close

Вход

Забыли?

вход по аккаунту

?

Sierra Nevada winter storms: A study using microwave radiometry, ice crystal and isotopic analysis technique

код для вставкиСкачать
IN F O R M A T IO N T O U S E R S
This manuscript has been reproduced from the microfilm master. UMI
films the text directly from the original or copy submitted. Thus, some
thesis and dissertation copies are in typewriter face, while others may
be from any type of computer printer.
The quality of this reproduction is dependent upon the quality of the
copy submitted. Broken or indistinct print, colored or poor quality
illustrations and photographs, print bleedthrough, substandard margins,
and improper alignment can adversely affect reproduction.
In the unlikely event that the author did not send UMI a complete
manuscript and there are missing pages, these will be noted. Also, if
unauthorized copyright material had to be removed, a note will indicate
the deletion.
Oversize materials (e.g., maps, drawings, charts) are reproduced by
sectioning the original, beginning at the upper left-hand corner and
continuing from left to right in equal sections with small overlaps. Each
original is also photographed in one exposure and is included in
reduced form at the back of the book.
Photographs included in the original manuscript have been reproduced
xerographically in this copy. Higher quality 6" x 9" black and white
photographic prints are available for any photographs or illustrations
appearing in this copy for an additional charge. Contact UMI directly
to order.
U niversity M icrofilm s International
A Bell & Howell Information C o m p a n y
3 0 0 North Z e e b R oad . A nn Arbor. Ml 4 8 1 0 6 -1 3 4 6 USA
3 1 3 /7 6 1 -4 7 0 0
8 0 0 /5 2 1 -0 6 0 0
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Order Number 92S0122
Sierra Nevada w inter storms: A study using microwave
radiom etry, ice crystal and isotopic analysis technique
Demoz, Belay Berhane, Ph.D.
University of Nevada, Reno, 1992
UMI
300 N. Zeeb Rd.
Ann Arbor, MI 48106
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
University of Nevada
Reno
SIERRA NEVADA WINTER STORMS:
A STUDY USING MICROWAVE RADIOMETERY, ICE CRYSTAL
AND ISOTOPIC ANALYSIS TECHNIQUE
A dissertation submitted in partial fulfillment of the
requirements for the degree of Doctor of Philosophy,
in Physics
by
Belay Berhane Demoz
Joseph A. Warburton, Dissertation Advisor
May 1992
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
The dissertation of Belay Berhane Demoz is approved:
Department Chair
Dean, Graduate School
University of Nevada
Reno
May 1992
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
Acknowledgement
I would like to express my gratitude to Dr. Joseph Warburton, my research advisor and
committee chairman, for his support, encouragement and guidance throughout the course of this
work. H e conceived the idea of using stable water isotopes in studying ice-phase winter storms.
The advice and numerous discussions as well as the assistance given by Arlen W. Huggins
and Richard Smith in locating some of the data I needed is much appreciated. They taught me
much o f what I know of the D RI radiometer construction and operating principles.
I am grateful for the advice, suggestions and discussion I received from Dr. Ed.
Westwater.
I was fortunate to spend some time with him at the National Oceanic and
Atmospheric Administration office at Boulder. The knowledge I gained in ground microwave
radiometry, theory and instrumentation, from him and his group was very helpful.
Dr. Paul Neil and Dr. Kleppe, members of my committee are also to be thanked for
reviewing the manuscript and their encouragement. I would also like to thank Dr. Randolph
Borys, Dr. Richard Pitter, Dave Mitchel and Renyi Zhang of the Desert Research Institute.
Microphysics discussions with them were very valuable.
At last but not least, I would like to thank my family for the support and encouragement
they gave me to pursue my studies. I owe much to them.
This research was supported by the National Oceanic and Atmospheric Administration
under the NOAA/Nevada cooperative agreement.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
ABSTRACT
An observational study has been made of ice-phase winter storm clouds over the Sierra
Nevada mountains.
In Part I, two microwave radiometers, one designed with a spinning reflector to shed
precipitation particles while the other radiometer’s reflector was fixed, are compared. The
absence/presence of contaminated periods in the data was attributed to difference in design.
These apparent contaminated periods lead to lower correlation coefficients between the
radiometers. Comparison of radiometer and rawinsonde resulted in a correlation coefficient of
0.97 for the spinning reflector as opposed to 0.8 for the fixed reflector radiometer.
In part II, stable water isotopes were used to study mesoscale and microscale storm
modifications by the Sierra Nevada. Initially, a low level warm front lay across the region and its
elevation lowered with time from 2.5 km to 1.7 km. This decrease o f frontal surface height was
accompanied by a steady increase in the SlsO values. In the pre-cold frontal period, the 5lsO
values at the upwind site signified warmer origin ice crystals than the downwind site. This is
explained by orographic effects and the production of supercooled liquid water at low elevations
on the upslope side. The S180 value peaked around -13 %o which translates to an "equivalent
temperature" o f -10.7°C for ice phase water capture at the upwind site. At the downwind site,
this was some 5 to 6 centigrade degrees colder.
During surface cold front passage, the
differences in 5lsO at the two sites are small probably because, during frontal passage, the
orography plays a less significant role in the precipitation production process.
In Part III, observations of precipitation rates, ice crystals, wind and supercooled liquid
water (SLW) upwind and downwind of the Sierra Nevada are presented. Observations show that
the stage of development of the storms was important in the liquid and vapor development. High
SLW, and increased riming were located before the frontal passage. Duration of SLW as
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
observed by the radiometers, was always shorter over the downwind station. Heavy riming was
associated with precipitation decrease while high precipitation rates were correlated with high
number fraction of aggregate crystals. Aggregation was found to be an important process for
precipitation development over the downwind station.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
vi
TABLE OF CONTENTS
Abstract
List of figures
List of tables
PAGE
I N T R O D U C T I O N ..............................................................................................1
PART I.
1.1
1.2
PART n .
Objectives
.
.
.
.
Data sources and instruments .
.
1.2.1
Rawinsonde
.
.
.
1.2.2
Precipitation gauges
.
.
1.2.3
Ground ice crystal sampling
.
1.2.4
D R IK ,- band radar
.
.
Introduction .
.
.
.
Theoretical background
.
.
Instrument description .
