
(Schmugge,
1979).
moisture changes,
The best
method to
as inputs to
minimize
surface
roughness is to use cross-polarized active microwave data, which the­
o r e tic a lly isolates the volumetric ( d ie le c tr ic ) effects while minimiz­
ing the scattering
(surface roughness)
that were developed were unable to
alone.
e ffe c ts .
Other combinations
remove the effects
of
roughness
Attempts to remove the roughness e ffe c t also diminished the
vegetation e ffe c t.
A second problem dealt with spatial
resolution.
If
large areas
of the world are to be covered in a short time period, s a t e l l i t e sys­
tems w ill
spatial
be required.
The question arises
as to what should the
resolution be and should the resolution be sim ilar for each
frequency.
V is ib le /in fra r e d data often has high spatial
resolution;
passive microwave data has low resolution while active microwave reso­
lution can be controlled by system design and processing.
Many fie ld s
around the world are too small to be seen even by Landsat.
Conse­
quently, by increasing spatial resolution to allow analysis of in d iv i ­
dual fields
implies extremely large amounts of both v is ib le /in fr a re d
microwave and active microwave data processing.
resolution, knowledge of composite
(fie ld s
With lower spatial
of d iffe re n t
crop types,
soil moisture, and surface roughness) returns within the cell size is
required.
For example, what effect would the return from a 32-hectare
f i e l d have on the composite return of a 10 km resolution c e l l , and can
cla s s ific a tio n
and biomass information be extracted from the larger
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
155
size
cells?
Consequently,
future
studies
are
needed to
find
the
proper resolution size for re a s o n a b ly accurate estimates of vegetation
using visible/infrared/microwave data.
Advantages of using microwave systems are obvious:
independence
of weather and sunlight and the opportunity fo r fewer passes with the
v is ib le /in fr a re d systems due to higher c la s s ific a tio n accuracy.
Both
reasons are advantageous over present v is ib le /in fr a r e d systems devel­
oped during the LACIE period.
Some foreign agricultu ra l areas that we
have previously been unable to monitor from a s a t e l l i t e due to cloud
cover could be monitored in the future.
two-fold:
(1)
The fin a l
an improved world-wide a g ricu ltu ra l
results would be
production system
which would prevent another event such as the U. S./Soviet Union wheat
crises which occurred in 1974, and (2) domestic food supply planning
would be more e f f i c i e n t
better
domestic
as better production estimates would induce
storage
and
production,
and
s t a b iliz e
commodity
prices.
Consequently, active microwave sensors need to be seriously con­
sidered as additional
sensing tools in evaluating agricultural
With
data,
the
additional
potential
world
food
disasters
areas.
may be
averted.
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
156
REFERENCES
1
Allen,
W. A ., H. W. Gausman, and A. J. Richardson.
1970.
Mean
effective optical constants of cotton leaves.
J. Opt. So'c. Ani.
60:542-547.
2
Bauer,
M. E.
1975. The role of remote sensing in determining
the distribution and y ie ld of crops. Adv. Agron. 27:271-304.
3
Bauer,
M. E ., L. P. S ilva, R. M. Hoffer and M. F. Baumgardner.
1977.
Agricultural scene understanding.
Final Report.
LARS
Contract Report 112677.
4
Beckman, P. 1966. Scattering by composite rough surfaces.
Trans, on Antennas and Prop. AP-13:1012-1015.
5
B illin g s le y , F. C ., M. R. Helton, and V. M. O'Brien, eds. 1976,
LANDSAT follow-on: a report by the applications survey groups.
Vol. 1—Executive Summary. NASA JPL Tech. Memo. 33-803.
6
Blanchard, A. J. and S. W. Theis.
1981.
Volumetric e ffe cts in
cross-polarized airborne radar data.
Intern. Geosci. and Remote
Sensing Symposium, IEEE Washington, D. C. Vol 1. pp. 629-630.
7
Blanchard, B. J . , J. L. Nieber, and A. J. Blanchard.
1979.
Continuation of measurements of hydrologic soil cover complex
with airborne scatterometers. Remote Sensing Center Final Report
3496. NSG-5156, Texas A&M University. 43 pp.
8
Brakke, T. W., E. T. Kanemasu, J. L. Steiner, F. T. Ulaby, and E.
Wilson. 1981. Microwave radar response to canopy moisture, leaf
area index and dry weight of wheat, corn, and sorghum.
Remote
Sensing of Environ. 11:207-220.
9
Chance, J. E. and E. W. LeMaster.
1977.
S u it's reflectance
models for wheat and cotton: theoretical and experimental tests.
Appl. Optics 16:407-412.
10
Choudhury, B. H . , T. J. Schmugge, R. W. Newton and A. Chang.
1979. Effect of surface roughness on the microwave emission from
s o ils . J. Geophys. Res. 84:5699-5706.
11
Clark, B. V. and R. W. Newton.
1979.
JSME scatterometer data
processing.
Remote Sensing Center
Final Report 3337.
NAS914875. Texas A&M U n iversity. 110 pp.
12
Claussen, J. P ., R. 0. Stroud, B. V. Clark, B. R. Jean and R. W.
Newton. 1979.
The system and hardware design of real-tim e fan
beam scatterometer data processors. Remote Sensing Center Final
Report 3556. Texas A&M University. 275 pp.
13
Cooley, W. W. and P. R. Lohnes.
1971. M u ltivaria te data analy­
sis . John Wiley and Sons. New York. 364 pp.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
IEEE
157
14
Davis, B. R ., E. 8 . Lipscomb, and S. J. Knight. 1965.
Terrain
analysis by electromagnetic means:
laboratory investigations in
the 0.76- to 5- micron spectral region.
U. S. Army Tech. Report
3-693. Vicksburg, Miss. 165 pp.
15
Deering, D. W., J. W. Rouse, J r . , R. H. Haas and J. A.
Schell.
1975. Measuring "forage production" of g r a z i n g units from Landsat MSS data.
Tenth In t. Symp. on Remote Sensing of Environment
Vol I I .
pp. 1169-1178.
16
deLoor, G. P ., A. A. Jurriens, and H. Gravesteijn.
1974.
radar backscatter from selected agricultu ral
crops.
Geosci. Elect. GE-12:70-77.
17
Donald, C. M. 1963.
Competition among crop and pasture plants.
Adv. Agron. 15:1-118.
18
E llerm eier, R. D., D. S. Simonett, and L. E. Dellwig. 1969. The
use of multi-parameter radar imagery for the discrimination of
terrain characteristics.
Center for Res. in Eng. Sci. Report
61-69. Univ. of Kansas, p 127-135.
19
Fenner, R. G ., G. F. Pels, and S. C. Reid.
1981.
A parametric
study of t i l l a g e effects on radar backscatter. Intern. Geosci.
and Remote Sensing Symposium. IEEE. Washington, D. C. Vol I I .
pp. 1294-1308.
20
Gates, D. M.
1980.
York.
611 pp.
21
Hartigan, J. A.
1974.
New York. 351 pp.
22
Heydorn, R. P ., R. M. B iz z e ll, J. A. Quirein, F. M. Abottee, and
C. A. Sumner.
1979a.
Classification and mensuration of LACIE
segments.
LACIE Proceedings of Tech. Sessions. Vol. I. NASA/JSC
p. 73-86.
23
Heydorn, R. P.,
M. C. Trichel, and J. D. Erickson.
1979b.
Methods for segment wheat area estimation.
LACIE Proceedings of
Tech. Sessions. Vol I I .
NASA/JSC. p. 621-632.
24
Holliday, R.
1960a.
Plant population and crop y ie ld :
(Review A r t ic le ) . Field Crop Abstr. 13:159-167.
25
Holliday, R.
1960b.
Plant population and crop y ie ld :
(Review A r t i c l e ) . Field Crop Abstr. 13:247-254.
26
Jackson, T. J . , T. J. Schmugge, and J. R. Wang. 1981.
Effects
of vegetation on passive microwave estimates of soil moisture.
Intern. Geosci. and Remote Sensing Symposium. IEEE. Washginton,
D. C. Vol I . p. 375-387.
Biophysical Ecology.
The
IEEE
Springer-Veriag.
Clustering algorithms.
New
Wiley and Sons.
part
part
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
I
II
158
27
Janza, F. J.
1975.
Remote Sensing, Vol I .
Interaction
Chapter 4.
mechanisms.
p. 75-179.
In Manual
28
Johnson, S.
1967.
Hierarchical clustering schemes.
t r ic k a 32:241-254.
29
Kauth, R. J. and G. S. Thomas.
1976.
The tassel led cap--a
grpahic description of the spectral-temporal
development of
a g ric u ltu ra l crops as seen by Landsat.
Proc. Symp. Machine
Proc. of Remotely Sensed Data. LARS. IEEEE. Cat No. 76.
30
Kirdyashev, K. P ., A. A. Chukhlantsev, and A. M. Shutko.
1979.
Microwave radiation of the earth's surface in the presence o f a
vegetative cover. Radio Eng. and Electronic Physics. 24:37-44.
31
Kubelka, P. and F. Munk.
1931.
Ein beitrag
Farbanstriche. Z. Tech. Physik 12:593-601.
32
Lundien, J. R.
1971.
Terrain analysis by electromagnetic
means.
Tech. Report 3-693.
Report 5.
U. S. Army Waterways
Expt. Sta. Vicksburg, Miss. 85 pp.
33
MacDonald, R. B.
1979.
Experiment design.
Proceedings
Tech. Sessions: LACIE Symposium. Vol I .
NASA/JSC p. 1-2.
34
McFarland, M. J.
1976.
The correlation o f Skylab L-band
brightness temperatures with antecedent p re c ip ita tio n .
Preprints
Conf. on Hydrometeorology. Amer. Met. Soc. Ft Worth, p. 60-65.
35
Morain, S. A. and D. S. Simonett. 1967. K-band radar in vegeta­
tion mapping. Photogrammetry Eng, 33:730-740.
36
National Academy of Sciences. 1977. The world food and n u tr itio n
study:
the potential contributions of research.
National
Research Council. Washington, D.C.
37
National Research Council.
1977.
space for earth resource surveys.
Washington, D. C . , 139 pp.
38
Newton, R. W.
1977.
Microwave sensing and i t s application to
soil moisture detection.
Remote Sensing Center T e c h . Report
RSC-81. Texas A&M U n iversity. 500 pp.
39
O 'N e il l, p. 1981.
41
O rlo ci, L.
The Hague.
42
Park, J. V. and 0. W. Deering.
1981.
Relationships between
d iffu se reflectance and vegetation canopy variables based on the
rad ia tiv e tra n s fe r theory. NASA Tech. Memo. 82067. 19 pp.
zur
of
Psychrome-
Optek
der
of
Microwave remote sensing from
National Academy of Sciences.
Personal Communication.
1978.
M u ltiv a ria te analysis in vegetation research.
Amsterdam. 451 pp.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
159
42
Peake, W. H ., R. L. Riegler, and C. H. Schultz.
1966.
The
mutual inte rp re tation of active and passive microwave sensor
outputs.
Proc.
of
Fourth Symp. on Remote Sensing of
Environment. Ann Arbor, Mich. p .771-777.
43
P otter, J. F . , E. M. Hsu, A. G. Houston, and D. E. P i t t s . 1979.
Accuracy and performance of LACIE area estimates. Proceedings of
Tech. Sessions: LACIE Symposium. Vol. I .
NASA/JSC. p. 527-574.
44
Richardson, A. J . , C. L. Wiegand, H. W. Gausman, J. A. Cuellar,
and A. H. Gerberman.
1975.
Plant, soil and shadow reflectance
components of row crops.
Photogram. Eng. and Remote Sensing
41:1401-1407.
45
Richardson, A .J. and C. L. Wiegarid.
1977.
Distinguishing
vegetation from soil background information. Photogram. Eng. and
Remote Sensing 43:1541-1552.
46
Rouse, J r . , J. W., R. H. Haas, J. A. Schell, and D. W. Deering.
1973.
Monitoring vegetation systems in the great plains with
ERTS. Third ERTS Symp. NASA SP-351. Vol. I p. 309-317.
47
Schmugge, T.
J. Appl. Met.
48
Schwarz, D. E. and
discrimination and
Proc. 5th SRSE.
49
Sibley, R.
1973.
Microwave emission and scattering from
vegetated te r ra in .
Tech. Rpt. RSC-44, Remote Sensing Center,
Texas A&M University, College Station, Texas, 148 pp.
50
Simonett, D. S ., J. E. Eagleman, A. R. Erhart, D. C. Rhodes, and
D. E. Schwarz.
1967.
The potential of radar as a remote sensor
in a g ricu ltu re .
Center for Res. in Eng. Sci. CRES Report 61-21.
Univ. of Kansas.
51
Swain, P. and 8 . Davis.
1979. Fundamentals o f pattern recogni­
tion in remote sensing.
In Remote Sensing: the Quantification
Approach. Chp. 3 p. 136-187.
52
Tucker, C. J . , J. H. Elgin, J r . , and J. E. McMurtrey I I I .
Temporal
spectral measurements of corn and soybean
Photogram. Eng. and Remote Sensing 45:643-653.
53
Ulaby, F. T. and J. E. Bare.
1979.
Look direction modulation
function of the radar backscattering coefficien ts of agricultural
fie ld s .
Photogram. Eng. and Remote Sensing 45:1495-1506.
54
Ulaby, F. T . , P. P. B atlivala and M. C. Dobson. 1978. Microwave
backscatter dependence on surface roughness, soil moisture, and
soil tex tu re , part I :
bare s o i l .
IEEE Trans. Geosci. Elect.
GE-16:286-295.
1978.
Remote sensina of surface
17:1549-1557.
soil
moisture.
F. Caspell.
1968.
The use of radar in
id e n tific a tio n of agricultural
land use.
1979.
crops.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
160
55
Ulaby, F. T . ,
P. P. B a tliv a la , andJ. E. Bare.
1980.
Crop
id e n tific a tio n with L-band radar.
Photogram. Eng. and Remote
Sensing 46:101-105.
56
Ulaby, F. T . ,
T. F. Bush, and P. P. B a tliv a la .
1975. Radar
response to vegetation I I : 8-18 GHz band.
Trans. IEEE on
antennas and Prop. AP-23: 608-618.
57
Ulaby, F. T. and T. F. Bush.1976a.
Monitoring
wheat growth
with radar. Photogram. Eng. and Remote Sensing 42:557-568.
58
Ulaby, F. T. and T. F. Bush. 1976b. Corn growth as monitored by
radar.
IEEE Trans, on Antennas and Prop. AP-24:819-828.
59
Ulaby, F. T . ,R. V.
L i, and
K. S. Shanmugam.
1981.
Crop
c la s s ific a tio n by radar.
Intern. Geosci. and Remote Sensing
Symp. Vol I . p. 638-647.
60
Waite, W. P. and H. C. MacDonald.
with K-band imaging radar.
GE-9:147-155.
61
Wang, J. R ., R. W. Newton, J. W. Rouse, Jr.
1980.
Passive
microwave remote sensing of soil moisture: the e ffe c t of t i l l e d
row structure.
IEEE Trans, on Geosci. and Remote Sensing GE-18:
296-302.
62
Wolfe, W. L. and G. J. Z issis.
1978.
Office of Naval Research-ERIM. 1665 pp.
63
Wortman, S.
1976.
1971.
Trans.
Food and Agriculture.
Vegetation penetration
IEEE Geosci.
Elect.
The infrared handbook.
Sci. Amer. 235:31-39.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
161
APPENDIX A
DATA QUALITY, CALIBRATION, AND OMISSIONS
At both Dal hart
and Guymon, data were deleted for various rea-
sons--quality and excessive a ir c r a f t a ttitu d e parameters.
t e r defines the questionable sensor and soil
methods used for correcting the data sets.
This chap­
moisture data and the
Each sensor system and
soil moisture w ill be discussed in d e ta il.
NS001/M2S
Most of the v is ib le /in fr a re d data was good quality at both Dalhart and Guymon.
One of the exceptions was the
excessively
water absorption bands (bands 6 and 7) on 8/14/80 at Dal hart.
noisy
Since
no means was possible to correct the data, they were eliminated from
further data analysis.
Also, at Dal hart band 1 data for fie ld s 6 , 8 ,
10,12 and 22 were deleted due to unstable c a lib ra tio n .
With the exception of band 9 (0.77-0.86 ym) M2S data at Guymon,
the calibration
lis t s
information
proved to be quite
the equations used to convert
values.
raw d ig ita l
stable.
Table Ala
counts to radiance
Note band 9 had three d iffe re n t equations applicable at d i f ­
ferent periods of the experiment.
All of the working NS001 bands had less stable calib ratio n in fo r ­
mation
at
Dalhart.
Table Alb l is t s
d ig ita l counts to radiance values.
the equations used to convert
Note that several bands had d i f ­
ferent calibration values on each flig h t day.
Calibration of the thermal band proved to be d iffe r e n t for Guymon
and Dalhart.
The c a lib ra tio n ,
using the PRT-5 data, showed that at
Guymon the low temperature calibration black body aboard the plane was
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
162
TABLE Al.
a.
b.
Equations used to convert raw NS001/M2S d ig ita l counts (DC)
to radiance values, R,
(10-1+ watts cm“ 2 ster_1) for
Guymon (a) and Dalhart (b)
channel 4
R = 10 ‘ a T -° - 4 * (°C-12)
7
R=
8
R = — ggfr—
9
R=
6 ,^
9
R=
6 - ^ -q--0- 4 *(DC-10)
9
R =
-4
6- ; « 0x l ° *(DC—17)
channel 1
R =
1,907104 *(DC-1) (8/14 & 8/16 ( F i t 1))
1
R =
1
R =
1 -*™
2
R =
4 , ^ x l ° 4 *(DC-21)
2
-4
R =------4 , ^ x l °--* ( DC-21) (8/18)
3
R =
3
R =
4
R =
-4
■1'- *4| -X-1- * (DC-9)
4
R =
-4
* (DC-9)
4
R=
,-4
“ f f 10 * (DC- 8 ) (8/18)
9*23Q10 ^ * (DC-13)
~* (DC-14)
-,-4
1Q--- * (DC-12) ( 8 /2 , 8 /5 , and 8/ 8 )
(8/11)
(8/14)
-4
M f e -1-0-- *(DC-1) (8/16 ( F it 2))
-4
*(DC-1) (8/18)
(8/14 - 8/16)
-4
*(DC-29)
-4
5 , ^ x l ° *(DC-29)
(8/14-8/16)
(8/18)
(8/14-8/16 ( F it 1))
(8/16 ( F it 2))
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
163
TABLE Al.
(Continued)
-4
5
R =
5232X '°
* ( dc~8) (8 /1 4 -8 /1 6 ( F i t 1))
5
R =
5 f f 7x10
-4
* (DC-9) (8/16 ( F it 2))
5
R =
5 io47Xl-g
-4
*(DC-9) (8/18)
6
R =
2 2 2 2 1 0 3 * ( DC~1 2 )
(8/ 16)
_3
6
R=
T&T~
* (DC-12)
(8/18)
_3
7
R = 1iioX^° *(DC-16) (8/ 16 &8/ 18)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
164
too high while the high temperature calib ratio n black body was measur­
ing the proper temperature.
This implied that low surface tempera­
tures were as much as 5°C too high.
tion occurred.
At Dalhart, the opposite
condi­
The low temperature calib ratio n black body was reading
the proper temperature while the high temperature c a lib ra tio n body was
reading 5°C too low, suggesting that
high surface temperatures were as
much as 5°C too low.
The normalization
are as follows:
( f l ig h t 2 ),
solar correction
August
14,
5.7;
1.1; and August 18,
factors
August
1.0.
solar correction factors are August 2,
16,
(cose-j)
(flig h t
for Dalhart
1 ),
two data
sets,
the
Guymon data
and
For Guymon, the normalization
1.7; August 5, 1.6; August 8 ,
5.0; August 11, 1.0; August 14, 1.6 and August 17, 1 . 6 .
the
2.0;
set
required
To normalize
a m ultiplication
factor of 1.3 to roughly match the radiance values at Dalhart.
Scatterometer
Due to excessive a ir c r a f t ro ll and d r i f t , several look angles had
to be eliminated at Dalhart and Guymon due to the uncertainty of the
cell
being within the f i e l d .
At Dalhart,
a ll
active microwave data
from one f ie ld had to be e lim in a te d --fie ld 16 on 8/18/80.
Also, data
at 40° and 45°
fie ld s
look angles o ff nadir from several
other
8/18/80 were eliminated due to excessive d r i f t (Table A2).
on
At Guymon,
flyin g conditions were much worse; consequently, data from more fie ld s
needed to be deleted.
A complete l i s t of omitted look
angles are
given in Table A3.
from 8/11, 8/14, and 8/17/78 were
most ques­
Data
tionable.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
165
TABLE A2.
Date
Questionable scatterometer data fo r Dalhart
F ie ld #
8/14/80
All data is good
8/16/80
All data is good
8/18/80
L12 R2
L12 R2
L l l R3
20,8,18
14
16
Questionable Analysis
45° ( d r i f t
9°)
40, 45° ( d r i f t 11°)
All Angles
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
166
TABLE A3.
F ield #
Date
Questionable Analysis
O
o
8/2/78
Questionable scatterometer data fo r Guymon
LI
L2
LI
L2
R1
R1
R2
R2
2 , 4 , 6 ,7 ,8 , 2 x , l x
1 0 ,13,14 ,15 ,2a ,2x ,lx
2 ,4 ,6 ,7 ,la ,2 x ,lx
15,17,2a
8/8/78
L2
L2
L4
LI
R1
R2
R1
R2
17, lx
2A
26
2,6,7
a ll
a ll
a ll
a ll
8/11/78
LI
L3
L2
L4
LI
R1
R1
R1
R1
R2
6 , 8 , 2x
a ll
a ll
all
a ll
a ll
L3 R2.