.
.
2.3.1 USBR radiometer
.
.
2.3.2 D RI radiometer
.
.
2.4
Radiometer comparison study .
.
2.5
Discussion of results and conclusions .
2.6 Summary .
.
.
.
.
3.1
3.2
3.3
3.4
3.5
PART IV.
4.1
4.2
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
2
5
5
6
8
8
.
.
10
12
17
19
21
23
. 3 3
35
RADIOMETRY.
2.1
2.2
2.3
PART III.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
STABLE WATER ISOTOPES.
Introduction .
.
.
.
.
Experimental design and Instrumentation
.
The storm o f 26/27 March 1985. - A case study
3.3.1 Results from Kingvale .
.
.
3.3.2 Results from Hobart Mills
.
.
Discussion
.
.
.
.
.
Summary and conclusion
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
37
40
. 4 3
43
47
50
57
STUDY OF WINTER STORMS: UPWIND AND DOWNWIND OF THE CREST
OF THE SIERRA NEVADA.
Introduction .
.
.
.
.
.
.
The storm o f 18/19 December 1986: - A case study
.
4.2.1 Synoptic structure iind winds .
.
4.2.2 Radiometer measured liquid water
.
.
.
4.2.3 Precipitation microphysics
.
.
.
A)
Results from Tahoe Dormer .
.
B)
Results from Kingvale .
.
.
4.2.4 Discussion of results .
.
.
.
.
.
.
.
.
.
.
.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
60
. 6 3
. 6 3
68
. 7 6
. 7 6
. 7 7
80
4. 3
4. 4
4. 5
4. 6
V.
The storm of 19/20 December 1986: - A case study
4.3.1 Precipitation and microphysics .
.
.
The storm of 22/23 December 1986: - A case study
4.4.1 Precipitation and microphysics .
.
.
The Storm of 03/04 January 1987: - A case study
4.5.1 Precipitation and microphysics .
.
.
Summary and conclusion
.
.
.
.
.
.
.
.
.
.
.
89
95
100
108
113
118
124
SUMMARY, CONCLUSION AND RECOMMENDATION FOR FUTURE
W O R K . ..................................................................................
.
134
REFERENCES
APPENDIX
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
viii
LIST OF FIGURES
Figure 1.1
Contour map (A) and surface plot (B) of the study area showing major
topographic features.
Figure 1.2
Temporal plot of data from co-located precipitation rate measuring devices.
Figure 2.1
Microwave atmospheric absorption in clear air, clouds, and rain, V is absorption
coefficient and "e" is attenuation coeff. From Westwater and Decker(1977).
Figure 2.2
Side view of DRI (top) and USBR (bottom) radiometers.
Figure 2.3
Precipitation and 3-hourly temperature record at Kingvale, California on 5-6
December 1986
Figure 2.4
Integrated cloud liquid water and precipitable water vapor as measured by DRI
(A) and USBR (B) radiometers.
Figure 2.5
Brightness temperature values between 1300 and 2400, 5 December 1986 for
USBR and DRI radiometers.
Figure 2.6
Scatter plot of brightness temperatures for 20 GHz (x-axis) and 30 GHz channel
of USBR (A) and DRI (B) radiometers.
figure 2.7
Scatter piot o f USBR versus DRI measured values o f integrated cloud liquid
water and precipitable water vapor.
Figure 2.8
Integrated cloud liquid water and precipitable water vapor in wet conditions
made by DRI (top two frames) and USBR (from Jacobson et al. 1986)
radiometers (bottom four frames).
Figure 2.9
Scatter plot of rawinsonde and radiometer measured precipitable water vapor for
DRI (top) and USBR radiometers.
Figure 3.1
500 mb analysis of height (solid) and temperature (°C dashed) at 0000 UTC 27
March 1985.
Figure 3.2
Temperature sounding at KGV, 1500 26 MArch 1985.
Figure 3.3
Temporal analysis of the 26-27 March 1985 storm as it crossed the central Sierra
Nevada. Time-height radar reflectivity profile over KGV (A). Minimum
detectable signal over KGV is 8 dBz. Precipitation rate (B) and 5180
measurements (C) at KGV.
Figure 3.4
Frontal positions and time-height cross section of equivalent potential
temperature (K) from 1600 UTC 26 March 1985 to 0500 UTC 27 March 1985
overlaid on 5lsO values for the same duration.
Figure 3.5
Wind speed, wind direction (A) and radiometer measured integrated cloud liquid
water and precipitable water vapor(B) for 26 March 1985.
Figure 3.6
Precipitation rate and 5180 measurements at HM on 26-27 March 1985.
9
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Figure 3.7
Isotope derived "temperature" according to the Equation 5180 = 0.9T - 3.4 of
precipitation that fell on sites KGV (dashed) and HM on 04 March 1985.
Figure 3.8
Frequency analysis of the S180 collected at KGV showing the equivalent
temperature for 1984/85 and 1985/86 seasons.
Figure 4.2.1
Surface weather map (top) and 500 mb (bottom) layer heights at 1200 UTC,
December 18,1986.
Figure 4.2.2
Measurements of wind speed and direction at Slide Mountain weather station
(A) and wind rose analysis (B) for December 18,1986.
Figure 4.2.3
Time versus height cross section from KGV soundings (A); Relative humidity
(solid), temperature (dashed) and (B) equivalent potential temperature contours.
A full wind barb is 10 m/s. Shaded areas are where 30Jdz < 0. Time increases
from left to right.
Figure 4.2.4
Radiometer measured integrated cloud liquid water and precipitable water vapor
at KGV (A) and TRK (B) for December 18,1986.
Figure 4.2.5
Spatial distribution of radiometer measured integrated cloud liquid water at TRK
on December 18,1986.
Figure 4.2.6
Total precipitation (mm) accumulated on December 18,1986 at stations around
TRK (A) and across the barrier (B).
Figure 4.2.7
Frequency analysis of the radiometer measured integrated cloud liquid (mm) and
precipitable water vapor (cm) at TRK and KGV on December 18, 1986.
Figure 4.2.8
Kj-band radar echo top height versus time at KGV on December 18,1986.
Figure 4.2.9
Cross section of the clouds over the barrier showing radar echo and temperature
structure on December 18,1986 at 1700 UTC (from Deshler et. al., 1990).