L2 R2
8/14/78
19 ,2 2 ,lx
2 x,
24,25,27
4 , 6 , 7 ,1A
22
(-9 ° d r i f t )
(-9' d r i f t )
(-8 d r i f t )
angles
angles
angles
angl es
angl es
LI R2
4
a ll
L3 R2
L2 R2
19
13
LI R3
L3 R3
L? R3
a ll fields
lx
13,14
15
L3 R1
L4 R1
L3 R2
o
<3*
L4 R2
a ll
45° (-4° d r i f t , 4° r o l l )
a ll
a ll
o
8/5/78
( - 8 ° d r i f t , 2 ° rol 1 )
10,17
2A, 2X
24,26,27
10
8/17/78
,45°
45°
45°
45°
,45°
45°
40° ,45°
40° ,45°
a ll
a ll
45°
( - 8° d r i f t , 3e r o l l )
(9° d r i f t )
(9° d r i f t , 3° r o l l )
( 11° d r i f t )
angles
angles
(9° d r i f t )
L4 R2
lx,19,20
24,25,2x
35° ,40°,45° (-12° d r i f t )
35°,40°,45° (-12° d r i f t )
a l l angles
40°,45° (-10° d r i f t )
45° (■-9° d r i f t )
LI
L4
L2
L4
2
2x
2x
2x
40°,45°
40°,45°
40° ,45°
40°,45°
R1
R1
R2
R2
21,22
2 x ,24,25,26,27
21,22
*delete these same fields for passive data
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
167
sets.
There appears to be cross-over in the P band data collected at
Guymon and Dalhart.
returns
with
Figure A1 represent
look
angle
for
the
same
perpendicular to the f l i g h t lin e .
polarized
data
increasing
roughness
look
20°
angle
can
data;
be
fie ld ,
and cross polarized
IX,
which
Any rapid
d ire c tly
cross-polarized
increase of
attributed
This characteristic
(Blanchard and Theis, 1981).
data
to
large
suppresses
the
scale
roughness
Consequently, the rapid increase
Figures A2a and
large increase in a0 at the 15°
for the cross pole data compared with the lik e pole
for the f i r s t four f l i g h t
tille d
a 0 with
band l i k e and cross pole responses from a milo f i e l d (25)
at Guymon.Note the absence of any
look angle
rows
is most apparent in
in a0 should not appear in the cross-polarized data.
A2b show P
had
Note the large increase in the l ik e
look angle.
characteristics.
lik e -p o la rize d
e ffe c t
at
like
days.
In the late r f lig h t s the rows
data
were
and the row height was increased causing a larger increase in
a0 at 15° look angle in both lik e and cross polarizations.
example of data with minimum c ro ss -talk.
This is an
The cross-polarized data
should have smaller decreases in a0 with higher look angles.
however, the P band response for f i e l d IX in figure Al.
Note,
At the 15°
look angle, a large increase in a0 occurs in both lik e and cross pole
data.
This
suggests
cross-polarized data.
excessive
cross-talk
between
the
lik e -
and
No attempt has been made to t r y and correct for
the cross-talk in the P band cross polarized data.
In addition, note
the
data
between the
as much as 5 db difference.
For these
firs t
o° differences
in the P band cross
and f o u r th -- flig h ts
polarized
reasons we questioned the 0.4 GHz data, especially at Guymon.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
168
GUYMON
F IE L D I X
Run
Freq.
I
0.4 H H
0.4 H H
0.4 H V
0.4 H V
-10
-20
-30
-40
5
15
25
35
45
LOOK AN GLE IN DEGREES
FIG.
A1
F ield IX (sorghum) P band lik e and cross pole response with
rows perpendicular to the f l i g h t l in e .
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
169
G UYM O N
10 -
F ie ld 25
D a te
0.4 HH
% o f F ie ld C a p a c ity
85%
160
70
0 -
100
135
45
-1 0
-
o
CO
/
-20
/
-
.- .It ? * *
’/
S
-3 0 -
-40 -
15
25
35
45
LOOK ANGLE IN DEGREES
FIG.
A2a
Scatterometer response from the P band lik e pole system
over f i e l d 25 (sorghum) with rows perpendicular to the
f l i g h t l in e .
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
170
GUYMON
20
F ie ld 25
0.4 H V
D a te
Vo o l F i e ld C a p a c ity
8/2
—
8 /5
—
8/11
160
70
8/8
10
857o
..
100
8 /1 4
135
8 /1 7
45
SIGMA
0
-1 0
-20
-30
5
15
25
35
45
LOOK ANGLE IN DEGREES
FIG.
A2b
Scatterometer response from the P band cross pole system
over f i e l d 25 (sorghum) with rows perpendicular to the
f l i g h t lin e .
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
171
Figure A3 represents lik e and cross polarized returns from the C
and L band scatterometer for f ie l d 25 (sorghum), at Guymon.
was t i l l e d
tio n .
with
rows perpendicular to the f l i g h t
A slig h t increase in return at the 20°
band lik e
pole and cross
pole is evident.
The f i e l d
line and polariza­
look angle for the L
The increase
suggests
again that some
cross-talk mayexist between the
polarizations.
Note
the absence of
cross-talk in the C-band data.
A slig ht increase in
the lik e -p o larized data at 10° look angle off nadir is not evident in
the cross polarized data.
These data suggest that the other frequen­
cies have some degree of cross-talk, but on a much smaller scale than
the P band data.
Since scatterometer power was lik e ly d iffe r e n t for the Guymon and
Dalhart data sets and no means exists
for externally calibrating the
system, normalizing the two scatterometer data sets proved to be quite
d iffic u lt.
Figures A4 through A7 represent scatterometer responses
for each frequency from two bare fields having approximately the same
surface soil moisture and roughness at Guymon ( f i e l d
14) and Dalhart
(fie ld
of L band l i k e
19).
Note the extreme difference
in
s h ift
polarized data between the d ifferen t frequencies.
difference exists between the two data sets
addition, the s h ift in the
constant nor is
A4 and
and in
As much as a 15 dB
in some instances.
In
lik e polarizaton for a l l frequencies is not
i t even in the same direction.
Note that in figures
A6 f i e l d 14 is higher than 19 while in Figure A5 i t is lower
Figure A7 they are a lik e .
The fa r look
the most comparable between data sets.
angles appeared to be
Since the differences between
data sets are not constant with look angle, normalization of the data
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
172
G UYM ON
F IE L D 25
1.6 H H
1.6 H V
4 .7 5 H H
4 .7 5 H V
<t
-10
-20
-30
5
15
25
35
45
LOOK A N G L E IN D E G R E E S
FIG.
A3
Scatterometer response (C and L band l i k e and cross pole)
from f i e l d 25 at Dalhart on 8/16/80.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
173
F ie ld N o. /Freq.
20
% of F ie ld Capacity
DALHART
19
13.3 V V
G UYM O N
14
13.3 V V _______________ 90
907c
10
SIGMA
0
-10
-20
-30
5
15
25
35
45
LOOK ANGLE IN DEGREES
FIG.
A4
Scatterometer response (K band lik e pole) from f i e l d 19 at
Dalhart on 8/16/80 and f i e l d 14 at Guymon on 8 /5 /7 8 .
Soil
moisture conditions were approximately 90% of f ie l d
capacity.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
174
F ie ld No. /Freq.
DALHART
19
4.75 H H
90%
19
4.75 H V
90
14
4.75 H H
14
4.75 H V
90
. . . .
90
SIGMA
GUYMON
% o l F ie ld C apacity
-1 0
-2 0
5
15
25
35
45
LOOK ANGLE IN DEGREES
FIG.
A5
Scatterometer response (C band l ik e and cross pole) from
f i e l d 19 at Dalhart on 8/16/80 and f i e l d 14 at Guymon on
8 /5 /7 8 . Soil moisture conditions were approximately 90% of
f i e l d capacity.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
175
F ie ld N o . / Freq.
10
% of F ie ld C apacity
DALHART
19
1.6 HH
90%
19
1.6 H V
90
GUYM ON
14
1.6 H H
90
14
1.6 H V
. . . .
90
0
SIGMA
-10
-20
-30
-40
5
15
25
35
45
LOOK ANGLE IN DEGREES
FIG.
A6
Scatterometer response (L band l ik e and cross pole) from
f i e l d 19 at Dalhart on 8/16/80 and f i e l d 14 at Guymon on
8 /5 /7 8 .
Soil moisture conditions were approximately 90% of
f i e l d capacity.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
% o l F ie ld C apacity
F ie ld N o ./F req .
DALHART
19
19
GUYMON
14
14
0.4 HH
90%
0.4 H V
90
0.4 HH
90
0.4 H V
....
90
-10
SIGMA
-2 0
-30
-40
.
N.
-50
5
15
25
35
45
LOOK ANGLE IN DEGREES
FIG.
A7
Scatterometer response (P band lik e and cross pole) from
f i e l d 19 at Dalhart on 8/16/80 and f i e l d 14 at Guymon on
8 /5 /7 8 . Soil moisture conditions were approximately 90% of
f i e l d capacity.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
177
proved
unsucessful.
However,
one normalization
technique
used to
compare information within a data set was a data combination using a
ct° difference between two look angles in the same data set.
Since o°
is based on the algorithm of a, a difference implied a ratio between
a— a
common
normalization
technique.
It
was
believed
that
this
technique provided much information on vegetation while minimizing,
soil
moisture
and
surface
roughness
e ffe c ts ,
depending
on
the
frequency and polarization.
Passive Microwave (MFMR)
Since the passive microwave radiometer was oriented at a constant
angle
(3°
from n ad ir),
MFMR data.
3.5°
the
roll
would imply questionable
Consequently, anytime the airplane had roll
f ie l d
deleted data.
mon--L
any excessive
average MFMR data was deleted.
greater than
Table A4 l is t s
With the exception of data from one f l i g h t line at Guy-
band data
on 8/11/78 had highly e rra tic
brightness tempera­
tures on one occasion—brightness temperatures were quite stable.
highly
the
variable
brightness
variations in the f i e l d .
temperatures
indicated
local
The
unmeasured
Therefore, the following fields at Guymon
were deleted from further analysis: fie ld s 10, 13, 14, 15 and 17.
Soil Moisture
Each sensor has a d iffe re n t cell
data,
soil
moisture
f ie l d
averages
size.
were
Consequently, to compare
determined
for
the
area
observed by each sensor by averaging only one sample located within
the
observed area.
Unfortunately,
in
some cases,
averaging
point
locations of soil moisture proved not to be a r e lia b le fie ld average.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
178
TABLE A4.
Guymon and Dalhart questionable MFMR data
Date
Fi el (i #
% Roll
8/8/78
L2 R1 IX
5.3
8/11/78
L3 R1 IX
LI R2 6
L4 R2 24
4.9
-5.1
4.9
8/14/78
L2 R1 10,17,2a
5 . 4 , - 8 , - 5 .6
respectively
4.9
-4 .8
L4 R1 27
L3 R3 IX
8/17/78
L3 R2 22
5.0
8/18/78
LI R1 16
6.3
These fie ld s were deleted from the MFMR plots due to excessive r o l l ;
d r i f t was not a factor.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
179
For instance, several
rows were irrig a te d and seen by the sensors but
not sampled within the f i e l d .
Also r a in fa ll events occurred at Guymon
between sampling periods—on 8/2 and 8/8/78.
An attempt was made to
correct the soil moisture by adding the amount of r a in fa ll or i r r i g a ­
t io n , assuming complete i n f i l t r a t i o n .
did a good job.
In some cases, this correction
But in the end the questionable soil
was deleted from the data set.
moisture data
The fie ld s at Guymon with deleted soil
moisture data was for 8/2: 22, 27, 20, 25, 19, 24, 8/ 8 : lx , 2x, 2, 10
and 8/17: l x ,
(lin e 2 ).
With the deletions, c a lib ra tio n s , and normalizations the Guymon
and Dalhart data sets were complete as possible.
Data for the s ig n if­
icant channels are presented in Appendix B and C.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
180
APPENDIX B
DALHART DATA SET
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
181
iV
1
**
o
X
10
•
M
>
©
*
•
O
m
mm
mm
mm
•
o
■0
•
O
•0
•
o
*
*
O'
•
o
9
•
O
o
n
9
•
O
eo
eg
o
•
o
K
N
♦
•
O
©
ro
00
•
o
N
eg
o
•
o
n
eg
o
•
o
♦
♦
•
O'
N
O
•
eg
eg
•
N
•
N
mm
©
r»
•
9
00
00
00
•
o
eg
©
K
10
•
N
K
♦
•
0
0
0
•
0
K
N
•
©
eO
eg
•
©
9
n
•
n
N
f0
•
eg
o
9
0
eg
0
K
•
0
00
•
0
eg
•
r*
mm
9
©
eg
•
©
eg
1
mm
M
00
•
eg
eg
o
s
•
©
00
♦
©
t
©
00
eg
•
eg
•0
•
n
9
e
O
o
r>
O'
•
o
#
©
9
w
©
•
o
©
•
©
©
*
O'
•
o
mm
C
•
eg
t
i
1
1
tm
9
♦
*1
1
1
1
1
1
1
1
I
•0
*
*
w>
DALHART DATA SET
CROSS POLE 5 DEGREE LOCK ANGLE (OBI
CROSS POLE 40 DEGREE LGCK ANGLE (OB)
CROSS POLE 5 DEGREE LOCK ANGLE (OB)
PU S-P
BAND
BAND
BAND
COS«L
C40-L
a
0*
a
o «
mm •
X
Ui u
10
10
UI
•J
3
10
ml
z
<
£
3
o
J
o
K
I*
<
a
z
o
z
UJ
x
00
Ui
w
#
X
*0 UJ
♦ o
♦ z
o
00
<0
10
UJ
J
z
mm
*
3
■
0
(0
z
Ui
s
0*
a
UI 10
a UI
3 3
tm «J
10 4
M 9
a
X o
X
z
M
UI
o 00 10
z a Ji
M 10
z
X
o z u
© © 00 o 00 >0
♦
N M s z
Q < fc* >• UI
Ui Z *0 X u i <0
-1 < UJ H X UI
o UJ 3 ez
a
• UJ O J a
a
♦ > UI O Ui
> > ex
to
10 • a
o 10 < o z <0
Ui y a
s
j
o
V V) 3 X
0 W a fM 00
o £ m o
a
z < O IL o UJ
< Q z 10 II a
UI z
03
m 0. < e
a O a a X
it o iu tm (O
o (/) a n
♦ Z ii •0
a II •M >
10 > tm
•o a
N
UI
Ui
a
o
Ui
O
•
o
01
N
•
10
m
N
>
a
s
©
♦
•
o
1
1
1
1
I
I
1
1
f
1
1
•
9
•
o
K
•
©
in
n
N
•
O'
•
m
n
N
9
O
•
©
mm
O
.1
n
u
n
♦
9
•
eg
eg
V)
•
N
m
w
•
eg
N
©
CM
•
©
N
©
•
©
eg
1
mm
*0
O'
•
eg
«
*9
©
•
m
O
m
O
K
•
M
eg
©
©
mm
•
eg
«>
•
•
eg
tfi
I
©
•
«•
©
•
9
<n
f
1
o
Hi
mm
mrn
f
h
A
(M
*0
mm
t
mm
•
©
n
•
eg
UJ
J
a
CD
3
H
(0
W
X
Ui
X
9
II
a
a
a
u
n
m
©
*
a
©
o
a
mm
•
mm
•
©
•
eg
eg
©
♦
•
eg
•
eg
♦
•
eg
eg
I
i
•
•
I
•
•
9
1
1
1
i
1
1
1
t
8
1
«
1
1
1
1
f
1
o
#
u
10
a
u
>
<
Q
9
•
eg
<g
mm
•
K
1
10
•
«•
«
m
•
eg
g»
«
<g
O'
•
eg
«)
1
eg
©
•
10
O'
o
•
1
1
mrn
o
N
O'
©
•
o
in
•
0
•
9
■
0
0
00
•
•
111
•
9
©
©
1
I
•
0
00
00
•
IS
0#
•*
©
—
©
«*
©
♦
©
00
0
»*
0
•*
X
>z
a
X
o
3
«
o
3
<
©
3
<
©
3
<
©
3
<
©
3
<
©
3
<
©
3
<
a
j
Ui
—
~
u.
s
00
—
N
•*
K
•“
©
-*
O
0
0
©
0
{
i
I
c
1
1
1
I
1
1
•
|
1
I
«
I
I
•*
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
182
mm
♦
BAND
iD O >
ID S )
ANGLE
LOC<
UJ
J
c
a
CROSS
5
Li
o
POLE
DEGREE
UJ
UJ
rr
o
♦
P05=P
(O U i
ANGLE
AN C LE
PO LE
CROSS
0 AND
SET
DATA
5 O tG RE E
LOCK
40
DEG REE
LOCK
PULE
DALHART
CROSS
BAND
C 0 5 -L
C4 0 = L
oj
A
*
Q •
♦
Z
UJ
J
J ©
z N<
X
K B
U
(J
J ♦
W
©
C
c
u
o
z
<
aj
a.
ti
o
♦
a
m
©
UJ
j
*
A
m
©
UJ
j
z
■i a
in
z •©
z
i UJ
t
X
w
«U
X
o
z ►Q M in
o M
mm a
X
z
UJ X
* Q UJ
Z Q M
Z c
o
mm
M in
2
Q z u
o «•
©
K M e
O < H H
Z >■* < u
< UJ
s
VS UJ 3
• UJ o J
* > UJ g
> >
• cc
© < a z
J UJ u
in 3 z
a L> * OJ
z
a I
< o u. o
03 2 in <1
UJ z M
a* a . < o
o « a z
o UJ © l/ l
in CL n
z 1?
>
ti
n > Vw
© a.
oj
«
©
in
UJ
g
j
<
>
o
z
M
in
m
M
z
v-
z
UJ
in
UJ
a
a
UJ
a
in
o
o
M
a
UJ
a
t
<
I
OJ
-0
«J
O
•
■*
O'
*0
0*
•
o
m
N.
0'
u
o
©
in
o
•
r*
c
©
•
mm
o
r»
n
e
mm
rn
©
•
mrn
N
in
in
•
o
'c
♦
UJ
•
o
©
CM
mm
A*
•
0
mm
tn
<v
•
©
♦
•
m
CM
•
hm
01
•
©
♦
K
•
O
®
in
•
mm
mm
tn
n
i'#
•
M
©
o
©
•
•*
in
0*
©
•
o
mm
>
a
©
N
CM
•
mm
in
m
N
m
©
m
•
©
mt
(U
•
O
• OJ
•
4
«0
9
t
O
•
O
s
9
+
•
O
•
n
OJ
©
<
«
U1
eg
14
tn
•
in
OJ
01
*o
•
®
in
o
•
o
9
eg
•
9
N
4
•
9
•
fm
me
•
1*1
eg
w
•
OJ
<U
•
l*>
©
K
<
•
•
•
l»>
O
♦
a
©
mm
•
Ol
eg
0
•«
o
•
o
mm
%
n
n
•
©
1
in
©
a
©
©
•
©
mm
*
O
•
v«
«
m
•
(M
(VI
N
OJ
•
o
Of
«
O
*
U
©
mm
•
♦
Hi
•
N
O
•
©
n
•
o
m
l
mm
O
•
r*
♦
1
f)
®
•
l*>
m
1
«o
<4
•
:n
•
*
••
I
mt
mm
N
X
<
a
x
m
a
UJ
UJ
B
II
a
o
z
u
•
1
9
•
1
1
•
1
1
1
•
•
1
1
«
1
f
«
1
9
9
in
©
u
0
N
•
UJ
OJ
•
in
©
•
©
mm
>
*»
O
OJ
NJ
tn in
«
• ■
9
9
0*
•
CD
mm
z
in
OJ
•
©
•
r>
o
I
1
1
«
i
>
<
a
OJ
*
•
OJ
N
1
*
mm
•
K
•
in
1
+
••
•
•0
m
rn
O
•
o
mm
f
1
©
—
—
©
•*
•
®
•
o
mm
9
mrn
me
•
0J
"•
©
tn
«
mm
VI
1
9
©
•
©
I
©
••
©
©
«
o
m*
9
•
9
*
•
m
01
Q
N
•
9
mm
♦
V)
•
mm
tit
•
01
*
•
<0
mm
1
<9
tn
•
♦
•0
o
-
I
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction
DALH4RT DATA SET
‘ .C0 5=L
BANDCROSS PULE S DEGREE LOCK ANGLE 1DBI
C40=L
EANDCROSS POLE *0 DEGREE LCCK ANGLE IDtt>
P05-P
BANDCROSS POLE 5 DEGREE LOCK ANGLE <08)
P40=P
BANDCROSS POLE 40 OEGREE LOCK ANGLE (OB)
ZIS. Z20# Z35=NSOOl BANDS 3* 4# AND 6 < 1 0 *« < -4 ) MATTS CH»*<-2)
PVI=PERPENDICULAR VEGETATION INDEX CDIMENSIONLESS)
I V I 3 TRANSFORMED VEGETATION INDEX (0 1 MENS1UNLESSI
S *01=0-2 CM VOLUMETRIC SOIL MOISTURE <X)
PERIODS REPRESENT MISSING VALUES
CROP«WEEOS
I
MONTH
DAY
C05
JM
AUG
14
•
3M
AUG
16
- 4 .SI
3M
AUG
16
•
JW
AUG
10
•
4W
AUG
4*
prohibited without perm ission.