Figure 4.2.10
Relative numbers (in percent) of observed ice crystals on December 18,1986 at
TD. Heavy rimed, warm origin, planar and other type of crystals (A) and rime,
aggregate and fragments (B).
Figure 4.2.11
Precipitation rate measurements on December 18,1986 at KGV (A), Donner
Grade (B) and Tahoe Meadows (C). Donner Grade and Tahoe Meadows are
the nearest precipitation gauge stations to TRK and TD respectively.
Figure 4.3.1
Same as Fig. 4.2.3, but for 19 December 1986.
Figure 4.3.2
Same as Fig. 4.2.2, but for 19 December 1986.
Figure 4.3.3
Same as Fig. 4.2.4, but for 19 December 1986.
Figure 4.3.4
Same as Fig. 4.2.5, but for 19 December 1986.
Figure 4.3.5
Same as Fig. 4.2.6, but for 19 December 1986.
Figure 4.3.6
Same as Fig. 4.2.11, but for 19 December 1986.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Figure 4.3.7
Same as Fig. 4.2.8, but for 19 December 1986.
Figure 4.4.1
Same as Fig. 4.2.3, but for 22 December 1986.
Figure 4.4.2
Same as Fig. 4.2.2, but for 22 December 1986.
Figure 4.4.3
Same as Fig. 4.2.4, but for 22 December 1986.
Figure 4.4.4
Same as Fig. 4.2.5, but for 22 December 1986.
Figure 4.4.5
Same as Fig. 4.2.6, but for 22 December 1986.
Figure 4.4.6
Same as Fig. 4.2.11, but for 22 December 1986.
Figure 4.4.7
Same as Fig. 4.2.10, but for 22 December 1986.
Figure 4.4.8
Same as Fig. 4.2.7, but for 22 December 1986.
Figure 4.5.1
Same as Fig. 4.2.3, but for 03 January 1987.
Figure 4.5.2
Same as Fig. 4.2.2, but for 03 January 1987.
Figure 4.5.3
Radiometer measured integrated cloud liquid water and precipitable water vapor
over TRK at 03 January 1987.
Figure 4.5.4
Radiometer scan measurements made over TRK on 03 January 1987. Scan # 1:
2130-2200, #2: 2300 - 2330, #3: 0000 - 0300 UTC.
Figure 4.5.5
Same as Fig. 4.2.6, but for 03 January 1987.
Figure 4.5.6
Same as Fig. 4.2.11, but for 03 January 1987.
Figure 4.5.7
Same as Fig. 4.2.10, but for 03 January 1987.
Figure 4.6.1
Cumulative mass percentage for overall storm as a function of crystal mass: total
snowfall (1), riming (2), aggregation (3) and fragmentation (4) for 03 January
(A), 22 December (B) and 18 December (C). ( From Demoz et al. in
preparation for publication).
Figure 5.1
A conceptual model of a mature winter orographic cloud system over the Sierra
Nevada. See text for explanation of the different regions.
\
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
1
PARTI
INTRODUCTION.
Modification of weather systems as a means for additional water supply has been studied
since early 1950’s. As advances in new cloud physics instrumentation are made new concepts are
developed and tested. The ground-based and satellite techniques which were heavily depended
upon for weather modification research in the late 1950’s and the 1960’s gave way to the more
advanced and systematic ways of using aircraft instrumentation in the 1970’s. Most of the effort
was concentrated in the mountainous regions for enhancing the snowpack by modifying winter
orographic storm systems.
One such project was the Sierra Cooperative Pilot Project (SCPP). The primary project
objective was to investigate cloud seeding as a means of increasing snowpack on the Sierra
Nevada and develop a firmer foundation for operational weather modification programs in this
area.
This project used ground-based remote sensing devices, ground and aircraft in situ
observations and a numerical targeting model in randomized seeding experiments. Most of the
work done on this project was concentrated on the upwind side of the Sierra Nevada mountains.
The storm systems affecting this area were well documented and have been analyzed by a number
of researchers for the upwind mountain region. Very little effort has been spent in studying and
documenting processes that are responsible for the snowpack on the down wind side of the
Central Sierra Nevada. Some work is being done by scientists at the Desert Research Institute
(DRI) to study this specific problem. The work of Huggins et al.,(1990) and Mitchell (1989) are
examples. One of the aims of the present work is to characterize the storm systems which affect
this region microphysically and dynamically and to compare the upwind and downwind parts of
the mountain barrier. The work and results of this thesis forms a part of the overall effort in
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
achieving that goal.
The storms that affect the Sierra Nevada usually originate over the Pacific Ocean, with
the warmer, and moister air masses coming from the southwest and colder storms coming from
the Gulf of Alaska. Heggli and Rauber (1989) have recently categorized the types of storms
affecting the central Sierra Nevada into five major systems and discuss the moisture content
associated with these systems. They studied a total of 63 storms which affected the SCPP area
from 1983/84 through the 1986/87 winter field seasons. The classification was done according to
storm trajectory, wind flow and cloud fields as derived from 500 mb charts and satellite images.
The two main categories, which were either predominantly zonal or predominantly meridional
(type A and type B respectively) were subdivided into; developing storm embedded in strong
westerly or southwesterly flow (A l), moderate amplitude shortwave associated with an occluded
storm (A2), split flow in the middle troposphere associated with a dissipating storm (A3), cutoff
flow or large amplitude shortwave near 40° N latitude (B l) and large amplitude long wave
pattern producing cold northerly storms (B2).
In this study, storms from the above groups are selected.
The evolution of the
radiometer measured liquid water with respect to the cold front and the precipitation
microphysics will be presented. Most of the work will be focused into extending the observations
and conclusions made upwind of the barrier into the downwind side.
1.1
OBJECTIVES.
Specifically, the objectives of this work can be stated as to
I)
describe the temporal evolution of the integrated cloud liquid water and vapor from
storms affecting the Sierra Nevada. The effect of topography and winds on the cloud liquid water
and its relationship with the cloud microphysics observations on the ground will be presented.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
The integrated cloud liquid and vapor data used in this study, particularly on the downwind side,
are observed using a newly built ground-based radiometer. The performance of the radiometer
is also analyzed.
II) describe the microphysical observations from winter clouds and explain relationships
between riming, aggregation and fragmentation as well as their relation to frontal positions. A
possible indication of secondary ice producing mechanisms from the ground observed habits is
also discussed.
III) compare the upwind and downwind site observations of ice crystals and liquid water.
This comparison will be used to better understand storm efficiency as a result of the topography.
IV) use the 180 / 160 to gain insight into the modification of the frontal structure of the
storms as they cross the mountain barrier and explain the frequently observed colder crystals past
the barrier.
V) recommend further research based on the information obtained.
The thesis is organized as follows. Part I presents general introduction and motivation,
objective of the study, data sources and a brief description of the instruments used. In Part II,
the principles of microwave radiometry and a comparison of the radiometers used is presented.
Part III discusses the use of stable water isotopes in gaining knowledge of storm characteristics
in the Sierra. A case by case analysis of the storms that impacted the area on 18, 19 and 22
December and 03 January 1986/87 winter season is presented in Part IV. The main conclusions
and summary of each part and recommendations3 for future work will be presented in Part V.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
120:30
12 1 :0 0
1 2 0 :0 0
I
• TRK
Tahoe
39:00
•
0
N
> Auburn
10
rlacerville
Sacramento
cS
B
Figure 1.1 Contour map (A) and surface plot (B) of the study area showing major topographic
features.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
1.2
DATA SOURCES AND INSTRUMENTATION.
Data used in the case studies that follow were collected during the SCPP conducted in
the central Sierra Nevada. The map and topography of the project area is presented in Fig. 1.1.
The names and station ID location of the sampling sites is given in Table 1.1. The data used in
this study were collected during the SCPP years.
In the 1986/87 winter, two radiometers, one operated by the United States Bureau of
Reclamation (USBR) and the other by D RI were in use. The USBR radiometer was located at
KGV and the DRI radiometer was located at TRK. The full discussion of these radiometers is
the subject of Part I o f this thesis.
Part II and Part IV also used data from the winter of 1986/87 while the isotope analysis
data for Part III were from the winter of 1984/85. Several precipitation gauges over the entire
area, snow crystal and tracer sampling, instrumented aircraft and remote sensing devices have
been used over the years. In the next section a description of the instruments used is presented.
1.2.1
RAWINSONDE.
Rawinsonde data were collected during the 1986/87 winter at two stations (see Fig. 1.1a),
at Lincoln (LNC), California (57 m MSL) and about 90 km west of Kingvale (KGV). Both
stations KGV and SHR are upwind of the main crest of the Sierra Nevada and have a cross­
barrier alignment. The rawinsonde system used was the VIZ Acculock type and was launched,
generally, every three hours during the storm events. Soundings made during the SCPP were
contoured stored. The 3-hour contours used in this study are from an inventory of the data
cataloged at DRI.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
1.2.2
PRECIPITATION GAUGES.
A network of precipitation gauges was in operation during the 1986/87 winter season and
was operated by SCPP and DRI. The gauges used were mostly weighing bucket type gauges
designed and built by Electronic Techniques Inc. (ETI). They had a capacity of 200 mm with
digital resolution of 0.1 mm (Hemmer et al., 1987) and are fully discussed in Rilling et al., (1985).
In this work a 15 minute averaged data from these instruments were used. A summary of the
stations used is given in Table 1.1.
Table 1.1 A summary of stations used for precipitation sampling.
STATION
LOCATION
ID
NAME
LATITUDE
LONGITUDE
ELEVATION(MSL)
NW6
BLUE CANYON
39:16:33
120:42:28
1609
S it
SIERRA SNOW LAB.
39:19:30
120:22:04
2087
S13
KINGVALE
39:19:03
120:26:05
1856
S21
CASTLE VALLEY
39:21:03
120:21:13
2254
S29
HOBART MILLS
39:24:02
120:10:50
1783
S33
BOCA RESERVOIR
39:23:18
120:05:31
1691
S82
KINGVALE*
39:19:03
120:26:05
1856
S84
DONNER GRADE
39:19:57
120:17:25
1945
S86
GOOSE MEADOWS
39:16:45
120:12:50
1890
S99
YUBA GAP
39:18:56
120:36:48
1760
SN1
SAGEHEN CR
39:24:45
120:14:13
1931
SN4
BROCKWAY SUMMIT
39:15:43
120:04:18
2195
SN5
APOLLO WAY
39:16:23
119:56:21
2225
SN6
TAHOE MEADOWS
39:17:52
119:55:04
2604
SN7
Mt. ROSE RELAY STN.
39:19:23
119:56:17
2945
SN9
Mt. ROSE RESORT
39:19:42
119:53:07
2524
SND
DAVIS CR
39:18:17
119:49:50
1573
39:28:32
119:58:50
1890
SNH
FULLER LAKE
A high resolution Belfort gauge.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
Since the data from different stations is to be used a comparison of the co-located gauges
and two close-by stations is helpful. Such a comparison during this season (Fig. 1.2) shows a very
good agreement (correlation coefficient of 0.87) between the outputs of two gauges. Hemmer
et al., (1987) also showed that the correlation coefficient could be as low as 0.6 and as high as
0.99.
SB 2 (-high
Sjl3 (tipping bjuckejt)
TIME (15 MINUTE STEPS)
Figure 1.2 Temporal plot of data from co-located precipitation rate measuring devices.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
1.2.3
GROUND CLOUD MICROPHYSICS SAMPLING.
Ground photo micrographs of ice crystals and their melted drops were recorded at Tahoe
Donner by DRI personnel. Details of the method are presented in Part IV. Similar habit
classification was also done at KGV without the melted drops. Surface observations of wind and
temperature were recorded routinely by DRI at Slide Mountain weather station (SM).
1.2.4
DRI K.-BAND RADAR.
The DRI IQ-band (8.6 mm) radar was operated at KGV for the 1986/87 season and echo
top heights are derived from it. The DRI radar operated exclusively in vertical-pointing mode.
This radar had an effective vertical range of 8 km, 64 echo return gates with 125 m range per
gate. It uses a 1 meter diameter paraboloid antenna with a beam width of 0.5°. A table of the
radar characteristics is given below. The data are usually displayed in echo intensity (dBm) units.