POSi
P40
14.71
•1 9 .1 7
29.47
- 7 .1 7
2 . 19
9.8 8
16.58
2.153
1.066
13.2
«
22.69
• 3 .9 4
2 . 56
10.16
19.14
1 .931
1 .047
8 .9
•
20.64
- 7 .1 8
2 .7 7
12.26
20 .95
2.623
1.064
4 .7
14
.
.
.
•
.
•
AUG
16
•
21.57
-0 .8 2
2 .6 7
9.4 4
21 .65
1 .356
1.017
9 .7
4 'J
AUG
16
•
.
2 3 .9 7
-7 .3 0
3.2 4
9 .5 6
2 3 .5 3
1 .067
0.9 97
9 .7
4M
AUG
18
•
.
20.29
«
3.C4
10.02
22 .4 5
1 .439
1.017
3 .5
F IE L O
C40
S T *« (-1 ))
•
-4 8 .5 5
Z I5
Z20
Z35
«
«
.
PVI
TV I
a
.
•
•
SMO 1
.
•
00
CO
184
S
-
O
z
©
o
o
i
I
>
(ft
(ft
«
••
(ft
+
©
©
*
»
«r*c
©
N
>
a
o
«
OJ
n
©
©
o
©
0
*»
01
(ft
ol
©
o
N
O
OJ - *
I (ft
*• a -* s 0* ft
s
Oa
**c3s
'J
Z
<
UJ
-i U
J
©J
2 O
< z
*
UJ
*
-J
z z
U3
ft v*>
*- z
»*>
UJ
I*
2 u, z
0 20
u
o
Olu s W Z
a
o
© UJ O UJ
<
o
♦
—O
z o
X
z
UJ J
-
2 —(ft
a
(ft lw
z a (ft
—
ft —
z
z u
OJ
J
J «J
< U
J
UJ 3 2
j a
♦• lU
>o
uj S uj
©
(ft (ft u)
a©
(ft (ft ft (ft • z
o o c a
a 2 z a
0)
©
0)
©
> > 2
n < a z (ft
j UJ u o
3 az rg a«
o u u ft
3
U
o o o c Z - 0 12
z z z z < o U o UJ
< < < < 2 Z ft H a
U
J
© Q O S
- a < o
-j -j a a O lz. 2 Z
©
•*
9
01
♦
©
•*
©
OJ
©
OJ
*01
04
X
2
3
o>
a
a
o
o © - O- -z zH ft
«* (fl <T
I- U U
J
it
a
<
UJ U
J UJ z k- < u; (ft &
I
UJ
I o a o c
a o. a
0
> **
©
OJ
f t UJ f t
Z Z UJ
— (/> <
n ** :*
uo
•J
J
UJ UJ
JU
J UJ
UJ U
«
u 3 3
Z ^ J
<
* w * c
a
o
o -J D - j
©
©
o
o
♦
a
n
7
•0
•
OJ
J
©
I
I
ft
♦
•
0*
1
©
•
7
01
o
«
o
OJ
ft
1
ft
8
N
•
O
**
1
n
•
©
•
©
ft
r>
OJ
©
•
OJ
OJ
1
2
U
©
tn
o .01
a
m
•
OJ
OJ
01
•0
OJ
OJ
ft
ft
01
ft
«
OJ
«
♦
•
©
•
©
ft
n
OJ
©
ft
mt
K
•
OJ
OJ
•*
u
0* 0*
in
«
i
Z II
•I M
ft >
N
*
ft
©
“
M ii ii i> O U
J *- (ft
(ft O (ft o ft & II
o o i a.
•
©
mt
©•
0»
♦
I
OJ
a
(A
O
o
•n
i
*
•0
OJ
V
t
©
•
©
<
1
0)
•
©
♦
1
01
•
N
©
ft
o
OI
I
1
ft
ft
*
ft
©
*
1
N
•
o
N
1
•
K
•
•
00
©
I
©
ft
m
UJ
l
OJ
ft
#
©
1
©
•
o
©
I
«
ft
•0
©
t
©
ft
OJ
©
•
00
ft
OJ
OJ
1
©
•
■0
OJ
I
©
ft
©
OJ
1
*
•
•0
01
1
o
ft
©
OJ
1
©
ft
©
N
1
©
00
©
00
©
■0
©
■0
©
©
■0
©
©
<
ft
©
1
•
N
ft
©
ft
I
OJ
N
ft
N
>
<a
+-•
X
tz
a
s
3
<
o
3
<
O
J
O
J
>
>
©
Q
«0
♦
©
•0
©
©
«•
©
3
<
(5
3
<
O
3
©
3
©
3
<
(5
3
<
O
J
OI
>
O
J
O
J
>
O
J
O
J
>
<
<
©
3
©
<
3
<
©
©
3
<
3
<
3
<
©
3
<
©
3
<
©
©
>
©
>
©
>
©
©
3
a
•j
UJ
>
>
>
>
>
©
>
>
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
©
>
185
V
9
1
1
1
1
•4
o
X
IA
K
K
•
•4
n
•
N
K
0
0
K
0*
•
o
0
O'
0
m
*
•
•
r»
•
0
•
0
eo
•
K
0
•
CJ
%
0
0
8
1
1
1
1
9
1
44
>
*
1
h.
tn
O'
•
O
•
Cb
•
o
o
0*
•
©
K
0
O'
•
o
*
•
o
•
©
9
•
1
I
•
1
2
44
>
a
t
9
9
1
tn
o
o
•
1
44
9
44
o O 3
a w 3
C
a
i
•
*
z
*
*
u
o
J
*
3
a U 5
-j a - j
<
o
a
3
e
cj
cj
44
n UJ in
z a UJ
3
.J
2 *■*
UJ 3
•4 X
m
w o
44
0
1 X
«4 UJ X
*4
M
a
X 3
Z
j
M
-j a a
ii n
m ii
in o in tTi
o 0 3 ♦
ij u
\
o
o
a <
IX a
UJ
V) 44
X
u
k*
3 z
in a H
Z u *4
a '1 ■* >
•n a .
o
N
0
N
o
X
V)
•
O'
N
CJ
•
N
0
•
0
m
0
•
o
CJ
CJ
•
CJ
CJ
o
n
n
•
o
0
0
•
•
o
CJ
9
0
•
0
•
N
K
0
N
*
•
C
K
v4
m
•
0
CJ
CJ
•
r»
CJ
•
CJ
CJ
9
O
o
•
m
•
o
c*
H
0
N
•
CJ
•
CJ
O'
0
o
0
•
W
0
•
r>
CJ
N
•
0
44
44
44
9
1
1
•
0
•
CJ
0
CJ
•
0
0
0
c*
K
s
CJ
•
K
N
9
n
0
■
1*1
CJ
CN
J
•
s
*4
•
CJ
Cl
0
0
O
CJ
•
0
o
K
0
•
♦
m
o
0
•
CJ
•
•
0
n
9
•
CJ
*4
tn
3 tA
< j .J - j -J < UJ k- y
o UJ 3
X o a c a
j a a a a • UJ o _ j
<
♦ > UJ a
> >
o 0 in Ift in
<n in in in • x
< a X
o a a o
a a X X
-1 UJ u
u u u V n 3 X
3 V X Cj
44
Q a 3 o z
o 1
z z Z z < o u. o
< < < < 3 z <n <1
UJ z •4
a ao 9 OJ
•
0*
•
0
0
•4
>
w
UJ
3 < K
< UJ m
a UJ UJ UJ UJ Z
j
in
N
•4
X
in 0 0 0
0
B
1
z j
z tt
UJ
UJ * o UJ
x « z c
3 O 44 z
44
UJ a*
Q *4 z
o z
o
O 0 *4* 3
o
UJ
UJ UJ UJ u
in UJ X
X o X
< <0 Uj
H Uj o UJ
•
0 a
tn in
UJ in
. j UJ
UJ
UJ u 3
M
UJ -J UJ . j
o J
j
to .3
o z 3 z *4 Z
z 0 Z < k» u
<
<
< z
0
CJ
1
44
o
0
tn
<n
UJ
X
a
UJ
X
in
a
o
44
X
UJ
a
a
M
a
0
a
u
9
9
1
9
9
•
9
9
•
II
I
9
O
0
a
0
©
a
©
0
V
CJ
©
CJ
0
•
0
0
9
0
•4
1
*0 tf>
0 0
♦
K
CJ
CJ
*
•
9
0
9
n
0
t
44
0
I
0
i
•
tn
•4
1
0
CJ
1
o 0
in
* «
eo
N
o
(V
CJ
0
•
•
n
0
•
0
0
CJ
0
t
•
CJ
N
•
9
r*
«
CJ
0
9
•
0
0
9
•
II
9
♦ 0
o
*9 «
tn
m
9
II
9
V
9
o
3
<
O <3
3
3
< <
0
N.
•
CJ
*
I
CJ
cj
t
•4
>
<
a
z
3
X
tn
o
O 13
3
3
<
<
O
3
<
0
o
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction
O ALH AR T
DATA
SET
C05=L BAND CROSS POLE 5 DEGREE LOCK ANGLE (08 )
C40=L BAND CROSS POLE 40 DCGREE LCCK ANGLE (DID
PU5=P BANO CROSS POLE 5 DEGREE LOCK ANGLE (OBI
P4 0=P BAND CROSS POLE 40 DCGREE LCOK ANGLE (Dill
235=NSOO 1 BANDS 3 . ‘V. AND 6 <10*41-41 WAITS C H t tl-a )
PV1*PERPENDICULAR VEGETATION INDEX ( 0 1MENSI UNLESS)
TVI^TRANSFORMED VEGETATION INDEX ( 0 IMfcNSIONLESSI
SMO1=0-2 CM VOLUMETRIC SOIL MOISTURE (X)
PERIODS REPRESENT MISSING VALUES
CROP=M ILLET - - - - - - - - F IELD
MONTH
CAV
COS
C40
P05
P40
215
Z20
2 35
PVI
TV I
SM01
prohibited without perm ission.
03
AUG
14
-1 9 .8 0
- 5 1 .3 0
10.40
-2 0 .7 0
1 .37
8 .0 9
•
2 .149
1 . 100
9 .4
03
AUG
16
-1 9 .5 5
- 5 2 .6 8
20 .90
- 6 .9 6
1 .T6
11.40
16.10
3.177
1.110
6 .3
03
AUG
16
-2 0 .2 1
- 5 2 .0 0
2 2 .46
- 5 .9 6
1.97
11.10
16.90
2.863
1.095
6 .3
03
AUG
18
-2 1 .2 2
-5 4 .1 3
2 0 .6 0
- 7 .5 9
2 .1 8
12.98
18.89
3.458
1 .101
2 .2
04
AUG
14
- 2 1 .5 8
-5 3 .9 6
•
.
2 . 16
12.29
.
3.192
1.096
2 .9
04
AUG
16
-2 1 .9 0
- 5 1 .0 7
2 2 .80
—6.4 8
1.C4
12.07
16.61
3.390
1.112
5 .8
04
AUG
16
-2 2 .2 2
- 5 3 .6 9
2 2 .63
- 5 .7 5
1.*i6
11.36
16.79
2 .986
1.098
5 .8
04
AUG
18
-1 8 .9 1
- 5 2 .1 5
15.71
-1 2 .3 0
1 .83
11.20
16.05
3.031-
1.104
1.7
186
187
o
*
0
M
>
M
>
&
OJ
•
0
0
«•
0
•
9
X
©
a*
•
•*
•
0
©
0
•
K
mrn
9
N
0
•
♦
9
•
x
mm
01
•
Ml
•4
•
<
©
•
X
0
N
X
9
X
mm
•
♦
•4
0
©
«
m*
mm
•
•
•
0
o*
K
9
fm
O
•
♦
•
0
•
♦
©
•
©
0
•
N
o
o
•
•
©
mm
n
n
0
•
©
(Vi
9
©
N.
•«
•
®
N
mm
mm
9
K
•
•
•
•
>
K
©
0
•
*0
©
0
•
♦
0
•
0
♦
•
m*
♦
0
©
0
OJ
01
0
N
©
0
01
0
OJ
0
•
•
•
9
0
©
0
•
mm
0
9
©
m
©
•
♦
©
©
•
0
mm
0
•
X
OJ
mm
9
0
•
©
•
Oj
CJ
©
0
9
mm
art
9
•
0
N
«
♦
K
0
©
01
mrn
mm
9
1
4*
•
♦
0
*•
0
mm 0
9 o
Q
sm
*4
OJ
mm 1
am 9
a U *
*
o
X
UJ
UJ
UJ . J UJ J
J 3 J •3
o z o z
z < z X
<
X
V
if o V o
a 0 u o
a
H
UJ
0
<
l»
<
Q
►
s
X
r
j
<
o
J
o
J
«j
u
0 .
0
M
0
0
iA (A 0
(A 0 0
a O a
cc a x
u 0 u
a
z
x
Q
Z
o
a
mm
0 0 0
Z a 0
u
z
o
a
z
*
o
0
3
X
mm
0J
•
•
e
n
K
0
©
•
©
•
©
0
9
o
«
N
OJ
•
9»
•
0J
•
•
o
«rt
0
K
t
©
•
«
o
•
OJ
mm
mt
9
•
mm
mrn
0
mm
fm
04
•
o
9
9
0
0
1
•
0J
•*
mm
mm
O
0*
♦
mm
fm
•
o
•
•
in
•
m
mm
■«
0
•
0J
K
OJ
•
OJ
m*
0
0
©
0
mm
mrn
©
•
O
«•
9
t
«4
art
•
0
©
♦
9
•
O
•
•
9
9
o
©
9
•
0J
•
0J
mm
9
fm
•
mm
m
•
OJ
©
«
•
OJ
mm
©
0
♦
•
0J
3
w ml
mm 0 X
9 •4 >
o
X 0
X
Z
0 ml *4
m
m0
o
z o 0
M 0 mm
X
z U
o M k»
a Z
u* »> 0 .
< u 0
0 z 0
0 3 a
0 J a
UJ a 0
> > a
mrn
z
a
o
u
O
♦
a
II
a
♦
0
•
01
1
n
N
•
M
1
e
•
©
*
0J
N
N
•
•
0
mrn
1
o
ao
•
o
fO
©
•
n
f
0
9
h*
r»
•
0
©
o
0
O
&
u. o 0
0 II a
rn
z m
< 9
a X
o
X
w
tm 0
II
■4
>
k
M
0
o
•
9
OJ
>J
©
9
9
K
0J
9
N
•
mm
•
r•
0
•
n
art
OJ
0J
0J
0J
0J
0*
0J
o
0J
<r
l
4T
•
0
9
mm
9
9
O
«
•
©
n
1
o
©
•
©
«
•
04
*
•
0
mm
•
K
mm
m
♦
•
K
n
•
•n
1
•
#
©
•
©
oj
1
mm
•
1
o
m
9
mm
o
9
fm
««
©
mm
9
*
0
•
0
1
0
•
r*»
♦
•
m
m
OJ
0J
9
04
1
1
N
•
0
I
•
©
o
a
u
0
• X
0
o 0 X o X 0
X
0 0 u o
a
0 0 3 Z
O U 3 04 *4
mo
o Z m
X
s
o
OJ
N
e
Z M
•4 O «*
z X Q
0 z
0
•* a
• j ,j a a o 3
M i> ii u u 0
0 © 0 e 0 a
O x a x Z •i
u u &> a tl m
0 >
9 a
X
o
•
«*
Q
1 X
UJ
0 4* 0
UJ UJ UJ UJ ♦ O
UJ a u zc • z
4* c X o ©
o UJ © aJ
UJ o UJ O
z
o
o
o
©
c 0 •4
H
0 ♦ 0 X
O X
UJ UJ UJ UJ z
X 0
J J J
o Q 0 Q
0
• 0
a a a a
♦ >
j
©
9
•
9
0
0 0
J Ut
z ml a*
0 0
<
*
0
0
•
*«
»• z
mm
9
ce
•
©
K
♦
•
0J
0J
0
O
•
0*
mrn
m
m
1
*•
©
0
•
9
•«
•
9
•
mm
d
0J
mm
9.
©
t
0
mm
0J
0J
©
0
0J
04
0J
©
N
9
•
•
O
••
mrn
0
N
•
OI
9
r tf
t
9
9
mm
K
0
•
K
♦
n
♦
•
m
l
♦
•
♦
1
0
o
mm
•
©
0J
9
m
•
o
9
•
«*
•4
1
1
Ik
0
•
0
«4
1
9
9
•
9
mm
1
K
•
1
mm
1
©
fm
•
©
o
o
0
«
«
•
«
9
•
©
0
•
1
K
©
•
9
1
©
•
9
•
1
mrn
0
•
mm
9
•
04
mrn
•
9
9
0J
K
n
•
N
•
0
0J
N
N
•
©
•
Q
n
•
*
©
©
•
K
««
•
©
•
O
*0
0J
9
mrn
•
•
0
01
■4
©
9
•
OI
9
1
mrn
9
•
9
mm
1
«
OJ
K
•
mm
9
1
9
0
•
©
«4
1
•
0
M
>
<
o
I
•
1
1
z
*•.
z
1
Q
X
i
o
I
1
m
w
9
•*
©
3
<
9
©
*
■*
•"
©
©
©
X
<
<
—
mm
©
<
3
3
©
©
©
o
o
o
3
©
3
9
*•
♦
©
©
©
<
<
<
3
o
©
O
mm
mm
mrn
3
3
•4
©
••
9
©
©
©
X
X
X
04
9
mm
mrn
©
<
3
mm
O
*«
©
3
3
•a
«4
•4
3
9
9
"•
©
•4
©
©
©
<
<
X
OJ
Oi
"•
mm
0J
■4
3
3
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3
4'
•i
9
4*
i
X
a
a
>m
a
<
X
UJ UJ UJ
•J J J
a ◦ o
*
a
o
a
a
a
a a a
Q o o
S X a
u u u
a
o
z
<
o
J
It
a
9
U
a
o
z
<
a
j
ii
o
z
<
o
a
m
9 a
« o
UJ a
X
u
a
9
O
I
I•
z
a:
o
VJ
n
a,
a
a
w
r»
•
tn
•
9
-o*
N
h»
N
•
>n
a
4
a
•
in
©
N
#
•
9
o
fw
4
»n
4
o
44
•
o
Of
N
©.
•
9
CJ
a
•
o
44
a
rw
4
9
in
CJ
•
o
m
n
9
tn
m
*
♦
44
tn
N
C
f*M K
N
©
9
<0
©
a
N
4
44
atf
CJ
c»
a
4
9
9
C
J
0
•
a4
©
K
•
♦
•
0
o
4
a
CJ
4
9
44
0
a
9
e
4*
4
n
94
9
CJ
9
4
©
♦
9
4
9
CJ
44
4
o
K
K
4
9
9
N
t
CJ
K
n
4
©
©
o
4
C
J
48
O
4
©
©
CJ
♦
N
4
44
44
O
9
••
©
m
4
•n
CJ
©
CJ
4
n
N
9
4
©
N
4
a
©
N
4
©
4
CJ
©
4
o
a
K
4
9
44
©
♦
©
O
4
in
m
n
CJ
4
CJ
44
•
©
4)
44
1.173
©
4.602
N
a
44
4
9*
•
a
M
N
O
4
e
9
9
4
o
tn
44
4
m
44
9
4
in
l
CJ
!A
4
a
9*
•
4
«4
9
9
4
m
OJ
I*
©
4
«
CJ
9
44
4
»n
©
•
9
©
N
4
44
♦
1
©
4
m
<e
♦
•
♦
♦
1
a
N
•
in
■4
|
9
#•
a
n
C
J
1
m
9
♦
4
N
a
tn
©
o
9
©
in
©
4
■4
4
44
4
©
p>
o
4
o
n
44
f
o
?l
N*
4
«
44
1
9
at
•
o
9
o
9
©
4
4
©
1
4
CJ
1
9
0
•
9
♦
o
<u
•
o
n
r*
wo
•
o
o
9
Q
4
o
CJ
9
4
a
CJ
h»
4
m
1
•
1
4
Cl
•
©
♦
•
44
9
9
o
•
a
0.C9
“
•
a
o
a
9
♦
w
1
9
•
O
CJ
S
•
in
9.35
4
a
44
a
»*>
N
tu
f X
e UJ
H a
o
a
a
a
©
-9.43
« .