The relation of the echo intensity to the returned power is given by;
P-101og10P
where Pr (in mW) is the power returned to the radar for a nominal transmitted power of 1 mW.
The characteristics of the DRI K, - band radar are given in Table 1.2.
Aircraft data, standard meteorological weather charts and C:band (5 cm wavelength)
radar data were all used when necessary in obtaining information on the case studies analyzed.
But since the information used from these instruments is minimal and their description available
in the literature a full description is not given here. The data from all the instruments used in
SCPP was routinely reduced and stored. Most of the data is available from DRI.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
9
Table 1.2 CHARACTERISTICS OF THE DRI KA - BAND RADAR.
ANTENNA
DIAMETER (m)
BEAM WIDTH (°)
GAIN (dB)
NOISE TEMPERATURE (°K)
TRANSMITTER
WAVELENGTH (cm)
FREQUENCY (MHz)
PEAK POWER (KW)
PULSE WIDTH (fts)
PULSE REPETITION FREQUENCY (Hz)
RECEIVER
NOISE TEMPERATURE. (°k)
NOISE FIGURE
BANDWIDTH (MHz)
DYNAMIC RANGE (dB)
NOISE POWER (dBm)
MINIMUM DETECTABLE SIGNAL(dBz)
EQUIVALENT REFLECTIVITY FACTOR
RAIN @ 25 km (dBz)
1.2
0.5
48.
0.86
34860
100
0.5
1000
298
2.7
10
80
101
-111
-25
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
10
pa rt
n
RADIOMETRY
2.1
INTRODUCTION
Passive microwave observation of the atmosphere was first reported by Dicke et al.,
(1946) based on theoretical predictions of 0 2and HzO absorption made by Van Vleck which were
reported later in Van Vleck (1947).
Since then the use of microwave radiometers for
measurements of atmospheric water vapor and liquid water have become increasingly important.
This is specially true of weather modification projects which have used ground-based microwave
radiometers for continuous measurement of vapor and liquid. Reports made by Long and Walsh
(1984), Holroyd and Super (1984), Askne and Westwater (1986), Sassen et al., (1986), Heggli et
al., (1987), Heggli and Rauber (1988), Reinking and Meitin (1989), Westwater and Kropfli
(1989) and Canavero et al., (1990) are some of the studies that have used the dual-channel
radiometer to investigate different atmospheric phenomena.
In the early 1960’s, Barret and Chung (1962) and Meeks and Lilley (1963) suggested ways
of determining profiles of HzO and temperature from measurements of atmospheric radiation
in the 1.35 cm line of H20 and 0.5 cm line of 0 2 absorption bands respectively. These early
studies have a sound physical basis.
But the nature of the problem of getting these
measurements is "ill-posed", i.e., it requires a priori assumptions to solve the mathematical
equation and achieve physically useful results. Twomey (1966), Westwater and Strand (1968) and
Rodgers (1976) give a good discussion of the formulation of the mathematical problem.
The Rayleigh attenuation and emission of clouds in the microwave region and the
methods of dual-channel radiometry for measuring liquid and vapor which Staelin (1966)
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
11
proposed were laid out by Weswater (1977, 1978).
Design and operation of dual-channel
radiometry have been presented by Hogg et al., (1983) and Guiraud et al., (1979).
The methods of estimating liquid and vapor can be either physical or linear statistical
inversion technique. A very good comparison of the different methods used is given by Wei et
al., (1989). Error estimates associated with the retrieval processes are given by Westwater
(1978).
A good agreement was also found (Westwater et al, 1990) when ground based
observations were compared with theoretical atmospheric emission and attenuation models of
Liebe (1989) and Waters (1976).
For temperature profiles, most researchers use multi-channel radiometers and a dual­
channel radiometer for vapor and liquid measurements.
These radiometer measured
temperature, vapor and liquid compare fairly well with radiosonde data ( Decker et al., 1978,
Hogg et al., 1983, Heggli et al., 1987 and Blaskovic et al., 1989) and the accuracy of the method
was discussed by Westwater (1978) and Snider et al., (1990).
Several weather modification projects have used and are still using radiometers in their
study of supercooled liquid water content.
The SCPP, designed to study precipitation
enhancement potential of winter-time storms in the central Sierra Nevada, have made use of two
radiometers: one operated by the United States Bureau of Reclamation (USBR) and another by
the University of Nevada, Desert Research Institute (DRI).
The USBR radiometer was
calibrated and tested against the National Oceanic and Atmospheric Administration (NOAA),
Wave Propagation Laboratory radiometer. The results showed a good agreement and were
reported by Heggli et al., (1987).
In this work, we present a comparison between the USBR and the DRI radiometers.
In the following sections a brief theoretical background, design and construction, and comparison
of both radiometers are presented.
The data for this study were selected from SCPP
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
measurements made at Kingvale (KGV), California on 05-06 Dec. 1986. As shown in Fig 1.1,
Kingvale is located approximately 8 km upwind of the Sierra Nevada Crest at an elevation of
1857 m MSL.
2.2
THEORETICAL BACKGROUND
The radiation incident at the earth’s surface from zenith is given by
7J_0- /(!exp[-j’o'Y(z>fe]+/*5(z>YCz)exp
*v,Z> 2 /jv 3
c*
(1)
1
*
ex p « --l
Where B is Planck’s function at temperature T and frequency v, h is Planck’s constant,
c is the speed of light, k is Boltzmann’s constant and y is the volume absorption coefficient of
the atmosphere. The first term represents the extra-terrestrial radiance, Ie before entering the
earth’s atmosphere, and the second term is the black body source term of the atmosphere. The
quantity in square brackets, is the optical thickness of the atmosphere (t). This equation does
not include scattering.
In the microwave region (1 to 100 GHz), the Rayleigh-Jeans law of radiation is valid (i.e.
hv/kT < < 1, is assumed in Planck’s function: the function is then expanded in a MacLaurin
series where only the first two terms are considered) and can be used to convert the above
equation into one containing brightness temperatures. The resulting approximation is the well
known Rayleigh-Jeans law;
B(v,I) - 2
c2
For hv/kT < < 1, the brightness temperature is linearly related to the thermometric temperature
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
13
through the emissivity (Tb =«T, for a gray body).