©
a
o
AJ
1
>
<
x
k-
Z
U
X
4%
.
o1
<
♦
a
•
h»
m
f
9
I
a
4
>
■4
1
©
a
®
4
CJ
CJ
CJ
CJ
o
4
9
N
Cl
9
a
N
o
o
*
©
a
o
9
CJ
4
m
©
N
©
4
•4
©
4
©
a
s
©
4
44
CJ
f)
9
4
©
CJ
01
©
■a
7
f
CJ
1
4
in
CJ
*
N
»n
in
a
a
a
9
0
7
1
©
©
©
0
9
o
«
9
mt
<*
4
CJ
CJ
N
4
o
a
a
18.11
UJ
UJ
a
0
UJ
a
t
1
«
II
a
N
-42.02
*
w
a
UJ a
a UJ
3
w -J
a <
(4
o
X •o
z
-J M
M a
o a
a «4
X
u
•4 H
s Z
k- UJ
UJ a
X UJ
3 a
a
d UJ
> a
<0 >o
©
©
■4
17.77
X
K a
u Q
o j
J
UJ
UJ u UJ
(A u X
X 0
< 0 UJ
h- UJ a
< a
9
o
0 «
3
A
«
44
■4
-»
4
a
«4
©
«
h*
(A
44
«•
4
©
12.82
■4
>
a
0J
I a a
3 w a A
Q • UJ a
^7 » - j UJ
X z j
UJ w a z
44 o
a a ■4
z w z a
«c
UJ z
< X 52*
X X ■4 X
a
o *4
W Q
0
-j ♦
I X
UJ
UJ K
UJ • a UJ
» z 3
X
3 c M Z
M
U ■8
Q
z
a z
9 o 44 Q
k- •4
*
a < k»
UJ z k- <
< UJ k0 UJ
o
• UJ 9
a
> UJ
a
>
• X
a
a
□
< O
X
J UJ
u a 3 X
o U X
Q z
a
Z < a u.
a
< © z
UJ z
a
*«• & <
o
a
a a
n o UJ |4
9 a a *
♦ z it
>
a 1
n •» >4
•
in
•4
>
M
>
3 4%
© O s
3
Q
+*
w
UJ
U -J UJ
J 0 .J
9 z J
z < z
<
<
CJ
12.7
188
3
9
3
9
3
*
CO
N
a
I
OJ
I
©
44
o
3
9
3
<
<
*•
o
oj
o
o
3
<
o
3
<
9
3
<
O
O
<
o
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
<
0
3
<
9
o
189
9
•
0
©
■
N
O
•
04
a
•
o
a
•
0
a
•
0
mm
M
>
0
♦
9
•
O
0J
0
9
•
O
0
0
9
•
O
0
0
* •
o
K
♦
9
•
O
©
0
9
•
O
0
0
9
•
O
a
o
oi
.
o
04
0
ml
•
o
9
OJ
o
•
o
«
0
0
•
O
0J
OJ
0J
•
o
0
mm
•
OJ
K
OJ
•
0
04
*
s
•
0
0J
0
0
•
N
«4
9
•
N
0
M
>
a
~
C
M
I tn
B
w
tn
9 • UJ tn
^
*t ZJ UJ
j —
UJ
Z X
OJJ
UJUJ u o O
•m
0
«n
5
-J
3j
0 Z o z ~ z <n u j <n
mU
J z s UJ
z < z < p✓
▼
••J 3 9
*
*
•«
X *• J
•»
<
o
a
K
M>
Q ~ C9 tn
u
o
—
o
X 19
N» J UJ JUJ v1 XUJ XJ JM
z
t li UJ UJ UJ UJ * 9 U
0 UJ S UJ a • z 9 - t n
o
a
z
a
t
n
0 e ••
< 5 u o u — —tn—
H UJ a u
j9 ~ Z
*
<
azu
o 0 » o - t0 aO o♦ a
♦
w— 3 Z
0
M
9 < N H Ul
UJ UJ UJ Z M < u i t n
a
Z UJ
< -J a
J a
-I - J < UJ Wuj 9
a
1 a
t9 J
a 0 u
j< a a. a O
j ai2ii
> >1z
Q V) tn tnV)
tn tn tntn • a
o
o o o
<
9 z tn
cc o
U 9
a Xw tn 3-J UJ
w ur u
x
a
a : o j us
9 9 9 9 9Z 0-» a I a
Z Z< Z< Z < 9 U o UJ
a o <9 0 UJZ z - a
- a., < ©
^
aae
a s zV)
H If II o U
© tn o <n aJ ft
« o ♦ z »l
0
a
N
0
01 am
0^ 04
9
•
04
0
9
a
mm
m
o
©
.
o
o
•
0
04
04
©
•
0
0J
a
K
«
0
01
N
04
»
0
0
0
•
a
0
0
•
a
a
0
•
a
K
♦
•
OJ
Q
0
r*
•
0
OI
•
0
0
0
a
0
9
04
♦
0
»
o
a
a
9
9
0
O'
CM
0
9•
O
a 0
01 o
0
©
K
a
<r
«
9
•
04
0
01
0J
a
0
«
O
0k
0.110
-
o
s
0
OJ
a
•
9
•
0
0J
0J
N
mrn
m
a
OJ
20.35
I
tn
•
0
mm
N
o»
■4
•
a
0
0
0
0J
a
e
0
0
•
9
mm
•
0
•
0
9
9.26
9
<
o
0
9
s
r*
0
0
•
•
a
o
•
♦
m *
o
OJ
0
4.16
I•
V
I
«
•
II
•4
•
aft
o
a
u
©
♦
a
0
O
a
a
«
0
OI
0
0
9
0
t
I
1
1
©
a
o
a
a
•
e
0J
•
e
0
e
mm
•
0
04
6
mrn
0
0
o
0
04
0
o
«♦
•
•
•
o;
•>
«
N
4
1
O
0
1
mt
04.
01
>
0
•
•
«4
*•
e
K
•
or
mrn
mm
1
1
O
•
0
O
•
0»
**
01
9
0
o
j
N
<0
e
cv
I
(0
<0
0J
7
7
7
0
0
0
OJ
0
a
04
0
3
<
0
3
<
9
04
a
♦
0J
19.27
0
0
0J
OJ
9
*
■
04
•
0
0
9
a
K
•
•
*
OJ
OJ
OJ
1
«>
1
a
1
0
tn
9
a
•
s
<
-17.89
t
UJ
AUG
uj
Q
04
N
1
0 X x n •+
in >
a
9
at
*
I
■9A
0
o
M
M
•
in
♦ 0
«
n 9
mm
I 1
0
94
n -n
«4
I I 7
T
1
1
1
«
a
•
i
t
i
i
<
m
m
Q
Z
z
a
X
a
ji
u
0
3
0
0
3
<
0
3
<
**
0
3
<
0
*
3
<
N
0J
*•
—
0
0
<
3
<
3
mt
0
3
<
0
3
<
0
3
<
—
*•
0
3
<
0
3
<
13
►
IM
9
n
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
190
*
n
•
mm
<6
a
01
10
a*
•
o
o
«
o
•
o
N
s.
o
•
©
01
0*
•
0J
•
n
mrn
e
OI
•
o
OI
a
in
•
tn
10
©
•
a
*
*1
•
a
m
m
tn
tn
a
a;
•
04
V
•
*
ro
0J
•
tn
*
*
•
*
t
I
I
04
•
O
•
f0
<0
•
©
CJ
oj
01
©
•
o
9
mm
•
K
•
01
n
n
©
N
04
©
o
O
a
•M
©
OJ
*
•
o
©
mm
•
01
•
a
o
o
•
©
*
a .
mm
CM
04
0J
<o
K
•
4»
0>
©
O*
a
•
*
mm
•
•
mm
CD
*
a
©
•
©
•
CO
a
a
•
o
9
a
a
m
©
©
n
01
•
o
•
o
O
a
a
©
•
N
©
m
a
o
K
N
o
©
•
o
•
04
m
•
•
04
©
•
©
a
•
01
OJ
•
©
<M
©
•
••
OJ
©
©
*
*
a
to
O
•
o
c
©
•
a
9
0j
a
a
oi
•
0J
©
04
2 .0
N
©
10
©
•
O
mm
a
m
•
*
*
©
«
O
a
>
mm
mm
*
m
©
•
©
a
IO
•
0
a
*
•
©
N
04
OJ
0.931
9
•
♦
0.012
F*»
•
*
28.31
O
t/>
X
© o m ZQ'J)
w — — {/>•*
z W X UJ.</>
< J J J -J « UJ N* X UI
Xo o o o
13 III 3 S
^ Q. & 0. a *•
0 (A 0 © <
n
> > a
tn on in •ft • a
o o o o 01 x o x u»
j
g
y f al
s
ss
3
g u u V 01 a x
o u a oi —
Q O O a z - o i l
z z z z < a u. o uj
X X < x a z ca ii a
(DOC CO UaJ z< *o•
S UJ U i UJ UJ
I
I
•
I
UI '
a
x
o
ii
a
a
a
u
•
OJ
♦
N
n
o
v
a
!> S 5
g
j& a S
M >1 II II o
•
CM
•
01
vW
1
N
mm
9
•
O
04
o
•
•
10
•
••
N
04
01
OJ
a
m
*
in
I
a o in o tn a h
o * o * z <| -4
oi
oj
in
I
«
1
oi
a
*
©
*«
i
SI
CO
m
rn
mm
0
mm
«
•
•
a
«%
l
OI
01
a
a
o
a
•
a
©
©
o
a
©
N
•
a
FO
K
a
•
K
a
•
i
I
1
7
o
o
•
a
a.
a
0
©
a
o
•
K
o
•
•
s
o
1
1
©
o
•«
N
•
N
0J
Of
C
o
•
OJ
l
1
am
o
e
•
in
in
•
OJ
•
*
•
•
a
O
V*
•
©
•
OJ
n
•
o.
0»
a
•
0
04
•
0
01
©
•
0»
m
tn
0
o
0J
a
o
•
••
n
mm
•
04
•
«
CM
01
©
0
1
«
<r z x
UJ >■ to
(0
mm
•
0
m
1
•
am
9
04
©
o
a
mm
n
•
a
a
1
0J
•
©
a
1
•«
f»)
N
m
•
a
•
a
04
a
•
©
o
©
a
*
a
mm
I
o
•
Of
o
Of
04
OJ
N
a
•
Of
©
♦
I
a
1
•
to
a
•
4.13
N
«
•
s
-23.34
SOS
<17
(JU
^ Uj O lu Q ^ Z
X
o z u
XO
©
o
©
© a - o — :-a a w*
K- X Z
M
o < >- H UJ
01
14.94
~ o
O J o
X ©
J
J JT X
h>
UJ
uj x
z
UJ ^
uj uj ui uj UJ • O UJ J M
Z
O
n
|/l
v) uj e uj c *
0
49.69
~ I (A -*
— a -• a ~ (A (A
D O © Q • UJ <A
» • J UJ
Q ** © %
mm
v
X Z J a z x
Ui
UI © M
O —
UJ - J UJ -J
j J J © in (A —
© Z © z » z m uj <n
Z < * < ►
- u j z a uj
< z uj a a
X
<
X X - x K J
« M U> «
X O * o
o o u o -* v a « >
9.79
CJ
t j u a. a ii «• >
>
in
m
M
4.
k«
*
<0
0
n
01
o
©
X
u
z
o
I
I
3
uj
mm
o
a
a
I
^
oj
I
*
•
n
<0
1
1
mrn
*
mm
©
a
©
a
x
a a
mm
mrn
•
•
a
0
a
OJ
•
a
04
0*
m
•
a
©
•
1
l
1
1
*
a
a
9
©
a
©
©
3
3
mm
©
©
©
©
©
©
<
x
<
<
<
<
a
a
a
a
a
a
<
04
9
mrn
a
n
X
a
a
a
-
0
1
©
©
©
©
x
x
k
mm
a
N
a
©
-
"1
o
•
a
I
©
a
x
a
x
n
I
a
o
OJ
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission.
I
a
O
O
J
I
a
mrn
0J
191
APPENDIX C
GUYMON DATA SET
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
192
©
•
N
1
-
K
K
N
•S
>
■ft
>
a
©
CM
N
X
«
K
—owo—
X
0*
—
z
tz
UJ
©
oo <
> u,
3X
o
ODD
a
a
ii
©
OJ
©
ft
9
CM
•
©
•«
•
—
n
X
•4
•
—
X
—
ft
■ft
©
©
••
•
aft
CM
CM
CM
CM
C
•ft
©
■«
•
9
O
©
■■
X
©
■ft
X
©
■ft
X
c
«k
ft
•ft
■ft
■ft
©
•ft
••
•ft
—
—
X
■•
s
ft
CM
X
9
O
•
CO
X
9
o
©
a
CM
9
9
n
9
n
9
X
©
N
X
©
K
©
CM
9
©
—
©
©
••
©
©
CM
CO
CM
CM
CM
OJ
CM
©
CM
CM
CM
•
©
CM
•
©
ft
©
©
ft
ft*
•
X
9
•
N
•
©
©
•
K
•
©
©
•
X
•
ft*
H
e
X
•
•
ft
ft
ft
•
•
•
•
«
•
•
•
•
N4
O
J
x•
X
N
X
o
Ik
-J
X
Ik
M
4
O
a
u
i
f
a
f
UJ
i
•
i
©
©
©
•
©
<0
a
©
©
o
o
o
n
o
X
CM
n
•
©
w
X
1
X
1
O
©
•
<*>
9
&
o
©
•
•
o
o
o
o
K
ft
n
•
9
X
©
9
X
1
l
1
•«
9
•
X
CM
t
O
CM
•
X
<M
1
o
*
r
•
X
N
t
O
©
CM
•
X
X
1
ft
X
CM
•
OJ
H
OJ
1
I
o
u
©
CM
1
z
o
■«
01
•ft
CM
««
CM
CM
CM
©
©
©
©
©
o
<
©
D
©
o
X
©
o
©
I
7
■ft
©
I
n
X
©
0J
ft
M
X
•ft
ft
■ft
N
1
•ft
•4
e
X
•
X
X
1
O
9
O
X
ft
N
X
1
••
•
©
n
9
m»
©
o
X
ft
♦
OJ
OJ
t
1
X
CM
I
•ft
<4
CM
N
©
o
o
•
(\J
I
'—
X
©
X
1
•
X
1
I
•
©
X
1
Oft
1
Oft
•
r>
CM
I
-
N
■ft
s
CM
CM
©
©
o
X
o
©
o
X
a
>
I
©
I
ft
X
OJ
(M
©
X
X
a
©
©
CM
I
7
©
o
m
K
o
©
©
©
X
©
0-
.
X
©
Q
o
tv
«
f
1
«
I
I
I
.
©
.
— z
z
X te* W
0 UJ tu UJ UJ o *• < UJ
1 J J -J -J z UJ
> a o
O UJ
UJ o
■a & 4 Q.
©
>
© in V) Ui K
a
U) © © ©
a a o o * x a X
a a a a x j u u
u u
© U 3 CM
- O f
a a a o
z z Z z z a u. O
< < X X X z © ||
a a
UJ z —
a x ©
J
J & Q. I/I a a x
11 * HII x uj ^ ©
a »
o
©
•
©
•
o
-• © X
o *• Z
v o z u
~ o o
1
ft o
©© zII
O
X O X
X
0*
N
■ft
o
X
O » *
W O
x ©
* X
z
UJ I UJ X
UJ m Q W J •
C * z a - ©
ww
CM
•
CM
©
-J *■*
X
r>
N
n
•
N
CM
•
©
•
O
>44.00
o -J O
J
j
UJ
UJ UJ UJ
UI
ui a
O
uj a
UJ
o
a
e
x ©
u
O
©
•
<A
•
o
•«
-© —
O~
J
*
UJ O
«a
u©
u 20
o
N
•
CM
©
ft
*
z g z M
xa ¥ U9 O
u o U (9 •
X
X
•
0J
9
ft
9
</) —
©
© z 4 2 IA 2 *0 Uj V)
z < Z < » UJ Z X UJ
X
>• Z UJ 3 O
X
« < —•£ *• ^
gj
©
N
CM
©
ft
>22.70
uj
UJ
Ui « i
M
'N
; CM
ft
•
CM
o
ft
©
ft
n
N
2
i
•
r>
•
S
<\J
a
•
ft
—
•
(DO** UJ ©
2 w * J UJ
V
• z -J **
W
J
•
0*
s
■«
N
AUG
-
•
X
1.155
in
©
•
CM
.<M
X
2.205
.
**
o
.
I
2 9 .0
I
X
a
X
w
z
u
X
©
D
<
*©
o
<
1
<
<
©
o
<
©
o
<
©
D
<
©
o
X
<
©
©
<
o
X
o
d
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
193
A
m
m©
0 a
a
UJ
UJ - j
©
z
<
0
z
<
o
3
©
-J
©
Z
<
X
X o X
u Q u
3 J o
J
J
*
UJ
Ui © Ui ©
(ft tu 3T ©
a O a
< 0 u ©
u. o ©
< o
Q
©
o
©
©
z
o © UJ UI
X J .J J
> 3 o 3
a a. a
3
0 V) tn
in tn tn
o o o
a x a
u <j <j
a
z
<
c
3
Z
*
Q
Q
Z
<
©
j ■J a
m H H
© O ITS
o <t Q
u u a.
4%
A f t <n
tn
a
Q
©
e -J
* z
UJ I a
u 44
5
tn
z tn z
< N ©
K X
* < 44
o * o
44
u
.J a
♦ X
©
UJ
mo
© m
X * z
o « M
© O
Q «4 z
a
M
o
♦ 9 ft<
© 0
© z ©
o < o
a
©
• >
© N
tn
X
• <
a
X ♦ ©
u
tn ©
Q a 44
Z z o
< < z
Q Q ©
a
0 . in X
it a ©
o z a
*
it
a. m
ft >
N a
m
rn
M
44
©
©
3
►- ©
tn <
>
m
ma
X 0
X
z
*4
UJ
a 4<4 ©
z a ©
M tn
z
z u
0 44 w
X z
©
< © ©
w X ©
UJ 3 a
o
a
UJ d ©
> > X
a
UJ
z
a
0
u»
in
z
<
a
N
N
•4
>
©
0
a
CM ■4
t X
O w
II a
x
U
o
X
tn
O
«■
N
•
•
0
©
•
O
i
t
i
I
i
•
•
©
•
N
•
O
ft
•
P»
•
•
•
•
0
*
s
•
♦
I
o
©
•
■4
ft
0
9
•
©
0
I
©
o
o
•
■4
ft
t
o
><
a
X
h>
z
u
X
•
©
V
44
0
0
ft
•
o
•
ft
1
in
0
1
9
44
44
44
""
"*
0
«4
0
•
0
44
©
♦
©
0
4
©
4
ft
ft
ft
©
4
ft
©
r*
ft
«
©
4
4
ft
4
0
4
©
0
0
6
•
P*
ft
•
4
9
•
©
9
ft
ft
1
4
©
©
4
o
tn
♦
1
4
o
o
4
9
«
e
©
4
ft
ft
1
ft
ft
1
0
ft
«4
ft
1
K
4
o
4
•
o
4
ft
1
o
©
4
0
44
4
4
o
o
0
4
©
©
•
♦
4
0
1
1
1
♦
o
©
44
©
4
©
4
48
ft
1
K
♦
O
©
•
4
©
♦
4
♦
1
o
4
o
©
1
f
0
4
♦
©
1
•
©
1
©
4
©
o
•
84
4*
o
O
44
©
•
•
O
0
4
©
•
©
ft
4
mm.
©
9
4
O
4
©
4
©
•
9
4
00
N
O
0
o
«4
0
4
0
•
0
4
ft
«
•
©
N
ft4
ft
o
9
0
4
ft
«
At
o
•
0
-4
0
4
9
1
9
0
4
0
■4
I
•0
1
ft
■4
ft
44
—
1.185
N
48
4
ft
3.361
©
N
44
4
00
•
O
1
o ©
© ©
•
•
0
©
1
9
©
•
©
ft
44
•
0
9
ft
ft
1
©
©
•
0
1
ft
44
©
44
©
•
O
•
4
•4
ft
1
ft
©
00
O
©
©
©
•
©
•
ft
©
«
•
O
•
1
©
-
.44
■
♦
©
N
r*.
ft
ft
©
©
©
©
0
3
*
0
3
<
0
3
<
0
3
<
0
3
<
0
3
<
0
3
<
0
3
<
0
3
<
0
3
<
0
3
<
0
3
<
0
3
<
tn
3
<
0
3
<
0
48
0
«8
ft
ft
ft
ft
ft
ft
ft
ft
ft
ft
ft
ft
ft
ft
ft
ft
ft
ft
*
I
j
1
1
1
1
1
1
1
J
1
•
ft
©
44
0
©
ft
©
N
•
©
•
44
AUG
3
X
©
ft
•
©
•
00
O
4
©
4
4
44
•
©
•4
N
0
•
K
9
l
©
o
0
cm
ft
•
©
e
♦
u
•
o
•
1
in
o
a.
©
©
m
m
Z
1
1
I
•
t
1
1
1
•
©
ft
o
*
&
i
i
i
I
'
t
ft
«
n
ft
0
•
0
©
•
O
i
<
u.
©
<
JL
J
«J
J
a.
o
a
©
•
ft
©
0
ft
•
cm
N
©
X
©
ft
©
•
CM
tn
•
s
©
**
tn
©
UJ
j
z
0
■4
tn
z
UJ
a
M
o
•
N
44
114
*
I*
(A
© ©
CM
♦ ©
ft
in
ft
K
44
4
**
-
>
a
•
ft
N
44
—
N
•
8.99
•
44
N
48
ft
•
0.45
<
c
48 *
ft
00
e
0.82
m
©
■ 4
0
•
-19.82
N
O
0
-3.74
m
48
•
•
>44.72
m
44
•
©
>22.21
m
*
•
©
00
0
1
— ©
o
J
UJ
*4
u_
0
•4
r>
84
0
84
0
00
0
M
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ft
ft
194
Cj
O
z
3
•
•
«
44
"*
o
o
o
o
3
Cl
9
CJ
mrn
mm
•
•
•4
3
44
•
mrn
•
3
9
•
9
•
3
3
mrn
n
M
«
•4
©
3
©
•4
44
•
•
M
CJ
•
9
in
CJ CJ
3
mm
mm
•
CJ
©
•
00
9
•
#
9
•
3
N
9
44
•
44
CJ
9
44
•
**
CJ
o
44
•
44
O
44
«
44
n
o
3
•
3
m
o
9
•
n
mm
•
I
->
&
>
*4 X
is
m
—
©
mm
•
•
CJ
©
•
©
©
»
•“
•"
©
♦
•
•
©
mm
•
K
tn
4%N tn
4* a
9 V tn tn
* o o O +0 UJ tn
•
J
Q 44 o
«0 44 • -jz U
-j •s
UJ UJ X o z M
UJ -J U
J «J u mm o 44
tn 44
-J •3 J 3
’3 z 3 z 9 z tn tU 3
Z < z <
z X U
J
k* z UJ 3 3
<
<
*
X « 44 X »• 3
X o * a N © 44 3 <
44 o »« >
a o u o
*0 a
o -1 □ j a
♦ X
-1
j
X 3
4» UJ UJ « UJ X
Z
O
J UI U UJ UJ «0 O UJ ml M
UJ X U
J a * z Q 44 3
cr o (X o * ■44 Z a 4
3
44 3 4
< 3 UJ 3 UJ o
u o UJ o aa z
X
< a
o z U
0© M
44
a
a
©
44 a z
m tn « ©
z
< 4* >• U
J
a tu UJ UJ tu © Is < UJ 3
z j -J j J z UJ fa* z UJ
> o o 0 a < 3 UJ 3 x
UJ 3 J a.