The Planck function which is an explicit function of temperature alone can be written as
a function of position through the dependence of temperature on height(z). Thus, the radiation
incident at surface is dependent on the temperature profile along the ray path.
This is
mathematically related to the brightness temperature (using KirchofPs law) as,
K - 2*^r><v)
The above discussion when used in equation (1) leads to the expression;
(2)
»o
where t, the transmittance is defined by t(0,z) = exp[-r(0,z)]. By introducing a mean radiating
temperature of the atmosphere Tn, equation (2) can be written as
r i-T4lt(o,«»)+rJ. / ‘(--|> fe-r4,t(o,»)+rji-/(o ,“)]
/o' et
(3)
which gives the following;
T -T
Tm Tie
(4)
x(0 , » ) - l n ( ^ L ^ )
Given T,,, and measuring Tb will allow t(0,» ) to be calculated from the above equation.
The cosmic radiation brightness temperature, T,*, may be set equal to 2.7 °K for frequencies
greater or equal to 10 GHz.
In the microwave region of the electromagnetic spectrum, water vapor, cloud liquid water
(excluding rain) and oxygen are the main absorbers (see Fig. 2.1). Therefore, the optical
thickness can be written as,
T" Tp+Tt+To
where v,l, and o refer to vapor, liquid and oxygen.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
14
: T =* 288.16K
I P» 1013.25 mb
■pVapor * 7.5 g/m3
*°Liquid * 0.1 g/m
■R * 12.5 mm/hr.
eRain
Total Clear
Vapor
Liquid
90
100
Frequency, (GHz)
Figure 2.1 Microwave atmospheric absorption in clear air, clouds, and rain, V is absorption
coefficient and "e" is attenuation coeff. From Westwater and Decker (1977).
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
15
The optical thickness is related to the density of the variables by the following equations.
r f 0 P ifyb-kyV
T*“iTp
Where M is the mass of water per unit volume of air,
is the mass absorption coefficient (in
units of n^kg'1), V is the precipitable water vapor(in units of cm) and L is the integrated liquid
water path(in units of mm).
The absorption coefficients are related (Staelin, 1966, Westwater, 1965 and Van Vleck,
1947) to temperature (k), pressure (mb) and water vapor density p (g/m3) by the following
equations.
For water vapor (cm'1):
644
a
T .-^ ( 1 * 0 .0 1 4 7 ^ )
4
i_______________________
(v-2123Sf*(,Ay>f (v+22^35)2+(Av)2
(7)
42.55xl0-9. pV-2-Av
j6/l
where the line width is given by
Av-2-58xl0'3(l+.0147pZ7p)----1L-=r.
(3T378)425
For cloud liquid (mm'1):.
[M x l(f
W
‘
*
w
W
(8)
10 x Aa
in the equation JL is wavelength, M is liquid water density of cloud (g/m3). This is valid for k
between 0.8 cm and 3 cm and is good for droplets small when compared to JJItt . Rayleigh’s
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
16
theory is applicable to cloud droplets from few micrometer to few hundred micrometers. In this
region attenuation coefficient is independent of drop size distribution and depends only on liquid
water content (Westwater, 1972).
The frequencies used for liquid water and vapor are 20.6 GHz and 31.6 GHz. Referring
to Fig. 2.1, 20.6 GHz is near the water vapor absorption line at 22.2 GHz and 31.6 GHz is near
the water vapor absorption minimum above the 22.2 GHz line. Thus the 20.6 GHz responds
mainly to vapor and 31.6 GHz to liquid. The absorption by liquid at 31.6 GHz is about 2.2 times
greater than at 20.6 GHz and absorption by vapor is lower by about 1.8 times under standard
conditions (Hogg et al., 1983). The choice of 20.6 GHz also minimizes the effect of pressure
broadening of the water vapor absorption line, i.e. absorption is nearly invariant with pressure
at the 20.6 GHz (Guiraud et al., 1979).
In the frequencies considered, excluding rain again, oxygen absorption needs to be taken
into account. Since oxygen concentration in the atmosphere is constant, its absorption can be
calculated from a standard atmosphere at 20.6 GHz and 31.6 GHz. The theory o f microwave
absorption by 0 2was given by Van Vleck has been verified by Whitehurst et al., (1957), Stafford
and Tolbert (1963) and Barrett et al., (1966) in atmospheric measurements. The 0 2 absorption
in microwave region is centered at 60 GHz and is caused due to magnetic dipole rotational
transitions. The absorption coefficient of oxygen is given by (Westwater, 1965)
fin
/ - &
E
Av-
«
Av-
i
(v.-vf+ A v* (v„+v)2+Av^
In the above equation (9), C is a constant depending on units of the absorption
coefficient a0, E„ is energy of the n111state, p2 is a constant and v„ is the resonant frequency and
Av„ is the line broadening of the n* state and yet unspecified.
For a discussion of the
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
parameters in 0 2 absorption Van Vleck (1947b) and Westwater (1965) may be consulted.
Using the above equations, the absorption measured at both frequencies may be written
as
(10)
The subscript "i" indicates the specific frequency. These equations can be solved for L
and V and give the following simultaneous equations.
(11)
The coefficients can be calculated from a climatology soundings by the method of linear
statistical inversion ( Westwater and Guiraud, 1980) or from the equations given above and are
geographically dependent.
A good discussion of the different methods of getting these
coefficients and their comparison is given by Wei et al., (1989). They found that the different
methods did not have large differences.
The uncertainty in liquid (AL) and vapor (AV) can be calculated from equation (11) for
a given ATmand ATb. For mean radiating temperatures of 266.23 K (20.6 GHz) and 267.34 K
(31.6 GHz), measured brightness temperatures of 195.5 K and 156.2 K, ATm = + 5 K and ATb
=0.5 K this would result in ±16% for vapor and ± 25% for liquid. If ATn=2 K and ATb=0.1
K is used, then the errors are less than ±8% for vapor and ±15% for liquid. An extended
discussion of the errors is found in Westwater (1978).
2.3 INSTRUMENT DESCRIPTION
Both the D RI and USBR radiometers operate at the same frequencies, 20.6 GHz and
31.65 GHz. The first frequency is near a water vapor absorption line while the second, 31.65
GHz, is located at vapor absorption minimum, and is more sensitive to emission by liquid water
than water vapor. The methods of reducing the observed brightness temperatures at these two
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
18
frequencies into water vapor and liquid water integrated depths have been explained by
Westwater (1978), Westwater and Guiraud (1980), and others.