3 a 0. Q. a
• > UJ O UJ
9
tn tn tn tn
> > X
tn tn tn tn
X
c Q o a • < O z 3
x X X a © J UJ 3 O
u VJ u u
3 X
o
tn W X CM 44
o a o o Q mm o 1 X
O
z z z z z a u.
UJ
< < < < < z 3 £ &
D u (D 03 o UJ z
a < o
-J J a a tn a X X
it II H * X UJ fri 3
w o tn o X a M
© ♦ 9 *
It it •#
>
9
X a 3
4*
m
CJ
co
N
♦
3
mrn
n
mrn
•
n
o
o
©
CD
•
CJ
n
•
CJ
CJ
♦
n
o
CJ
•
9
©
c*
N 9
© ©
O
•
©
•
9
HI
«
•
Cl
9
9
9
•
44
•
©
•
9
•
J
1
1
I
1
»
1
•
1
1
I
1
9
3
mrn
N
©
*
•
o
3
O
«a
m
3
•
Cl
CJ
*4
•
t
•
•
u.
J
<
u.
ml
<
o
X
u
1
1
1
1
1
I
1
s
1
1
1
«
I
1
•
O
•
mm
©
•
O
•
•
t
•
♦
©
•
n
t
o
•
mrn
3
•
•
t
*
V
©
O
o
m*
©
•
©
•
•
©
©
•
©
♦
o
o
3
3
©
•
N
3
a
1
N
1
z
-
o|
u
3
<
CJ
CJ
I
l
<0
•
•
tn
44
©
•
o
•
©
9
«
44
•
CJ
44
•
44
•
9
•
3
•
•
CJ
1
•
44
9
•
m
1
o
3
•
<o
♦
I
44
o
©
•
©
44
©
9
•
O
C|
•
3
•4
I
9
9
1
«4
4*
CJ
t
3
■4
o
o
©
o
3
*
t
©
*
9
o
3
•
N.
©
9
o
a
9
N
o
n
©
•
N,
•
©
•4
■4
1
1
1
♦
CJ
•
♦
1
$
9
©
©
*
4"
N
1
•
O
1
o
3
44
•
•
m
.
9
44
m
44
1
CJ
1
3
O
a
3
O
•
•
O
44
o
•
44
3
•
O
mrn
**
•
m
N
•
3
mm
•
e
K
•
O
O
•
K
n
X
1
•
Cl
©
♦
&
<
>
•
©
©
9
•
•
•
♦
1
O
9
•
*
1
9
I
r*
•
«*
C|
1
•m
•*
Cl
mrn
CJ
44
CJ
o
©
<
3
3
<
<
3
3
O
3
<
o
3
<
o
3
<
o
3
<
<
3
9
3
<
o
3
<
3
3
<
t3 9
3
3
< <
CJ
h>
S
CJ
K
N
CJ
CJ
CJ
N
N
CJ
S
CJ
K
CJ
N
N
CJ
K N
CJ
CJ
•
•
9
C|
•
9
«
m
»n
*4
O
M
3
x
><
o
Z
M
Z
a
X
©
«j
UJ
u!
CJ
CJ
CJ
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
195
in
•
CM
K
•
CJ
m
t
o
*
»
•4
44
•
CJ
tn
ot
•
o
CJ
tn
©
•
o
9
9
•
o
m
9
<0
ot
•
o
©
r*
9
S
4
•
o
•
O
Ot
•
o
4
©
*4
•
o
*
o
44
•
©
a
n
Cl
•
o
GO
n
Cl
•
o
o
O
•
C|
•
•
•
N
0|
C|
•
■4
CJ
4
•
«•
o
•4
M
in
o
•
<4
•
o
s
tn
44
>
H
44
>
a
**
«•
i
•
«
M
tn
tn
4*
S3
o 03
w o
v
UJ
-J UJ
13 . j
z t3
< Z
<
X
a X
a u
J o
J
UJ
UJ UJ
s UJ
to er
UJ 13
Q UJ
O
o
♦ in
UJ
J
o
a
UJ
-J
o
a
a
a
w
UJ
-1
N*
z
<
*
a
O
j
UJ
UJ
a
(J
UJ
O
</)
tn
o
s
V
tn
tn
o
s
u
UJ
J
a
a
tn
tn
a
s
VJ
a o
z z
X <
a as
o
z
<
03
j j
0 11
tn o
o *
u g
a
R
in
©
a
4»
** OI tn
c 1 tn tn
o «4 UJ tn
• j
UJ
« z .j
UJ Z o z
a
.J u
5 •*
3
Z to y tn
< *4 UJ Z
z UJ
X «< 44 T
o » o 44
w
o
o
V
•J m
4 X
UJ 9 UJ X
UJ
o UJ
a • z Q
13 4
Z
M
UJ O
a m* z
44 o z
44 o
o
M
4 0*
< *4
UJ Q ►- <
j Z UJ w
o < S3 UJ
UJ S3
a
• > UJ
>
tn s
tn
a
o • < Q
4
U
J
s
j
S3
D z
t/> S3 a
a Q 44 o
z Z Q u.
< < Z to
CO a UJ z
a <
a tn s Z
M X to
o z a II
4 ii ii ■4
>
a ■n
oj > M
*j a
m
CJ
IM
(0
4*
M
*4
uj
tn
a
3
►*
tn
M
o
Z
uj
3
J
<
>
S3
z
-)
~ 9
o m
tn * •
z
a
►tu
z
s
—
j
a
>
z
UJ
tn
UJ
a
a.
uj
;
i
UJ
cc
<
03
H
0.
o
a
u
e
4
a
•
•
o
in
M
•
r.fl
<n
#
9
•
CJ
4*
•
9
CJ
1
•
•
O
r%
*
o
m
•
■4
•
CJ
•
44
•
m
0t
•
o
•
t
•
o
K
t
CJ
1
•
o
CJ
•
to
•«
•
o
m
•
©
•
•4
•a
•
•
4
T
CJ
•
m
CJ
n
44
•
o
o
44
•
o
44
o
•4
•
o
CSJ
OJ
44
•
o
9
in
•
•
9
44
•
4
N
0t
•
■4
•
n
*
•
•
44
o
CJ
•
9
m
4
•4
•
o
O
Cl
9
©
•
O
CJ
9
©
•
O
9
9
©
•
O
«4
9
tn
a
•
o
9
©
O
•
o
9
in
44
•
o
9
9
44
•
o
9
©
Cl
•
o
•
©
©
•
9
•
Cl
CJ
•
4
•
o
♦
4
O
*
•
44
•
9
9
4
4
•
44
9
•
44
•
CJ
tn
•
44
•
•
•
44
4
9
♦
©
•
O
CJ
CJ
•
O
Ot
•
o
CJ
©
•
©
9
O
•
©
n
CJ
e
9
*»
4
©
•
O
9
4
ot
•
o
9
9
K
*
9
4
©
•
9
"*
9
9
■T
9
•
n
©
©
•
9
9
•
9
9
•
H
•
U>
4
•
.4
wm
in
o
a
S3
t
9
•
4
N
9
•
9
9
•
44
9
44
•
©
CJ
1
•
•
•
9
♦
•
•4
•
•
9
K
CJ
•
4
1
•
•
o
9
•
9
*
•
CJ
44
•
•
•
o
K
•
•0
o
9
•
in
4
4
1
i
9
tn
O
o
CJ
•
tn
CJ
«
O
44
4
4
CJ
1
O
n
•
o
cj
1
•
4
1
O
4
9
K
•
44
•
•
e
9
•
44
•
o
44
•
9
1
9
4
9
N
•
•
•
44
9
CJ
•
9
CJ
1
(V
o
•
N
9
•
9
1
9
9
•
O
©
•
O
CJ
•
N
N
N
•
4
4
O
♦
•
©
4
4
•
1
9
1
©
♦
•
O
N
•
9
cj
O
rs
•
4
CJ
1
•
O
©
•
4
4
1
■
z
z tn
U Q
O
CJ M
1 a
O UJ
ti a.
M*
o
X
tn
1
I
:
o
o
••
•
O
9
•
N
4
1
o
9
•
9
*
1
•
©
♦
9
•
o
•
©
Cl
1
O
K
•
9
CJ
1
O
tn
o
tj
o
o
•
tn
CJ
1
z
D
m
N
4*
CJ
■4
Cl
«4
CJ
m
m
CJ
>
<
CJ
01
m
m
©
00
44
44
44
44
4
44
44
13
13
©
3
<
<
D
<
<
oj
CJ
<0
9
o
♦
u
•
4
♦
m
•
*
CJ
t>
t
n
cj
•
•
4
44
•
n
CJ
1
•
CJ
CJ
I
O
9
•
in
cj
1
O
r*
•
9
cj
1
m
CJ
44
cj
K
K
CJ
Cl
44
44
•
•
CJ
4
CJ
1
44
«4
N
1
•
■*
N
a
CJ
44
9
9
©
13
13
13
D
3
<
3
<
<
9
9
9
•
O
44
N
*
o
i
i
i
i
i
i
i
i
i
i
i
i
i
I
z
*•
z
u
S3
S3
<
<
t3
3
X
CJ
CJ
CJ
13
D
<
13
D
<
13
13
19
13
<
3
X
3
<
D
<
D
<
Cj
' ‘J
CJ
CJ
CJ
CJ
CJ
3
13
13
3
z
Q
J
tu
U.
II
I
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
FIELD
MCNTH
OAY
RUN
p e r io d s
represent
COS
C40
P05
m is s in g
P40
6
AUG
8
2
•
•
•
0
6
AUG
11
1
.
•
•
•
6
AUG
11
2
»
•
•
•
6
AUG
14
1
- 2 3 .7 8
•
-1 *4 2
0
6
AUG
14
2
—2 3 . SO
-4 2 .7 0
.
0
6
AUG
17
1
-2 3 .7 0
-4 6 .6 0
- 4 .2 6
6
AUG
17
2
- 2 2 .7 0
-4 7 .1 0
•
0
10
AUG
2
1
-2 5 *0 0
•*47*80
•
•
10
AUG
2
2
- 2 3 .7 0
-4 6 *6 0
- 3 .7 9
-2 2 .7 1
10
AUG
5
1
- 2 4 .7 0
-4 8 .2 0
-3 .9 4
- 2 4 .3 0
10
AUG
S
2
- 2 4 .6 0
-4 5 .0 0
•
10
AUG
8
1
- 2 1 . SO
- 5 0 .0 0
- 3 .6 4
10
AUG
8
2
-2 2 .1 0
- 5 0 .3 0
•
•
10
AUG
11
1
- 2 2 .8 0
-4 4 .3 0
.
0
10
AUG
11
2
- 2 0 .4 0
-4 5 .5 0
- 5 .5 0
10
AUG
14
1
o
•
0
1
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
GUYMON DATA SET
COS-L BANO CROSS pace S DEGREE LOOK ANGLE ( 0 0 )
C40=L BAND CROSS POLE 40 OEGREE LOCK ANGLE 1081
P0 5=P BAND CROSS POLE 5 DEGREE LOCK ANGLE <081
P40-P OANO CROSS POLE 40 OEGREE LOOK ANGLE (OB)
ZIO. Z IS . ZZS^MMS BANDS 4 . 7,• AND. 9 1 1 0 **1 -4 ) WATTS CM**( - 2 )
PVI»PERPENOICULAR VEGETATION INOEX <01MENSIONLESS)
TVI stransformed v eg etatio n in d e x io im e n s io n l e s s i
S M O I= a -2 CM v o l u m e t r ic s o i l m o is t u r e < x >
- 4 6 .3 0
«
.
10
AUG
14
2
-21 .30
-3 .3 S
.
•
- 2 8 . IS
*
-2 J .S 4
- 2 9 .4 4
values
710
z ts
225
PVI
TV!
SMOI
-
.
.
0.256
0.965
6 .2
0.094
0.942
3 .9
0.054
0.942
3 .9
0 . 109
0.948
20.1
0.1 09
0.9 48
19.2
0 .1 10
0.945
5 .7
0.110
0. 945
6 .0
0.066
0.940
2 .2
0 .0 66
0.940
2 .2
O.ISS
0.950
10 .2
0 .155
0.950
10.2
0 .3 5 6
0.972
0
0.356
0.972
0
0 .042
0.936
5 .7
0 .0 4 2
0 .9 36
5 .7
0.049
0.926
3 .9
0.049
0.926
3 .9
1 .86
.
1.28
.
1.74
.
1.67
.
1.73
.
1.85
.
2 .3 6
.
2.0 0
.
2 .1 6
.
1.57
.
2 . OS
.
1 .89
.
2 .0 0
.
2 . IS
.
2.61
.
2 .2 7
.
4 .9 0
.
3 .6 6
.
4 .7 0 .
.
4 .2 5
.
4 .7 0
.
S .60
•
S .75
.
4 .8 0
.
Reproduced with permission of the copyright owner. Further reproduction
GUYMON OATA SET
C05=L B A N O CROSS P O L E S OEGREE L O C K ANGLE (OBI
C40*L BANO CROSS POLE 40 DEGREE LOOK ANGLE (OB)
P05*P BANO CROSS POLE 5 OEGREE LOCK ANGLE (OB)
P40*P BANO CROSS POLE 40 OEGREE LC'OK ANGLE (OB)
ZIO. Z tS . Z23=MMS BANOS 4 . 7. ANO 9 ( 1 0 4 * ( - 4 ) BATTS C M **(-2 ) S T * * ( - I > )
PVI=P£RP£NQICULAR VEGETATION INDEX (DIMCNSIONLESS)
TVi=TRANSFCRMEO VEGETATION INDEX ( CIMENS10NLESS)
SMO1 -0 -2 CM VOLUHETRIC SOIL MOISTURE (X )
PERIOOS REPRESENT MISSING VALUES
»—
F IE L D
MCNTH
10
AUG
17
10
AUG
14
DAY
RUN
CROP=>BARE — -
prohibited without perm ission.
COS
C40
P05
P40
ZIO
Z I5
225
PVI
TVI
SMOI
1
-2 2 .9 0
-4 6 .0 0
—4 . 0 6
•2 9 .3 5
2 .3 4
2 . 69
5 .7 3
0 .0 * 2
0 .9 3 7
3 .6
17
2
-2 3 .6 0
-4 7 .8 0
m
•
0 .0 * 2
0 .9 3 7
3 .4
AUG
2
1
-2 6 .1 0
-4 6 .5 8
•
.
0 .1 2 8
0 . 950
2 2 .3
14
‘ AUG
2
2
-2 3 .6 0
-4 7 .1 0
-3 .7 1
-1 9 .2 3
.
•
•
0 .1 2 8
0 .9 5 0
21 . 1
14
AUG
5
1
-2 4 .0 0
-4 7 .6 0
-4 .1 0
-1 6 .7 5
1 .5 2
1 .8 1
4 .2 7
0 .1 4 7
0 .9 5 1
2 0 .0
14
AUG
5
2
-2 5 .7 0
-4 6 .5 0
.
.
•
•
0 .1 4 7
0 .9 5 1
2 0 .4
14
AUG
a
1
-2 4 .6 0
-5 0 .5 0
-6 .0 1
2 .3 0
6 . 10
0 .4 7 0
0 .9 7 6
1 0 .8
14
AUG
a
2
-2 4 .0 0
-5 2 .1 0
.
.
.
.
.
0 .4 7 0
0 .9 7 6
1 0 .8
14
AUG
ii
1
- 2 4 . SO
-4 6 .6 0
•
.
2 .0 7
2 .3 5
5 .5 3
0 .1 6 5
0 .9 5 1
5 .4
14
AUG
u
2
-2 3 .8 0
-4 4 .2 0
-6 .6 2
o. ies
0 .9 5 1
5 .3
14
AUG
14
1
4 .4 5
0 .2 6 6
0 .9 6 8
4 .0
14
AUG
14
2
-2 4 .3 0
-4 5 .4 0
-4 .9 1
-2 6 .5 4
.
•
•
0 . 2 e6
0 .9 6 8
4 .0
14
AUG
17
1
-2 6 .1 0
•4 6 .5 0
-2 .2 0
-2 7 .7 0
2 .0 3
2 .3 0
5 .6 8
0 .2 9 4
0 .9 6 1
3 .5
14
AUG
17
2
-2 7 .1 0
-5 0 .1 0
.
.
.
•
.
0 .2 9 4
0 .9 6 1
3 .5
17
AUG
2
1
-2 7 .1 0
-5 0 .4 1
•
•
1 .7 0
4 .6 1
0 .0 9 9
0 .9 4 4
3 .7
I7
AUG
2
2
-2 6 .2 0
- 4 8 .6 0
-5 .2 1
-2 9 .6 2
•
0 .0 9 9
0 .9 4 4
4 .2
17
AUG
5
1
—20 .00
-5 0 .5 0
-4 .9 8
- 2 4 . 57
4 .1 6
0 . 146
0 .9 5 2
5 .8
•
•
.
.
-2 6 .8 5
-2 5 .4 5
.
•
1 .4 3
2 .0 0
.
1 .5 4
.
1 .4 9
•
1 .7 2
.
1 .74
2 .0 2
.
1 .7 6
.
4 .0 3
.
R eproduced w ith perm ission o f the copyright owner. Further reproduction prohibited w itho ut perm ission.
199
I
I
00
o
9
•
z
n
M
>
9
9
•
n
9
♦
©
OJ
♦
9
9
♦
9
•
w
♦
O'
o
o
o
O
N
O
0
9
•
•
01
•
©
CM
•
10
o
•
o
o
(0
♦
9
0
©
s
s
K
©
9
«
9
»
o
•
O
o
o
o
o
•ft
•ft
O
©
N
O
©
f0
o
M
O
0J
o
N
at
0J
©
o
•
Q
CD
9
O
0
o
o
•
•
■
•
0
tn
«
10
©
•
♦
<6
•
*
n
•
(0
•
w
©
N
•
*
<n
OJ
©
•
O
1
1
1
1
1
1
00
(
w
•
•
fa
1
•
00 OJ V ) 00
00
4* O
©Q
o
V
Ui
UJ -J
J o
o z
z <
<
*
* a
o a
0 j
J
fa
UJ
u UJ UJ
V I UJ Of
a 9
< © UJ
fa UJ Q
< o
o
©
m ©
Z
O UJ UJ
x
j
.J
> a o
3 CL a .
©
■ft 9
a Q
a
w
UJ
UJ J
•j 9
© z
z <
<
X
u
o
J
I
(A
UJ
•
-J
•
z
Z o
M
w
(A
v> z
fa UJ
fa z
<
9 9
©
o
«J 00
©
UJ «
UJ UJ ft#
UJ Of ♦
*
x o
9 UJ o
UJ o 00
00
o
o
0 © 9
UJ 9
J Z
a ©
a
•
IA (A (A V ) fa
(A (A 0 ©
•
o a o a
a c
z a *
u t j (J v
VI
a 9 9 9 a
z Z Z Z z
< < < < <
u o 9 a m
J J
II It
IA o
O ©
9 (J
UJ
_j
a
a
a
it
9
o
a
a
n
o
©
a
(A
z
UJ
X
fa
00 a
fa
•
o
fa
UJ 0
a
uj
3
9
>
fa
IA
1
1
1
I
IA
1
>- j
(A ©
*0 >
a
X
z ©
z
UJ X
9 UJ J n
z 9 fa A
Z O (A
M V I ■*
z
z
o z o
• ft o fa fa
fa fa a z
< fa fa UJ
fa < UJ <A
UJ fa Z UJ
© UJ s a
UJ © « j a
> UJ □ UJ
> > a
a
< a Z (A
UJ U 9
j
O
3 z
L i a OJ 00
Q 1 z
9 U . O UJ
Z <A
UJ z £ a
0>
s s z
UJ fa V )
a II
ii fa
(A
Z
Z
II
0
N >
N a
«
(A
v>
<A
UJ
-1 00
Z M
a fa
1
1
I
1
1
1
»
1
1
1
1
I
•
0
o
1
1
00 00
K
©
O
a.
01
N
•
V)
m
>
00
H
O
00
T4
9
in
0
n
9
IA
•
•a
9
i
1
I
i
i
i
i
i
i
i
J
i
i
a
i
1
i
i
i
i
i
i
I
I
•
1
1
II
II
r*
o
0
o
9
OJ
•
«
9
ft
ft
0
•
•
O
©
eo
0
o
•
•
©
0
m
9
0
©
©
•
0
IA
1
m
0J
I
U.
•
00
n
ft
1
0
©
I
©
•
9
©
•
B
10
•
to
ft*
ft
©
CM
1
9
0J
ft
©
CM
©
t
o
to
ft
©
©
I
9
©
0
9
OJ
«
O
pft
•
9
OJ
1
9
m
0
9
OJ
1
o
OI
•
GO
OJ
«
o
r*
•
9
OJ
•
OJ
00
01
00
01
©
K
o;
O
3
o
s
0
K
OJ
1
o
■*
0
K
OJ
1
00
#
00
m
o
3
<
o
3
<
o
3
<
©
3
<
X
OI
X
N
X
PJ
X
X
X
0J
04
0J
t
•
©
0
©
ft
n
0
©
0
o
ft
ft
ft
n
0
•ft
o
OJ
ft
•4
©
•ft
0
ft
ft
•0
9
0
o
9
0J
OJ
■ft
fft
o
00
r*>
a
o
0
o
O
O
o
0
o
•
0
ft
ft
ft
ft
0
0
o
9
•
©
1
ft
ft
04
©
•
h*
©
1
o
©
0
FI
©
•
o
©
ft
9
©
1
o
0
0
o
F|
1
fa
M
•*
•
©
©
©
0
B
©
O
0
o
9
0J
OJ
ft
o
9
9
•
to
IA
K
•
9
•
M
0
ft
0
0
o
o
o
r
©
©
©
•
0
00
*
0
0
0
ft
ft
0
n
04
1
CD
©
0
©
1
|
o
n
*
©
♦
I
a
©
•
9
OJ
1
o
•0
•
©
N
1
o
9
i
•0
OJ
•ft
OJ
04
©
©
©
©
■■
©
3
<
9
3
<
9
3
<
©
3
<
O
3
©
X
X
0J
OJ
X
CM
X
0J
OJ
©
•
*
0
e
©
e
•
• ft
to
9
•
O
CD
•
04
0
*
01
OJ
1
IO
©
•
n
©
1
©
1
N
9
*
O
©
0
*
e
n
K
04
•
©
0
O
m
•
ft
o
04
ft
0
*
0J
©
0
9
©
•
0
O
O
0
Q
©
1
00
9
©
©
©
«
©
•
00
ft
O
o
©
©
O'
m
00
ft
©
©
9
•
o
♦
0*
•
©
00
©
9
•ft
1
1
0
0)
©
f
0
©
fa
m
o
■ ft
o
©
0
©
©
1
e
o
m
>
<
o
0
o
K
0
o
©
•
o
z
3
a
O
©
•
©
ft
00
o
©
ft
OJ
©
s
0
o
©
V
(A
9
U
ft
•
9
9
ft
©
e
•
•
00
•
o
00
n
o
H
«
IA
• e
a.
ft
0m
•
■ft
•
a>
K
©
•
N
©
a.
o
J
UJ
1
ft
•
o
z
fa
Z
CJ
z
I
N
©
O
0
o
9
m
•
•«
O
UJ
a
©
a
II
a
o
a
u
•
O
©
0
o
0
0
0J
1
oi
1
04
I
0J
■0
OJ
■4
•ft
©
©
■ft
N
N
• ft
04
—
9
3
<
9
3
<
9
3
<
9
3
9
3
<
9
3
<
o
3
<
X
X
OJ
X
01
X
OJ
X
X
OJ
OJ
X
CM
04
J
00
<
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
200
I s
m*
O
X
IA
M
>
M
>
a
w
©
tn
CM
N
h*
•
♦
©
•
tA
©
a
tn
0
0
0
•
0
O
O
O
•
O
0
0
0
0
0
0
•
0
O
O
O
•
O
•
•
•
0
0
0
0
•
0
•
•
mrn
mm
•
tn
•
in
CM
•
4
O
O
O
•
O
O
0
O
•
O
0
0
0
•
0
0
0
0
•
0
O
4
©
•
O
O
O
O
•
O
O
O
O
•
O
0
u
©
•
0
0
0
0
•
0
O
•
•
•
•
•
■
O
O
©
•
CM
«
+
O
•
tn
O
•
tn
©
a
©
O
0
in
©
a
O
0
m
©
•
0
©
©
a
O
©
UJ
0
•
O
tn
CO
0
•
0
r>
r*i
M
a
O
N
©
a
CM
a
N
©
a
CM
<
•
O
O
mm
mrn
•
0
•
n
©
©
a
©
n
a
©
O
a
4
*a
©
©
a
O
©
©
©
a
O
©
©
©
a
O
CM
©
©
a
P
CM
©
©
a
O
r>
r»
♦
CM
•a
a
0
«
CM
•a
a
O
9
©
CM
a
O
4
©
CM
a
O
©
©
a
CM
a
©
a
©
•a
mrn
a
O
a
©
©
a
CM
a
- * CM (A «*
a I (A (A
a a a •»
'Jj A .
o ~ o a ~« J UI
w
UI UJ X Z3_ O-JZ ~-*
UJ -J IU
J C
J d u V
i m
t i Z O 2 (A X IA Uj <A
z « z < *• u i z a u i
mm CO
<
<
X UJ 3
-
X
x o x
a a u a
OJ o j
m
1
HUH
O
IA
o< o
uua
O X
o
a
ii
a
a n
*>1 "•
>
> >•
N
M a
m
•
•
•
•
•
•
©
©
•
•
CM
tn
e
m
©
a*
a
mm
m
©
V
a
O
mrn
0
9
9
O
©
K
«•*
O
mrn
O
•
©
O
a
O
mi
a
a
a
a
mm
5.1 ^ U I «C
w Q « >
O
•
N
3
I UJ o
©
m*
X h- J
O
X
X P
«1 UI
J U I -I ui —
X
z
*•
UJ U I U i U I UJ « * a u j j
© j a uj t?
a * z o —-a
a o a UJ o« * • ■Z IA3 •©•
p
>
•
_i o m
2
X
o Q- o
<o
z o
_ o 1 K
o
i< ©
H *- S Z
< N K UJ
z
O U I U I u i UI Q N < UJ (A
X J -J J -J z UJ k X UJ
> 0 00 o < P ul 3 «
UJ
j a
3 a a a a
• > U
I>
1o>
ui
p
(A (A <A tA K
>
a
© © <A (A
a
OOO a • < a X (A
a s s a «• J UJ1u o
u u p w (A u3 a_ o
_
O
QQO Q m o a
z z z z z O IL O UJ
<
< < X < 2 IA II (L
0 0 a O CS UJ z a < o
J J n a (A a a x
H z u j * • ia
<
V)
n;
UJ
s
<
CO
0
♦
a
•
•
•
O
©
•
•
©
10
•
0
mm
•
•
a
•
mm
•
•
m
a
«
CM
1
a
m
HI
•
CM
CM
1
a
m»
9
O
a
O
CM
1
©
a
CM
CM
1
m
*
a
©
I
O
a
O
©
a
©
9
a
©
O
a
O
1
a
O
♦
a
CM
a
O
CM
a
O
©
9
©
a
a
N
♦
I
O
©
a
O
©
9
a
«
CM
0
©
a
O
*A
O
©
a
9
CM
1
O
©
a
♦
CM
P
3
P
3
n
a
0
u
a
Ml
O
a
•
a
•
O
n
•
tn
•
mrn
9
1
1
•
1
«
1
•1
1
9
1
i
•1
1
1
9
I«
O
<0
•
0
0
©
•
4
•
•
w
u
9
ui
0
u
0
©
•
♦
CM
1
CM
•*
•
m
<M
1
0
©
•
ui
AJ
I
Z
3
a
0
cC
•
O
in
1
O
©
•
©
1
0
4
M
9
0
«i
•
K
CM
I
-»
<M
mm
•
>
<
O
♦
©
a
9
4
•
0
©
•
tn
4
t
0
in
•
tn
4
1
O
•a
a
©
♦
9
O
4
•
4
AJ
9
0
«
•
♦
N
1
0
©
•
4
O
0
CM
•
9
©
—
P
3
P
m
mm
a
cm
9
a
♦
CM
1
a
©
O
a
'A
<M
1
•
n
9
a
0
©
a
•a
O
©
a
9
+
a
©
♦
9
•
1
0
9
O
CM
a
4
CM
f
M
«•
P
3
P
3
P
3
<
<
mm
m»
m
CM
CM
CM
mm
a
m
N
O
9
9
m
9
9
CM
©
a
•«
n
9
w
CM
a
9
A|
1
-
r
z
u
3
<
P
3
O
O
3
<
<
<
X
CM
X
CM
X
M
P
3
<
X
<
X
CM
X
CM
X
CM
P
3
«
P
3
P
3
<
<
<
X
CM
X
CM
X
CM
<
P
3
<
<
CM
CM
X
0
UJ
X
CM
X
CM
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
p
3
<
o
•
©
mm
•
©
••
•
©
CJ
•
©
M
>
H
©
©
©
•
O
©
©
©
•
e
iff
N
©
•
O
Iff
©
•
O
©
©
©
•
o
M
>
a
o
©
Q
•
O
o
©
o
•
o
©
©
•»
•
o
©
©
M
•
o
CJ
©
mm
CJ
N
•
m
•
CJ
•
e
©
•
©
*•
©
•
©
mm
N
•
N
K
•
©
©
©
©
©
•
o
©
©
©
•
O
O
©
©
. •
O
©
©
•
o
N
©
N
Cl
mm
m
O
N
(M
©
mrn
mm
t
©
mm
mm
•
O
Cl
©■
©
•
o
N
fs
O
•
o
IS
H
•
CJ
c*
o
CJ
•
©
•
N
•
•
O
©
•
mm
m
n
•
•
©
»
mm
•
•
O
mm
t
m
*
•
•
CJ
o
o
©
©
•
©
♦♦ ©
♦
© ©
©
«
K
O
•
©
©
CJ
O
o
•
Q
N
O
CJ
IP
©
0.077
mm
•
©
3.02
mrn
o
Z
Iff
0.946
201
mm
0
UJ UJ 0
Iff u i z
0
0
a 19 s
< 19 0 19
UJ a 0
« O
O
o
o
© * V) © ©
z
o tu 0 0 0 Q
X J J 0 0 Z
► o o o o <
3 a a a a.
•
19
<n iff iff in K
u i iff if f if f
K
M
<
*
0m 0
<
0
X
S
0
•
*
N
mrn
©
«
o
CJ
•
•
—
•
©
o
•
O
©
Z
M
Iff
Iff
mm
z
z
0
Iff
0
a
>
0
O
UJ
o
c
<D
<
£
a
ae
v
9
I
a
o
©
•
N
1
•
©
r*
•
t
•
•
o
©
•
CJ
Cl
o
mm
0.
m
© •
•
©
o
©
©
©
1
©
©
1
©
©
0
J
©
♦
I
o
o
CJ
©
r%
Cl
o
r*
o
o
©
iff
CJ
•
CJ
«
•
©
o
©
♦
•
O
•
©
I
•
•
•
©
•
•
Cl
1
o
o
•
•
•
O
©
•
©
©
»
o
©
•
©
©
1
o
©
•
a
«
O
CJ
•
n
o
©
1
©
1
o
o
©
•
9
in
i
a
•
•
Iff
♦
CJ
o
CJ
I
CJ
«
a
z
•
©
CJ
«
C|
♦
cj
1
c*
n
•
o
•
n
Iff
o
0
o
•
©
•
cj
mrn
CJ
-«
CJ
19
3
19
3
<
19
3
19
3
<
o
3
<
15 15
<3 <3
©
©
N
©
©
CJ
©
Cl
CJ
1
CJ
19
3
<
19
3
<
19
3
<
0
1«
0
in fff
Cl
N
I 1
©
CJ
1
©
1
•*
•
t
1
CJ
CJ
<0
Iff
CJ
CJ
1
-
z
u
3
<
i9
3
<
mm
mrn
m
mm
CJ
CJ
U
CJ
CJ
o
-J
0
19
3
<
Cl
<
CJ
<
CJ
19
3
AUG
X
H
Z
u
<0 ©
Cl ©
CJ
26
14
3
•
©
©
©
I
o
iff
K
•
©
©
t
o
N
•
Iff
1
©
1
•
•
Iff
Cl
o
•
mm
©
•
mm
1
iff
•
mm
♦
»
©
©
1.31
•
>
0
o 0 a a •
a z X a ©
w U u 0
if f
a o a a a
z z z z z o 0 O
< z Iff
<
<
<
<
0 a n © © 0 z
a < ©
0 j a a. I f f IE a z
n ii ti n 3L 0 *» I f f
© e in o X Q. H
o ♦ o © It n "0
tn m >
0 vj a
Cl >
si a
•*
CJ
©
o
mrn
o
O
*9
J
<
0 3
19 0 a
0 s 0
> > a
z
< Q z if f
0 0 0 a
u
3 X
0 a Cl M
Ma 1 a
19
•
1.12
Iff
0
3
©
if f
•
.
u *
o u a
o 0 O
0
J
Iff
««
N
.
t-
.<
0 z 0 z M
o
0 o
Iff mm
3
z Iff z Iff 0
© *» 0 Z <Z
0m z 0
3
* * © M 3 s»
a
Q M Iff
*0 o M
Q
s* o
0
© *
z
0 • 0 X
0 00 o 0 0
*
m
m
a
oe
z
19 © M z o
M Iff
0 O
3 mm z
00 0 z 19
O
o
-46.30
<
©
o
-25.30
o o a
o w o
00
0
UJ 0 U
0 o J
(5 z 19
z < z
C l Iff
1 Iff iff
0 if f
•
0 0
•
z 0
1
0
m
<
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
202
i
-o
X
V)
i
-
s
IP
CJ
N
H
(/J
a
o
10
UJ
J
9
z
<
X
o
o
H
UJ UJ
U) UJ
a
* 9
M UJ
X c
o
to
z
o UJ
X J
> o
3 a
o
9
ip
a
a
u
o
a
Q a
w a
**
u
-j Li
9 _J
z 9
< z
X
x
o X
o a
j o
~j
UJ
UJ UJ
X aJ
9 x
UJ 9
Q UJ
o
O
X 10
N
9 9
Q **
•
«
UJ X
J u
9
z tn
< p»
X X
CJ a 9
9
J
UJ
UJ
X
o
UJ
o
o
X
UJ
-J
□
a
UJ UJ
u J
o o
a a
in
tn
o
X
w
tn
tn
o
X
u
tn
tn
a
X
u
o
a
o
♦
l
•0
X
UJ
o
z
M
*
*
o
■* z
W o
0» M
X
Q K
Z UJ
X 9
UJ
• >
N
X
• X
X J
0+
tn
m
UJ
I
j
z
a
■■
tn
z
UJ
X
M
u
w
X
UJ
o
z
M
z
o
M
<
H
UJ
9
UJ
>
ii
m
■« >
N >
N
fi
•m
N
a
N
M
UJ <n
a UJ
3 3
•J
tn X
•• ?
a
X 9
Z
j ktf
to* tn
<n
o
tn m
X
u
M K
a z
K UJ
UJ <n
X UJ
3 a
J a
a UJ
> a:
9 X
Ut u
3 X
V> w a CJ
Q M a t
Z o IL 9
X z tn II
CD UJ z
a XO
tn X a X
a U
J tn
X a ii
z z z z
< <
a ©
•J •J a a
11 II n ii
9 o in o
o X 9 x II
9 9 a & n
X <
CD o
tn
tn
UJ
9
z
o
M
tn
z
UJ
X
tn
3
a
■*
a
UJ
a
O
N
i
u
X
<
©
II
X
D
c
9
O
X
a
•
•*
CJ
I
m
o
a
X*
•
in
1
mt
9
1
1
•
1
t
9
1
1
1
I
1
I
1
1
1
1
9
f
9
9
9
1
1
9
o
X
u
©
•
o
X
1
in
o
9
CJ
•
tn
c»
9
Z
3
N
X
*
1
9
I
1
1
i
1
•
•
1
1
I
1
1
•
«
I
K
1
1
I
I
9
>
X
O
S
*
X
l-
z
o
X
0
J
UJ
m
9
3
X
©
N
u.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
C05=L
C40=L
P05=P
BANO CROSS
BANO CROSS
BANO CROSS
POLE
POLE
POLE
S DECREE LOCK ANGLE (OB)
40 OEGREE LOCK ANGLE (OB)
S OEGREE LOCK ANGLE (0 0 )
mm
ci (A
a t V) (A
J (A
o wU
mm • J ui
mm
» z
Ui X o z *
J u
o
(A M
o
z <A £. (A ui V)
< k- Ui z s U
J
z Ui 3 3
X < M X k» J
u B o M IA <
mm a w* >
Q
wo
♦ X
X U
z
U
J « uj X
UJ Q UI -J M
•*
(A
2 * Z a
O ♦ M z a (A
Ui O mm U)
Q
z
X
mm o z u
O M o M k*
♦ 9 K Ma Z
J
< k- k* U
Ui o >- < UJ IA
J z Ui )• X uj
o < o u 3 2
.
a
Ui o J Q
• > Ui O U
J
<n H
> > a
tn
2
o • < o X (A
2 ♦
UJ u o
VJ
o
3
«Au 3 CJ M
o a
o « a
z z o u. 0 U
J
< x z <A II Q.
o a UJ z •«
&< O
a (A a 2 X
it X Ui M V)
o X a H
♦ tj ii ■*
% '/> >
Cl > m
Ma
U>
N
•
I
•
••
N
B
c
Z
a
a
u
ii
a
a
a
w
s
B
O
*
a
e
•
fA
•
<n
o
a
©
•
N
1
1
0
•
1
P
o
«
u
1
1
IA
O
V
B
B
f
1
•
B
B
1
1
f
I
I
1
1
1
1
i
1
l
«
1
•
i
8
I
B
•
1
1
1
1
1
I
1
1
I
z
3
IA
•
«•
f
•
•
©
•
©
«
O
. •
*A
•
o
•
IA
r>
•
•
CJ
•
•
1
1
©
•
Cl
•
in
•
*
•
©
•
•
1
<0
n
V
•
o
cv
•
>«0
B
•*
Cl
•
IA
•
O
•
eo
9
•
X
IA
•
♦
•
*
m
1
«•
CJ
m*
<M
M
B
Cl
Irt
tn
m
n
in
in
n
o
3
<
o
3
X
o
3
X
o
3
X
o
3
X
19
3
X
©
3
X
U
u
CJ
CJ
u
n)
u
IA
U
X
U
St
>
X
a
T
kz
2
O
3
<
a
-j
UJ
M
11
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
.....
I ELO
HCNIH
day
RUN
——
1
1
1
1
1
1
1
1
fl
Reproduced with permission of the copyright owner. Further reproduction
ZIO.
GUVMON DATA SET
C05=L BAND CROSS POLE 5 DEGREE LOCK ANGLE (OBI
C40=L BANO CROSS POLE 40 DEGREE LOOK ANGLE (OBI
P05=P BAND CROSS POLE S OEGREE LCQK ANGLE (OBI
P40-P EANO CROSS POLE 40 DEGREE LOOK ANGLE (OBI
Z15. Z25=MMS BANOS « . 7 . AND 9 ( I O » * ( - 4 l MATTS C H 4 *(-2 I S T * * ( - l l )
PV1 = PERPEN0ICULAR VEGETATION INDEX (01 MENSIUNLESS!
TVI= TRANSFORMED VEGETATION INDEX (01 MENSIUNLESSI
SMOI - 0 —2 CM VOLUMETRIC SOIL MOISTURE (X I
PERIODS REPRESENT HISS TNG VALUES
CRCP= SORGHUM (PARALLEL ROWS TO FLIGHT LINE)
C95
C40
P05
P40
prohibited without perm ission.
AUG
2
1
-2 6 .3 0
AUG
2
2
- 2 6 .5 0
- -4 4 .8 0
- 6 .8 0
- 2 0 .3 4
AUG
S
1
- 2 8 .0 0
—48.80
-5 .4 1
-2 2 .2 3
AUG
5
2
-2 7 .5 0
-5 1 .2 0
.
.
AUG
8
1
- 2 3 .5 0
- 5 1 .5 0
13.70
- 0 .9 0
AUG
8
2
.
.
AUG
1
1
12.60
- 0 .9 0
AUG
11
2
.
.
•
AUG
14
1
- 2 4 .3 0
.
-4 .0 5
.
AUG
14
2
-2 5 .8 0
-4 2 .0 0
.
.
AUG
17
1
- 2 5 .0 0
-3 0 .9 0
-4 .1 9
AUG
17
2
-2% .80
-3 9 .6 0
.
.
8
AUG
2
1
-2 8 .5 0
.
.
a
AUG
2
2
- 2 8 .8 0
- 3 5 .4 0
- 4 .8 2
-1 9 .C 6
8
AUG
5
1
- 2 8 .4 0
-3 6 .7 0
- 4 .7 4
-2 1 .1 1
8
AUG
5
2
-2 8 .5 0
-3 6 .0 0
•
.
8
AUG
8
1
- 2 7 .3 0
-3 6 .2 0
14.90
- 0 .3 0
.
>22.00
.
.
.
-4 4 .2 0
.
.
.
- 2 2 .3 9
210
Z1S
Z 25
PVI
TVI
SNOT
1.72
1.75
5 .5 6
0.7 42
I.O il
5 .2
•
0.742
I.O il
5 .2
0.7 78
1.‘ 029
10.4
0.7 78
1.029
10.0
4 .6 5
0.813
1.037
6 .0
.
0.6 15
1.037
6 .0
4 .8 5
0.8 26
1.034
18.8
•
0.826
1.034
18.8
0.554
1.052
7 .7
0.954
1.052
7 .6
5 .6 5
1.189
1.061
4 .5
.
1.189
1.061
4 .5
5.51
1 .475
1.102
2 .3
.
1 .475
1.102
2 .3
1 .379
1.098
7 .2
1 .379
1.098
8.1
1 .475
i
.1 n 2
4 .8
.
1.34
.
1.30
.
1.38
.
1.26
.
1.39
.
1.04
.
1.01
.
0 .9-0
•
1.36
.
1.25
.
I .33
.
1.18
•
1.30
•
0.9 2
.
0.9 1
.
0.8 0
4.8 0
.
4 .8 3
.
5 .2 6
.