The radiometers used have similar operating characteristics which are described by Hogg
et al., (1983). The equations relating the vapor or liquid to the brightness temperatures are
linear with coefficients calculated statistically based on historic meteorological conditions for a
given area. Westwater (1978), Hogg et al., (1983), Jacobson et al., (1986,1988), Heggli et al.,
(1987), Wei et al., (1989) and others have discussed the principles and methods of dual-channel
radiometry and some of the problems that arise in the measurements. Jacobson et al. reported
that wet snow leads to values of liquid and vapor which are cyclic (periodicity of v and both
channels out of phase by tt/2) when a radiometer was operated in scan mode. They also found
an optimum configuration of polarization that minimizes the problem of contamination. A
polarization parallel to the plane of incidence at 20.6 GHz aiid polarization normal to that plane
at 31.6 GHz introduces minimum (and almost constant) error.
The radiometers were calibrated by the "tipping curve" (Hogg et al., 1983) method. In
this method, absorption is computed at several antenna elevation angles each of which represents
an incremental number of atmospheres (secant of the angle). The results are plotted and the
fitted curve (which is a straight line) is then forced to pass through zero (i.e no atmosphere
corresponds to no absorption). The true zenith brightness temperature is then found from the
true zenith absorption by the following equation.
t
-
In (Tm - 2.9)/(Tm - T,)
In the equation, Tmis the mean radiating temperature of the atmosphere, Tb is the brightness
temperature and t is the absorption in nepers.
Condensation and melting of snow on the outer mylar sheets used as windows and
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
19
melting snow on reflectors are the main problems that can arise during periods of winter
precipitation. Currently, the NOAA/WPL group (Stankov et al., 1990) control the quality of the
radiometer data statistically.
2.3.1 USSR RADIOMETER
The arrangement of the USBR radiometer components is shown in Fig. 2.2A. The
receiver, feed horn, offset paraboloid and a mini computer (LSI-11/02) were housed in a
temperature stabilized trailer. The offset paraboloid and feed horn design are discussed by Hogg
et al., (1979). There were two flat reflectors on top of the trailer (azimuth and elevation flats)
fixed on two different bearings. The elevation flat, mounted 4S°C from the horizontal is rotatable
in elevation for calibration. Full-sky coverage by the antenna is achieved by rotation of both
bearings. Multi-layer mylar sheets (thickness 0.05 mm separated by a 50 mm air gap), protected
by the cowling, formed a low-loss window (Jacobson et al., 1988) for the reflected energy from
the exposed elevation flat.
The USBR radiometer had a bandwidth of 1 GHz, sensitivity of 0.26 °K at both
frequencies, and receiver noise temperatures of 680 °K and 725 °K at 20.6 and 31.65 GHz
respectively. The antenna beam width for both channels was 2.5°. Data acquisition was possible
at fixed azimuth and elevation angles and during continuous azimuth scanning at a fixed elevation
angle. A full 360° azimuth scan took about 12 minutes.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
20
SPINNING DISH
RECEIVER
CONTROL CONSOLE
GENERATOR
•i
Mobil* Scanning R adiom eter System
I
Double Mylar. Cowling
Window
Azimuth
Flat
and
Bearing
Mini- I
Computer
LSI-11
Elevation
Flat
20.8. 31.6 GHz
Radiometers
Offset
^aralffloid
Side
Figure 2.2 Side view of DRI (top) and USBR (bottom) radiometers.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
21
2.3.2 DRI RADIOMETER
The DRI radiometer has IF band pass of 50-550 MHz and a noise temperature of 445
°K and a sensitivity of 0.1 °K in both channels. The feed horn and offset paraboloid have a
design similar to the NOAA/WPL system with antenna beam widths of 2.5° at both frequencies.
The DRI design is shown in Fig. 2.2B. The receiver, offset paraboloid and elevation flat (or
reflector) are mounted on a common frame and turret assembly and can be rotated 360° in
azimuth. The flat spinning reflector can be tipped to provide viewing between the zenith and ±
90° from the zenith. The elevation angle is measured by an inclinometer mounted on the frame
holding the spinning reflector. The entire system is built into a van (see Fig. 2.2), with the offset
paraboloid and flat reflector on the roof of the van. The paraboloid is protected by an enclosure
which is open on only one end, while the reflector, which is exposed to the weather, is capable
of spinning (0 to more than 500 rpm) in order to shed precipitation particles. The reflected
electromagnetic wave from the offset paraboloid passes through a small mylar window located
just above the feed horn.
The body of the van is insulated and houses the receiver and computer which serves both
antenna control and data collection functions. The van is mobile and can be transported and set
up by one person with little effort. Software enables collection of tipping curve calibration data,
fixed zenith-pointing mode data and scanning data. The data collection rate during scans was
almost one record per degree with a full scan taking about 12 minutes (~ rate of .5 deg/s). All
the raw data are recorded so that they may be reprocessed later, if necessary, with different
calibration coefficients.
The main difference in the two systems is in the design to prevent melting or liquid
precipitation from interfering with the measurements. The USBR radiometer used multi-layer
sheets of mylar and a fixed elevation flat while D RI’s used an enclosure and a spinning reflector.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UJ
1
fi
I
uj
=!
- I
-4
—5
4:0
o:
X
3 .5
^
3 .0
^
2 .5
£
2.0
Q.
0 .5
0.0
0
3
6
9
12
15
18
21
0
3
6
9
TIME (UTC)
Figure 2.3
Precipitation and 3-hourly temperature record at Kingvale, California on 5-6
December 1986.
12
15
23
2.4
RADIOMETER COMPARISON CASE STUDY
The period selected for the comparison was 5-6 December 1986. The storm that passed
through the Sierra Nevada during this period was classified by Heggli and Rauber (1988) as a
moderate amplitude short wave associated with an occluded storm. The storm began at 2230 (all
times are UTC) on 4 December and ended at 1630 on 6 December 1986. The cold frontal clouds
moved into the experiment area after 0530 on 5 December. Radar echo showed the back edge
of the frontal band crossed Kingvale at 1600, 5 December. Radiometer measurements were
made throughout the storms while precipitation (see Fig. 2.3) was recorded from 0900 December
5 to 1300 December 6. The precipitation rate peaked at 3.5 mm/hr at 1345 on 5 December, then
gradually decreased with time.