5 .2 5
205
«
*»
w
*•
-J
0
•j
J
UJ
UI
0
X
a
z
U
I
z
mm
a a
*0
♦
i
0
wU
o
V u
o
z
«
o
tn *
a M>
w a
X
X
UJ X
a UI
<J
0
n
mm
m
**
cn
n
mm
•
mm
0
9
FI
•
o
o
o
•
e
o
0
0
♦
•
mrn
0
0
♦
•
mm
0
0
•
W
«
•
O
0
K
•
mrn
o
0
K
•
•
•
•
CM
♦
•
0
•
K
O
•
0
•
0
F•
0
•
o
s
•
0
•
•
•
O
9
•
O
•
0
H
•
O
•
O
H
•
O
•
0
0
•
o
t
•
•
*4
o
•
“*
•
0
9
•
o
•
9
•
O
•
o
0
•
o
•
FI
9
•
0
o
•
a
O
0
•
0
•
(
1
o
o
•
o
z
o
0
Z
y
a
ii
n
3
a
u
o
z
<
a
in
o
z
<
o
a
in
3
X
CM
1
o U. O
z tn It
UI * m
a < Q
s X X
u
O
mi «•
X
MO
H a UJ *m U>
O X a It
< f (I it
>
3 . •n
N > >•
N a.
o
mm
X
UJ
a
♦
#
0
m
0
n
mm
•
0
mm
mrn
•
mrn
•
l<*
e
o
0
•
a*
—
O
•
a
<
a.
•
—
•
•
•
•
•
•
1
e
n•
60
r»
I
ce
a
0
ii
a.
u
s
u
1
I
CM
•
a
3
X
0
0
©
o
F)
•
F»
FI
o
•
O
9
•
H
1
n
CM
•
9
CM
1
mm
T
*m
z
u
X
0
^
o
3
<
13
3
<
mrn
mm
0
3
<
l
F>
1
C
CM
M —
N
CM
0
1
0
CM
I
mrn
•
0
0
' ■«
«•
•
“•
4
n
mrn
•
«*
—
•«
o
0
•
"
mm
N
•
•*
•
o
9
•
0
•
CM
4
•
0
•
♦
O
•
o
•
4
4
•
O
•
to
0
•
o
•
CM
r*
•
F»
•
*•
—
•
o
o
•
•
CM
O
0
•
0
CM
1
«
0
F>
CM
•
■J
m
•
0
1
1
1
o
Cl
O
9
•
o
♦
1
CM
F»
•
•
O
mm
•
♦
r>
1
o
o
o
O
0
•
0
CM
1
•
4
N
1
•
CM
to
«
♦
••
•a
N
*«
N
0
3
0
3
0
3
0
3
«
<
9
<
•
e
K
•
0
0
I
e
0
•
9
CM
1
CM
CM
<
o
•
mrn
CM
•
N
n
1
o
9
•
0
N
•
mrn
I
1
o
o
•
4
•
m
*0
•
—
a
.0
•
N
•
9
n
CM
•
t
o
O
9
o
o
•
o
3
♦
1
0
ri
•
N
I
CM
mm
N
mm
CM
M
0
0
0
0
0
3
<
0
3
<
N
•
•
•
CM
1
•
•
o
o
1
CM
1
o
0
•
0
CM
1
mm
CM
mrn
mrn
mrn
mm
0
3
<
<
0
3
0
3
<
f%
•
0
•
4
CM
mrn
0
3
<
•
m
o
H
•
9
CM
1
<0
O
0
•
O
a
a
ji
u
0
©
9
•
1.136
ro
mm
•
9
•
1 .757
H
M
•
O
<M
mm tn
a tn
UJ o
Q
o
0
z
z
< <
03 o
J J
II
o
o *
ft
tft
CM
O
mm
•
**
M
o
u
3
trt z tn UI 0
tu z X UJ
X UJ 3 3
* mm X
z*
a
tn M
0 u o
M
z
W oz z u
©
o
a m w
o
MX z
n * 9
in ♦
z
» UJ
< ►
UI U I UJ U I O fc* < u i <n
a
z j -J j -J z UJ
X u i
>» o a o o < o UJ 3 a
0 Oi UaI
3 a a a a.
•
O
> UJ
K
in in
> > X
0 in tn0
c o a o • < Q X tn
♦
UI
u o
c s
z z
J
o
UJ
o
CM
o
mm
•
M
0
•
0
.
o
w
«*
CM 0
1 tn
UJ
*
mj UI
• z -j
U I UJ UI UJ U I mm
in UJ a UJ a •
z o ♦
3
«
O
O
e
•
o
0
•
0
14
o
a
m
UI
UI
UJ
UJ
-J 3 J 3
J z
0 z
z < z <
<
<
if
X
* u
a
u u u o
0
J
©
o
o
•
o
#
0
O
•
mm
CO
«
AUG
0
v
0
A
A
s
ffi 0
CM
0
•
0
.
a
«a
•
0
.
0
•
0
.
in
>
CM
•
0
.
0
0
o
CM
•
to
9
•
0
3
<
0
3
<
0
3
<
•
O
4
—37.90
4
9
•
*
0
•
0
-2 6 .C
C
c
N
•
4
N
1
1
•
0
*
a
-j
u
i
u.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2
0
O
X
0
206
-«
o
m
•
tn
•
X
tn
©
•
C|
e
o
o
I
>
C
©l
N
C
O
Jl oe
N o
©
•
X
CJ
o
o
o
•
♦
Cl
0
N
o
•
X
Cl
o
© o
o
o
e
©
o
©
©
o
o
o
o
© o
o
e o
o
o
o
o
o
o
o
o
N
**
0
0
Oa
o a C
a w a
CJ tn
I tn tn
» g UJ
• z g
UI
UI I 3 z X
UI J UJ g U M aM
o
•3
tn
j
g
o z o z !/> z tn g m
g z z UJ
z X z x
<
x
z g 3 3
X
x X M * fV g
X a X o M o mm tn X
mm U M >
a o g a
a
0 j o g mm
X X
Xo
j
g
UJ
tu 1 UJ X
z
UI UI UI UI UJ a g g mm
tn UJ a UJ a • z o M tn
M
©
»
tn
o
z
o
a
a;
M tn
< © tu © U
JO
U
J o UJ o
z
X
mm a z g
x o
o
M o M H
©
>
Q
o
M
V) X tn X © w a z
g g
Z
<
< g tn
J UJ UJ UJ c
a U
z J g g g z U
J k- X g
> o o a o < o g 3 X
UI o g a
3 a a a a.
• > g o g
O
tn tn <n «/} N.
> > a
a
tn tn u> tn
0 o o o • x a X tn
a a a: e X g g g o
V g g u
3 z
o
tn g a Cl
o o Qa a M a 1 X
z z r z z o u. o g
X x x x < z tn 1) a
Jz
o © o o 0 U
a X o
-J j a 0. tn a a X
It H tt H at uj H tn
in O tn o X a n
o * © x i i H ■«
g g a a • ji ** >
mm UJ tn
CJ > >*>4 a
UI 0
z
N
©
O
X
©
in
09
0.
Q
C
J
UI
J
J
X
a
x
«•
©
•
o
X
K
0.
1
o
a
a
3
Z
©
a
o
tn
©
t
•
•
X
1
o
tn
©
X
g
©
•
N
n
9
C
l
mm
tn
CJ
■«
0
•
•
•
CJ
CJ
•
X
1
o
X
•
<0
m
O
mrn
•
X
0
l
9
o
o
O
X
o
•a
•
X
0
I
0
•
•
•
•
•
0
Cl
1
mrn
©
K
•
UJ
n
9
©
©
•
0
0
•
©
X
•
tn
0
9
•
Cl
Cl
N
•
*»
X
0
•
•
mm
©
O
O
•
0
•u
fm
•
o
0
©
N
•
O
o
s
•
©
•
o
o
0
1
1
m
•
0
C|
•
•
0
©
•
©
C|
mm
©
©
O
CJ
-
•
•
•
0
•
0
©
©
©
•
o
o
o
•
O
•
0
0
t
0
C|
mm
X
o
•
©
M
1
©
Q
•
•
t
•
•
«N
o
o
•
0
0
1
©
o
O
©
•
0
0
0
•
0
CJ
1
©
o
•
X
1
•
•
K
©
©
©
•
C|
•
•
0
V
N
•
X
•
0
o
X
0
©
C.
•
mm
•
N
Q
e
•
•
m
o
N
m
•
«•
O
N
««
•
©
1
N
•
©
•
©
tn
m
©
mm
•
0
•e
•
o
X
N
I
•
o
•
X
mm
0
tn
•4
•
•
g
ui
3
O
0
n
tn
mm
K
©
*
e
0
•
N
•
0
1
N
©
«
CJ
9
X
©
•
' X
1
©
©
•
o
0
•
©
©
•
o
X
N
0
•
f
0
9
o
©
•
o
0
•
©
O
0
•
©
9
1
tl
a
a
a
a
g
o
0
O
g
•
©
Cl
1
•#
•
tn
N
1
PJ
•
X
C|
I
•
©
O
X
•
©
o
0
©
0
9
1
1
t
C|
1
U
1
o
©
C
J
Cl
o
•
0
0•
©
©
0
•
•
0
©
•
•
•
K
N
CJ
1
1
-•
N
•*
©
3
X
©
3
X
©
3
X
N
1
-4
o
r»
•
©
<N
1
C|
1
Cl
©
3
X
©
9
X
©
3
X
©
C|
x
u
z
z
u
X
g
3
<
g
3
X
g
g
o
g
3
<
3
X
3
X
3
X
g
g
g
O
3
X
3
X
3
X
3
X
O
J
UI
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
tn
tn
10
0
e
•
H
N
o
•
K
M
O
•
4«
s
0
o
•
9*
0
O
•
o
0
o
•
mm
>
©
O'
©
•
o
O
X
■a
•
e
X
4*
•
«•
N
©
©
•
O
s
©
©
•
o
©
O'
•
o
O'
©
•
o
n
0
N
N
•
0
N
•
0
•
©
0
•
0
•
0
©.
•
X
•
mm
mm
mm
—
«
©
mm
©
0
©
•
*•
X
©
•
X
e
•
mm
©
o
•
■4
O
0
•
e
in
n
•
©
©
©
•
Q
D
©
©
•
O
X
©
•
©
•
>
•
X
•
rnm
mm
mm
•
©
0
■4
©
o
«
0
O
J
r*
•
•4
0
C
M
■4
•
v>
©
O'
•
o
©
0»
•
o
o
©
ft
•
0J
e
©
ft
•
nj
X
•
X
•
0
O'
•
H
•
4.96
mm
N
©
•
♦
—
mm
1.069
*1
0
o
,•
mm
©
-0
•
♦
9
•
♦
1.099
0
N
o
•
0*
•
O
X
X
UI 5
o j
.J
UJ
UJ UJ
UJ X
3 g
13 UJ
UJ Q
o
o
© X
*0
X
1
x
»
•
O
*4
X
UJ
0
z
M
z
mm o
UJ ©
a UJ
g D
x mi
n
M 9
o
X iS
X
z
UJ mi * *
0
1A
z a 0
X t/t X
X
z u
o X X
X a Z
►- H UJ
< UJ ©
b» X UJ
UJ 3 a
o
a
UJ d UJ
> > e?
a
z
x
0
in u
Oo
z
<
0
z
<
a
J
X
a
z
UJ
z
a X
m i a a in a
H H u *t X UJ
0 O © o X a
O ♦ 9 ♦ <
1 n
o u a a •n x
0J >
N a
©
N
i/i
a
a
X O
JX
0 1 3
u . o UJ
© 1
1&
a
x
9
x
>
X
o
X
V)
z
o
o
mi
h»
lb
0
»
O
UJ
I
J
X
c
<
a
x
g
z
o
e
o
©
01
0
♦
mm
N
N
r>
0
O 1
in
N
CV
O
♦
a
V)
N
•
©
N
I
M
I
o
©
•«
0
©
CL
6
r*
m
O
♦
•
e
in
V
1
N
•
mm
tn
•
r.£
rt
I
9
r)
n
•
e
ip
1
•«
o
O
•
•4
1
•
©
♦
1
0
O
O
o
n
m
O
♦
mm
•
o
♦
tn
9
o
«o
O
n
♦
1
©
•
o
♦
•
o
o
n
•
©
t
•
ru
¥1
0
©
mm
I
©
o*
0 in
n
*
N
1.08
X
**
1.16
z
<
z
a
x
(0
z
UJ
X
44
tt
w
1
0
o
O
.
o
z
<
o
</>
<n z
«■ UJ
X
< M
3
10
o
♦
•
1
I
o
•*
•
e
o
♦
1
CM
•*
•
a
1
o
V
o
<r
O
eg
<M
♦
1
0
.
U
o
(M
I
o
o
-36.28
UJ X
mi
O
©
tu
z
O'
•
•
O'
n
f
•
©
r>
1
o
o
N
•
o
•0
1
1
V)
£
o
a
u
o
f*
tn
•
O
mm
«
■
0
1
1
1
9
•
N
<M
e
Z
g
a
01
m*
<M
•4
N
>•
0
to
©
M
•4
13
O
o
(J)
•
•#
o
<
o
X
z
a
X
g
X
+
N
N
\9
o
V4
©
o
N
*U
1
0J
I
•
in
“i
1
•4
N
•0
X
X
13
13
•
tm
g
g
.X
0
0
0
X
X
g
X
<
0
■4
—
0
0
-4
*•
•
n
-25.73
a
z
x
o
*•
X
mm
•
o
N
1
•
©
fU
1
N
C\J
1
ftl
mm
N
•m
s
rj
CM
0
0
©
13
o
13
C
13
©
o
J
X
X
X
•
mm
g
X
0
g
X
g
X
0
0
-4
0
•
g
X
g
X
•
g
X
M
•
N
1
0
0
x
•
h»
mm
g
X
g
X
II
UJ
M
O' K
X
u ; UJ Q K
j J Z UJ
X
a o
©
a a
• 8!
(0 (0 </> ©
tn © © (0
a
a o a o • X o X
♦ j Ui u
a X a
u u u 3
DX
X
IP
•
©
0J
•
0J <0 <4>
4*
mm o 4* 0 1 0 <0
a a o a mm UJ 0
• mi UJ
a mm o
X
X
# z mi X
o
z
N
<M
a
13
0
mm
M
>
K
UJ
UJ
j d
a Z
z <
x
X
* o
o u
o -i
.J
N
UJ
m UJ UJ
in u i s
a a
< v UJ
UJ a
< a
o
©
0 ♦
z
a uj UJ
X mi •J
> a o
3 a a
0
X
•
X
in
mm * *
x
•
X
X
AUG
m
rn
O
X
1 4 .S
207
0
j
UJ
M
0
mm
rnm
mm
<
ej
X
X
N
N
ej
OJ
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
<
w
X
0J
208
O
0
0
N
in
tn
in
(M
44
>
a
(N
*•)
•
M
■4
OJ
n
•
44
u
in
•
«*
in
AJ
N
IP
X
9
o
1
«4
9
«
a
4i
1
1
1
f
in
mm
«4 9
0 a o
o
o
V
UJ
UJ
UJ - j
J 3 -J
vJ z 3
2 < z
<
<
X
a z
O a w
o j a
•J
J
»•
UJ
UJ UJ UJ UJ
t n uj s UJ
s •3 2
< o ui o
UJ o UJ
< a
a
o
o
ip * 0
2
O UJ UJ
3 J .J
> O o
3 a a
o
V) 0
in
a o
a 2
u VJ
a
z
<
a
UJ
-J
a
2
in
in
o
2
U
o
z
<
o
z
0
o
<
- j -1 a .
n B if
m o uS
o * o
u u X
•
IP
-4
N
44
in
in
UJ
. j 44
z M
o 44
o
o
o
o
o
O
0
•
mm
•
•
•
•
•
•
•
•
•
•
O
n
•
•
•
♦
•
o
♦
in
z
UJ
X
**
o
u j in
a uj
9 3
»- J
m <
»
Q
x in
X
z
UJ «J
o * - in
z o m
44 i n *•
X
z u
o
M 2 Z
*m W UJ
< uj m
w X UJ
UJ s a
o ^ &
UJ a uj
> > 2
o
UJ
X
a
o
u.
in
z
<
2
H
H
m
>
x
u
in
o
a
(V «■
1
tn
1
2
O UJ
it a
o
X
in
44
UJ
z
mm
mi
IP
mm
N
z
o
mm
J
Ik
O
mm
M
o
N
m
X
o
e
9
pi
o
9
•
O
■ri
o
14
CJ
mt
mm
•
«
a
*
e
. ♦
a
•
K
«4
•
9
9
•
©
mm
o
0
1
1
2
-J
UJ
J
J
<
*
44
•
#
•
©
O
•
O
9
•
O
M
•
n
•
•
in
tn
«
tn
in
o
o
•
O
©
•
•0
0
a
u
*
N
9
AJ
N
N
z
3
«*
<M
m4
N
am
N
am
01
♦
♦
N
mm
N
0
0
0
0
15
O
3
15
•
o
t
I
i
0
0
•
0
•
N
•
1
mrn
CVJ
•
O
♦
2
a
in
u
&
o
a
u
0
*
2
<
a
4*
X
3
Z
o
•
o
o
(M
•
N
ip
2
•
IP
0
•
0
o
to
•
•
0
0
n
t
o
•
•
♦
1
9
1
o
N
•
OJ
mt
o
•
1
01
•
o
OJ
•
1
■
1
i
i
1
I
O
mm
M
O
OJ
9
O O O
I
f
44
♦
♦
44
44 N in
1 m
0
UJ
0
•
« 2
UJ X a
u ■■
in
3
z m z
« *4 UJ
X
< M
a * o
w
o
j
< X
UJ • UJ
Q
UJ
2 » Z
19 « M
UJ o
Q ■# z
44 a
**
O
♦ 9 ►»
<
UJ 0
J Z UJ
o
<
a
UJ
&
• >
in
2
in
• <
o
a * J
3
u
V) U
a a 44
z z a
x < z
o o UJ
a
a m a
UJ
ii
o X a
* tl n
a m a*
oj >
M Cl
♦*
1
i
i
9
2
>.
<
a
"
z
I
1
I
«
•
z
a
X
a
j
UJ
3
<
<
(5
3
•<
O
3
<
O
3
<
<
OJ
O
3
<
J
<
-
-•
X
X
<
3
<
M
X
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CRUP«=SORGHUM (PERPENOICULAR ROMS TO F LIG H T L IN E )
COS
C40
POS
Z2S
PVI
TVI
SH01
3.78
0 .4 6
7.31
2.647
1 .175
25 .0
•
2.647
1 .175
25 .0
7 .1 4
2.S40
1.170
2 6 .9
.
2 . 840
1 .170
26 .9
6.ao
2.488
1 .179
•
.
2.488
1.179
•
2.687
1.173
11.9
2.607
1 .173
11.9
2.738
1.175
6 .9
2.738
1 .175
6 .9
7 .1 0
2.568
1 .175
5 .6
•
2.568
1 .175
5 .6
9 .0 4
3.309
1 .179
24 .7
.
•
3.309
1 .179
24.7
0 .5 4
0.9 0
3.241
1 .177
21.1
.
3.241
1 .177
21 .1
12.05
4.653
1.189
AUG
2
1
-2 6 .6 0
IX
AUG.
2
2
- 2 6 .6 0
- 3 9 .1 0
- S . IB
-1 7 .0 4
IX
AUG
5
1
- 3 1 . SO
- 4 0 .8 0
- 4 .3 a
- 1 4 .8 3
1X
AUG
S
2
- 2 9 .4 0
-4 0 .4 0
.
IX
AUG
a
1
- 2 3 .1 0
- 3 6 .4 6
16.80
IX
AUG
a
2
-2 1 .60
- 3 7 .0 0
.
•
.
IX
AUG
It
1
- 2 7 .1 0
-3 1 .5 0
11.90
- 1 .0 0
o.ai
IX
AUG
11
2
- 2 3 . eo
- 3 4 . SO
.
•
.
IX
AUG
14
I
- 2 8 .5 4
- 4 .4 2
•
IX
AUG
14
2
- 2 6 .4 0
•
•
IX
AUG
17
1
- 3 4 .0 0
- 5 .0 0
1X
AUG
17
2
- 2 6 .6 0
-3 4 .7 0
•
•
IX
AUG
2
1
-2 7 .6 0
- 3 6 .3 0
•
•
IX
AUG
2
2
-2 7 .9 0
-3 0 .2 0
- 3 .6 3
-1 7 .9 1
1X
AUG
S
1
-3 1 .2 0
-4 2 .4 0
-5 .0 8
-1 8 .9 4
IX
AUG
s
2
- 2 9 .3 0
- 3 8 .5 3
•
.
»
IX
AUG
a
1
•
.
0.9 0
.
.
•
o
prohibited without perm ission.
Z1S
1X
•
.
P40
210
a
RUN
MCKTH
u
n
•
6
DAY
FIELD
0
«
<
N
1
Reproduced with permission of the copyright owner. Further reproduction
GUYMON DATA SET
C05>:L BANG CROSS POLE 5 DEGREE LOCK ANGLE (OBI
C40«=L BAND CROSS POLE 40 DEGREE LOCK ANGLE (0 8 )
P05*P BAND CROSS POLE S DEGREE LOOK ANGLE (OB)
PAO^P BAND CROSS POLE 40 DEGREE LOOK ANGLE (OB)
Z10. Z1S. Z25=MMS BANDS 4 . 7 . AND 0 (1 0 » 4 (-4 ) 4ATTS CM44I-2) S T * * ( - D )
PVI=PEGPENOICULAR VEGETATION INDEX OIHENSIONLESS)
IVS=TRANSFORMGO VEGETATION INDEX (OIHENSIONLESS)
SHO1=0-2 CM VOLUMETRIC SOIL MOISTURE (X )
PERIODS REPRESENT MISSING y4LUES
•
•
3 .9 0
—18.40
•
0 .8 3
.
0.70
0 .7 5
.
0.7 2
.
0.9 2
•
0.9 0
.
o.so
.
0 .4 0
•
0 .4 9
.
0 .4 8
.
0 .4 5
.
0.53
.
0 .5 5
7.4 7
.
7.57
.
.
210
9*
*
o
X
in
■»
>
>
■
1
1
•
i
i
99
99
«•
M
>
a
O
O
N
N
e
9
•
s.
K
■
99
•
9
ft
©
•rt
P*
M
«•
•
*
»
m
©
•
K
r-
*•
9
9
S
•
ft
•
•
9
9
•
•«
ft
9
9
K
•
ft
in
tn
•
•c
«
•
«
©
•
9
m
tn
N
99
•
99
O
9
99
•
©
in
•N
99
•
o
9
*
©
99
99
99
©
N
h»
M
•
99
99
9
99
99
•
99
•
ft
n<*
m
©
N
-O
V)
w
CM
O
V
N
•
ft
o
9
P»
K
ft
*
•
M
©
99
m
t
•
99
N
ft
X
•
■*
99
ft
ft
t
••
•
9t
9*
in
9
in
tn
9t
e
9*
9
in
•
a«
N
♦
•
N.
in
x
•
••
9
•
N
•4
N
9
99
9
99
•«
pw
in
9
•
t
o
9
9
9
•
ft
9
«•
••
**
9
•
cm
ft
in
in
ft
•
•
•
0
9
9
■«
* •
9
99
9
©
9
9
©
9
9
9
9
99
•
99
in
in
«•
©
9
ft
9
99
CM
o
9
M
•
•rt
©
*
e
m
t
0
o
99
*
99 a
•« f t m 09
s • j ) in
a
UJ <n
UJ
o 90 9 90 *
• 2 .j
90
90
UJ X 3 z
UJ
u 99 3
UJ J (JJ
in M
m
l o J •3
o z o z in z in
<
<
u z
z
z
<
X u
<
X
X < 99 3L
X a X 3 s 3 90
□ 9 a a
a
90
e J c j
# X
-j
UJ
UJ 1 UJ X
o UJ
UJ UJ u UJ UJ
in UJ s UJ s * z o
z
e o a c ♦
■a
< 3 Uj o UJ o
t z
UJ O u a m
9
0
< o
o z
o
99 o
o
e
o
in ♦ M •» 9 tm M
«c w
2
<
a UJ UJ UJ UJ O
z j j -J J z UJ 09
3 0 a 3 < vj
3
a a. a
UJ o
• 9 UJ
9
in in in in N
>
<n in in m
ce
• < o
a o o o
a
<f
s a X
. j UJ
w u u V
3 X
in O z
a o Q a Q
9
z z Z z z a u.
< < < < < z in
© o 9 o © UJ z
a <
j
J a a in a X
ii H ii
H a. UJ m
in o in o X a It
ii
ii 90
9 « o
•J u a X n 99 >
ft 9
N a
*9
a
.
A
ffl Q 9
UJ
z
in
«*
in
09
M
N
o
UJ
3
3
m
in
««
o
X
ft
UJ
3
3
«
>
Z
J «•
M
a w>
in ««
X
u
99
a
z
P
9 UJ
UJ -n
X UJ
3 a:
J fi.
o Ui
> a
X in
u o
o
f t 90
1 s
o UJ
II a
99
e
X
in
Z
O
u
N
o
*
M
>o
<0
©
o
9
in
(0
CD
o
o
O
*
0•
a
Q
C
-J
3
U.
o
z
ft
I
ft
o
*
I
I
a
a
©
I
©
a
UJ
9
m
ft
•«
1
©
e
o
in
9
99
as
in
m
«
o
ft
60
©
I
UJ
o.
G
9
9
in
•
•
3
Z
1
9
a
©
o
<*
V
©
1
n
o
«•
9
I
o
o
0
K
m
1
o
9
•
©
99
tn
©
o
in
o
•
o
in
n
n
•
©
n
tn
<
1
1
t
t
•
ft
v
I
•
o
*
1
VI
O
9
©
ft
9
K
•
9
o
m
u
•
a
o
<n
n
z
u
©
ft
u
1
o
©
(J
•
«
•
ft
•
©
9
ft
9
ft
I
ft
1
4)
M
>
v>
*
•
•
•
9
©
©
1
ft
■*
9
•
r>
ft
9
•
ft
♦
1
O
N
•
ft
•
•
in
CM
O
N
*
9
«
*
i
9
1
0
{
e
9
9
n
9
o
©
©
t
©
ft
ft
•
o
99
o
O
art
♦
•
•
ft
9
©
ft
in
CM
1
1
1
•
•
o
t
9
•
©
©
©
n
1
in
tn
•
ft
1
♦
ft
t
1
1
©
«
-
ft
•*
ft
^
9
3
<
9
3
<
9
3
X
9
3
<
9
3
9
3
<
9
3
<
9
3
<
9
3
<
<
9
3
<
9
3
<
X
•rt
X
—
9
•rt
9
•rt
9
9
v4
9
•rt
9
9
9
9
art
a)
•*
•«
o
3
<
o
3
<
9
3
<
9
2
<
9
3
X
X
X
■#
X
X
I
1
1
o
9
ft
ft
V
X
Q
z
z
u
<
X
Q
ml
UJ
9t
9*
rtrt
9*
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
9
3
<
211
m4
rt
•
9
o
X
(rt
O
rt
mm
0
■0
M
>
rt
4
M
>
a
B
o
~
UJ
-I
J
z
4
u
-J
©
z o z I/I
< z 4 ►<
►
x
* <
z o *
o ©
-J Q
ui
*
*
* * UJ *
30
0 U*
UJ Q
O
O
lf l 4
rt •*
Z
V) UJ A
z a uj
UJ
Z rt J
•* (rt <
Q w >
O
X ©
UJ X
z
-I rt
Q U
Z o m ir t
M Z O (rt
«0 in «
Z
o z
u
X
M
o
M
•J
* *
♦
©
-
a a a a
z
uj
UJ (rt
X UJ
uj 3 X
UJ N
©
UJ 0
• > UJ
dS
N 0! > > X
< o X rt
J Uj u o
o
V) (A © v*
</) </t
o a q o •
a a z x <
v u u w
07
a a a o o
z z z z Z
< < 4 4 4
O ffi o a o
3
u
M
Q
z
mt
z
9
s!
•
mm
4
•
eg
rt
9
•
rt
•
OJ
9
•
©
rt
rt
•
O
•
4
rt
•
©
mrn
9
•
©
•
rt
(0
•
o
©
•
•
o
a a
ii n II H
m
© ui o
o ♦ o 4
- >
u u o. a •rt
oj » rt
O
m»
2
••
mt
•
01
wt
••
•
01
**
4
•
©
0J
9
•
K
01
©
N
mm
•
mm
4
©
mm
•
mrn
4
©
mt
•
mm
©
©
M
•
mt
©
©
mt
•
©
©
mt
•
©
©
•0
•
mt
mt
•
mm
©
O
o
•
01
©
O
o
•
eg
rt
o
4
•
rt
rt
01
«•
•
01
rt
O
in
4
•
0J
O
rt
4
•
0J
f*
rt
rt
t
mt
rt
©
•
in
•
9
9
•
©
0
rt
9
0
4
mm
•
rt
n
mm
©
mm
•
mt
o
N
mm
•
mt
mm
©
mt
•
mm
in
wt
mm
4
•
OJ
eg
4
•
0J
rt
eg
4
•
OJ
•
4
O
•
•
0
4
•
01
O
9
•
©
•
OJ
mm
•
OJ
©
o
•
©
0
0
e
0
o
mt
mt
w
rt
0
o
0J
rt
mm
mm
0
mt
*
N
rt
rt
•
mt
©
mm
rt
0
mt
mm
rt
0
0
rt
N
rt
•0
*
m
rn
•
©
rt
«
o
•
o
4
•
o
•
4
rt
•
o
•
©
4
•
o
0
rt
rt
•
o
0
©
©
•
o
•
rt
©
•
o
•
rt
©
•
o
•
N
©
•
O
•
o
©
0
o
0
o
N
•
o
•
o
©
•
©
1
0
mt
OJ
•
K
0
©
0
rt
0
©
fa
0
O
©
9
0
0
0
N
0
O
•
4
9
0
O
0
mm
0
9
O
rt
O
4
rt
K
a
•
rt
a
•
1
01
9
©
.
o
a
a
<
0
m»
N
•
•
©
mm
u
M
Q
Z
UJ
©
o
1
1
a
rt
o
eg
1
7
a
«
UJ
o
3
r
4
®
rt
rt
l
o
O
©
0
mm
•«
8
1
4
«
©
©
eg
9
F>
I
*
PI
I
0
©
in
9
®
p>
1
4
rt
i
9
a
o
H
a
o
a
j
©
CM
Z
3
X
S
H
V
-
0|
rt
N
I
rt
rt
t
O
©
O
(VI
r>
I
rt
4
*1
I
4
-
OJ
0
0
4
4
3
0
3
4
o
©
•
O
©
n
4
•*
4
OJ
4
I
I
9
pj
n
•
n
I
0
n
UJ
o
j
©
"I
I
©
oj
N
N
-
N -
0
3
4
0
0
0
4
4
4
OJ
0J
0J
o
V)
9
eg
•
(rt
n
I
V
o
4
o
rt
a
M &
©
4
©
mm
•
0J
©
•
e
OJ
mm
•
K
m
9
OJ
O
I
a
«M m
I X
U. O UJ
rt n x
uj z
a. <
(rt a a
x u M
z a it
ii ii m
N
4
•*
4
«
mm
m
m
rt
o
•
rt
•
9
0J
o
rt rt ar t
rt 4 9 <
rt 4
UJ UJ UJ UJ o
- J - J .J J Z
a a a a 4
in
H
- o
u; a
<r o «
© uj ©
M
uj o
w
o
I
-J UJ
Z -I
o z
2d
UJ
UJ 1
UJ UJ UJ UJ «
uj or
i/s </i
uj
3
X0
n
o
j
•
•
(/> ^
n
a 55 ;
OQ
a
a
©
n
•
a
9
•
*•
rt
0J
N
rt
•
9
rt
0J
I
«•
©
eg
I
?g
eg
—
OJ
<M
<u
t
9
m
eg
I
0|
•
t
tn
vN
I
1
«
I
>
4
O
X
rt
Z
u
o
3
9
3
4
4.
0
3
<
Z
O
J
UJ
9
mm
O
OJ
0
3
0
3
4
0
3
4
0
3
4
3
O
O
©
O
O
O
O
OJ
OJ
OJ
OJ
OJ
OJ
OJ
01
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3
3
0
3
4
212
>
f-
•
O
(9
(9
trt
*M
•4
4
•
•
mt
CO
i
t
>
a
i
a
1
i
9
9
9
|
9
0
0
h*
tn
m
(Vi
N
•
rt
rt
«
M
9
M*
mm
0
♦
tU
0
91
mt
«*
.*
•
*•
•
o
o
•
e
m
•
—
mrn
(VI
0
(9
in
(V
9
<9
.*
s.
•
•
•*
mm
M
mm
<—
tn
tn
in
tn
rt
rt
9
4
0
•
•*
•
•
•
—
0
in
mm
4
•
a*
tn
•
•
©
•
♦
mm
rt
mt
mt
m
rt
•
«
•
«—
9
♦
•
**
—
•
9
9
•
♦
rt•
9
tf)
«
mt
i
mm
mm
•
•
•
4
•
-
in
tn
•
•
•
o
<9
a
<9
n
rt
in
«
in
rt
rt
«•
—
♦
•*
4
«*
<9
*
*
•
tn
a
u>
—
4
«*
•
♦
o
**
•
0
o
m
•
M
m
«•
•
««
mm
**
mm
—
a*
9
9
(V
O
0*
0
•
(VI
o
(9
4
V*
0
•
(VI
4
tn
0
•
0
(V
rt
*
(V
(V
in
«*
mt
•
*•
0
(VI
N
in
9
•*
o
•
a
•
•
•
<9
0
o
•
*
<9
o
rt
•
•
4
•
(VI
N
V)
a
o
0
0
9
9
l. I S l
M
0
O
<9
•
2.4,99
m
rt
O
(9
•
rt
•
rm
o
X
2 1 *1
i
i
a
o
z
4
ai
J
II
tn
o
o
•
J1
N
O
mt
M
N
rt
X
I*
M
3
«|
-J
4
;»
IL
X tn
U Q
O
M ■*
1 X
UJ
o
ii 0 .
o
x
w
o
o
mm
IS
•
•
o
•
rt
O
•
o
•
mm
O
•
—
ft
rt
•
O
O
0
a
O
a
a
rt
O
a
mrn
a
o
a
«•
•
9
(9
a
0
a
O
0!
a
•
o
•
•
0
rt
a
O
a
0
a
O
a
tn
«
a
O
a
a
9
O
«
«
e
O
o
s
mm
a
o
o
a
•a
a
0
0
a
4
a
a
a
a
n
a
rt
(9
rt
a
O
•
•
o
0
a
•
X
<
•j
3
U
o
z
5P
a
X
UJ
a
w
X
3
X
O
a
a
tn
H
£
3
X
O
1
f
9
1
9
!
1
9
1
1
9
I
0
•
•
1
1
9
!1
1
9
1
9
1
o
rt
•
o
V
•
•
o
0
•
0
•
e
0
u
o
rt
•
(VI
4
I
m
o
•
rt
4
1
O
rt
•
(9
4
9
m
o
U
O
tl)
•
rt
N
«
rt
0
•
<0
<9
1
O
Hi
a
rt
<9
•
N
a*
in
IS
D
4
tn
e
• a
z
3
X
>
4
O
X
rt
z
GJ
X
•
4
t»
a
0
—
1
O
■ 0
a
-1 *20
O
rt
tn
M
O
a
t
s
o
0
a
0
1
a
a
9
—
*
O
a
O
tn
a
tn
•w
a
O
*
a
®
rt
9
O
in
a
mt
4
1
O
9
a
9
rt
1
o
0
a
M
4
1
O
rt
a
O
4
«
O
CD
a
mt
4
9
O
9
a
O
0
9
o
rt
a
4
(9
•
O
4
a
rt
ni
«
o
®
a
0
<9
1
0
9
•
(9
N
1
Q
O
a
rt
(9
9
o
4
a
4
(9
1
O
©
a
in
<9
I
N
•w
<9
•«
<9
•w
(9
•4
rt
mrn
rt
(9
<9
0
in
©
©
mrn
mm
0
3
4
0
3
4
0
3
4
0
3
4
0
3
4
0
3
4
0
3
4
0
3
4
0
3
4
0
3
4
4
(9
0
(9
4
(9
m
(9
in
(9
0
N
0
(9
m
(9
tn
(9
0
(9
c
•
•
*
0
a
rt
I
•
•
0
0
a
rt
rt
1
o
©
a
9
rt
9
•
•
m
rt
a
rt
(9
I
O
rt
a
<0
*
1
N
m
<9
0
0
mm
m
mrn
mt
0
M
4
aw
9
■3
4
0
3
4
0
3
4
0
3
4
0
3
4
4
(9
0
(9
■
a
I*
O
a
rt
N
•
(9
a
a
12*00
O
z
J ft*
*m tn
0 <n
w **
X
u
N rt
ff Z
N UJ
U tn
Z UJ
3 X
J a
0 Ui
> Z
m
o
•
o
-3 7 .3 0
tn
tn
o
z
u
<n
UI
•
-2 * .3 0
0
aj
cc
0
N
VI
0
z
•
tn
J
2
to*
UJ UI
tn u i
a
< 0
^ u
4 Q
o
0
Z
O Ui
X j
> c
3 a
0
K
v
II
*
u
s
-J
mm
UJ
z
AUG
UI
-J
0
z
4
mm
CS
o
mm
25
?9 0 mm
9 tn tn
mm Ui tn
0 j Ui
• z J
UJ
UJ X o Z
O
J UJ .J u
**
3 J 3
z 0 z tn z V )
< z 4 * u z
<
z UJ
4
X
u * a * o —
u
u u 9
w
J
.J
0 4
UI
UJ 1 UI X
%r
u u i UJ
Q tu
oc UJ s 0 z Q
u X 0 ♦
Z
u u tu O
-•
o u i c a* Z
W a z
o
M O
o
o
4 0 4 0* r t —
4 rt
u u i UJ Q r t <
-J j
Z UI m
j
a o o 4 O UI
a a. a
UJ O
• > U
tn tn tn
>
m tn tn
tt
• 4 Q
s a o
X « a 0 J UJ
u u u
3 X
to u X
Q 3 a a
w
a
Z Z z z
u.
4 < 4 4 z tn
0 0 o a ui z
a <
J a a VO a a
II 1! H 0 UI
Q in O X a It
4 o 4 n ii
u a a n •o >
N > rt
■4 a
A
mm »
m o a
o
o
.
o
J
UJ
mm
u.
4
CJ
4
(9
4
(9
0
(9
9
9
9
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
213
«4
t
o
o
x
to•
I
«
OJ
(0
rt
♦•
to
N
CM
<0
(M
>
a
9
w
*
♦
to
tn
t
I
1
I
I
1
l
0m
O
J tn 0^
C
O 1 tn tn
UJ
a
4 UJ tn
. z
o Q cd a 4
A
n
m
i
U
i
•4
a
» 2 J
«# ^0
•I
UJ Mi z a z M
to
u M a 44
UJ
M
tn
z
J 3 3d
o z o z tn T? UJ v> o
z < z < *» UJ z 3 UJ •4
to z UJ 3 3
<
<
J
X
u.
X < M X to mi
u -J w
r t M «)
Wa
>
o
c Qa o
wa
to
Q J OJ
♦ X
z o
-J
tn
to
X
z
UJ UJ 9 UJ
44
j i
J UJ UJ 44 a UJ j
UJ UJ U
in UJ a uj a • z o •4 tn a
* M z a <n CL
oc rt a
M
*4
o
tn
u
< rt
rt UJ
X
to \M Q uJ O •4 z
<
<5
< o
o
o z u
M o M to
o
o
J
a
5
tn ♦ tn * * to t4 a z
Z m
< to UJ u
•4
U
J UJ UJ a to < UJ in
a
J to X UJ 5
X J J . j -J 2 U
> c Q a o < rt U
Z
J3 a
o
a a a a • UJ r t -J a
a
> UJ 3 UJ W
o
m in tn m N
> > X
z
UJ
in Cl wi in
3
o a o o • < a z tn
a
44
a
J UJ u o
a 3 3 X
u u U UJ
3 z
tn v a O
J 44 X
o Q a o a 14 a
3
s
Z
z Z z z z a u . o UJ
< < «< < < z to 1
rt
1
a
O
Ja G
J z ■4
a
Do <0 U
o
a < o
tn
J j a a V) 3 a X
J to tn
M ii H M z U
If
tn o m o s a tl
a
o * o ♦ If 'i ■4
a
w u a a «n ■« >
z
tj
N > to
Nl a
I
•
in
i
1
N
1
i
•
i
O
]
i
N
1
s
1
I
I
1
1
9
1
I
|
1
in
eg
N
M
«4
N
O
mm
N
O
<»
a
♦ S
tn
*
•
N
0
1
tn
•
•
tn
a?
•
o
•
rt
•
♦
•4
o
3
5
N
rt
•
a
•
to
•
•
to
O
a
*4
a
•
tn
rt
•
tn
*•»
I
*%
1
«4
©
•S'
•
«4
•
m
•
m
•
tn
a
•
o
*
o
m
1
ri
tn
©
u
»•
•
o
oi
!
o
•
o
N
z
3
X
©
10
to
•
o
•
rt
O
1
m
4
n
V
4
rt
l
rt
•
a
•
rt
•
M
•
tn
tg
•
*
N
«4
•4
N
>
<
Q
♦
to
•4
to
•4
Z
to
Z
a
o
3
o
D
<
o
3
<
tn
tn
<u
1
O
3
<
O
J tn m
Ui
CJ
N
1
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
214
VITA
Name:
Wesley Dean Rosenthal
B irth Date:
August 17, 1952
Permanent Address:
Education:
3020 Strauss C t., Lincoln, Nebraska
68507
High School Diploma, 1970, Lincoln Northeast High School
B .S ., Geography (sp e c ia lizin g in c lim ato lo g y), 1974,
U niversity of Nebraska
M .S ., A gronom y, 1976, Kansas State U n iversity
F ie ld of S p e c ia liza tio n :
Societies:
Applications o f remote sensing to p la n t/
water relatio n s h ip s , and so il and water con­
servation
Alpha Epsilon
American Society of Agronomy
American Society of A g ricu ltu ral Engineers
American Society of Photogrammetry
Sigma Xi
Soil Science Society of America
Experience:
1969-1973 (Summer), Student Technician, Department of Aqronomy,
U n iversity of Nebraska, Lincoln, Nebraska
1974 Student Technician, Department of A g ricu ltu ral
Engineering, University of Nebraska, Lincoln, Nebraska
1975-1976 Graduate Research A ssistant, Evapotranspiration
Laboratory, Kansas State U n iv e rs ity , Manhattan, Kansas
1976-Present Research Associate, Remote Sensing Center, Texas A&M
U n iv e rs ity , College S tatio n , Texas
This d issertatio n was typed by Ms. Linda Kocman.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Документ
Категория
Без категории
Просмотров
0
Размер файла
6 398 Кб
Теги
sdewsdweddes
1/--страниц
Пожаловаться на содержимое документа