The atmospheric retrieval coefficients used to calculate liquid and vapor depths were the
same for both the radiometers while the tipping calibration coefficients were determined
separately for each instrument on 2 December. The USBR radiometer operated only in zenithpointing mode. The DRI radiometer was used in zenith as well as scan mode. Even though both
instruments had user-adjusted sampling rates and averaging intervals (via menu-driven software),
the sampling rates used during this storm were not the same. The USBR radiometer recorded
averaged data every 2 minutes while the DRI radiometer zenith records were 10 minutes
averages. Each 10 minute DRI average was computed from about 350 point samples.
A few scans were made with the DRI radiometer at an antenna elevation angle of 25°.
These scan data were then averaged in post-processing and normalized to the zenith to give a
value for each 12-min scan interval. For comparison purposes the data from both radiometers
were averaged over a longer time interval (20 min).
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
24
20 MINUTE AVERAGE
0000 (GMT)
5 DEC 1986 to 1630 (GMT)
6 DEC 1986
r-Z
-s
0.80
2.00
Liquid
Vapor
A
precipitation
1 0.60
1.50 r»
S
(V
“1
C
u
0)
a
u 0.40
*4
>>
>
0
•o
T><fC+-1
II
o
1.00
*
r
o’ 0.20
3
0.50 i
I
U
0.00
I
1-------- 1-------- r
6
9
12
15
JQ 0.00
1---- I”
18 21
0
TIME (GMT)
3
9
12
15
20 MINUTE AVERAGE
0000 (GMT)
5 DEC 1986 to 1630 (GMT)
6 DEC 1986
r-sz
0.80
-
Liquid
2.00
Vapor
0.60
accumulated snow was cleaned
1.50 7
0.40
1.00
o- 0.20
0.00
0.00
TIME (GMT)
Figure 2.4 Integrated cloud liquid water and precipitable water vapor as measured by D R I (A)
and USBR (B) radiometers.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
O
Q.
C
o
CD
Q.
with permission of the copyright owner. Further reproduction prohibited without permission.
1300 -
2400 5 Dec. 1986
AT KGV
70
TB30
TB20
60
USBR RAD.
USBR RAD. DISH CLEANED
50
PU
40
W
30
H
20
CO
CO
rv
< —
< —
< -
'< —
10
w
H
70
g
60
rV
50
• TB30
v TB20
DRI RAD..
40
30
20
SCAN
15
16
MISSING
17
18
19
20
21
22
23
TIME IN UTC
Figure 2.5 Brightness temperature values between 1300 and 2400 5 December 1986 for USBR
and DRI radiometers.
K
26
Figure 2.4 shows 20-min averages of the liquid water and water vapor depths from both
radiometers for the storm period sampled between 0000, 5 December and 1630,6 December. At
the top of the plot the mode of operation of the instrument is shown. Short bars indicate the
scan mode of operation at fixed elevation (indicated by S on the upper right corner of the graph)
while longer bars are for the zenith mode of operation (indicated by Z). The spacing of lines in
the mode of operation shows that data were not uniformly collected by the DRI radiometer
throughout the comparison period.
Figure 2.4 shows that, generally, the two radiometers recorded very similar trends in
liquid and vapor depth. The DRI radiometer values were slightly greater than USBR values.
There were, however, several periods of disagreement, most notably in the vapor depth. These
periods, where prominent vapor (and liquid) peaks occurred are marked with an "x" on Fig. 2.4B.
These relative vapor maxima are absent in Fig. 2.4A. Aside from these obvious differences the
two time-sequential plots from both the instruments followed each other closely. The slight
consistent offset in both channels could have been caused by a slightly inappropriate tipping curve
calibration in one or the other of the radiometers.
The brightness temperatures between 1300 and 2400, 5 Dec. are plotted in Fig. 2.5 for
both radiometers. For Fig. 2.5, the USBR radiometer values are 2-min averages while those of
DRI are variable (2-min to 10-min at times). The "x" regions in Fig. 2.4 are also indicated by "<lines.
A correlation coefficient of 0.74 and 0.73 was found between the 20 GHz and 30
GHz brightness temperatures respectively. A scatter plot of the data in Fig. 2.5 is also made in
Fig. 2.6. The spread between the 20 GHz and 30 GHz of DRI radiometer is much less than that
of the USBR which is some indication that the USBR channels were affected.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
27
1300 -
o
o
#
2400
5 D E C . 1 9 8 6 AT KGV
5000
'ST
4000
2
UJ
I—
<f)
UJ
z
UJ 3000
t—
X
o
cc
CL
CD
2000
20 GHz
6000
o 5000
o
*
“'W
ST’
^
4000
UJ
(/)
C/5
3000
X
o
QC
CD
2000
20 GHz
1000
'1000
2000
3000
4000
5000
6000
BRIGHTNESS TEMP.(K)*100
27
Figure 2.6 Scatter plot of brightness temperatures for 20 GHz (x-axis) and 30 GHz channel of
USBR (A) and DRI (B) radiometers.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
28
20 min. average liq. water values in mm
correlation coeff. = 0.8250
m ean
c.
<u
0]
4-»
e
a
T3
<0
c.
oo
cc
Q
oo,
.2
-
oo
mean
oo
io@
USBR radiometer
20 min. average vapor values in cm
correlation coeff. = 0.6B26
1.5 -
c_
O
)
■
M
O)
s
o4
•»
■a
ai
c.
1.4 1.2
mean
-
a>
a
■are-
1
mean
1.2
1.4
1.6
i. a
USBR radiometer
Figure 2.7 Scatter plot o f USBR versus D R I measured values o f integrated cloud liquid water
ana precipitable water vapor.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
29
The isolated radiometer disagreements might have been related to ambient weather
conditions at the radiometer site. Surface temperature was greater than 3°C before 0600, 5
December, 2°C between 2200, 5 December and 0100, 6 December and decreased slightly
afterwards.
Continuous temperature records were unavailable between 0600 and 2200, 5
December but 3-hourly radiosonde profiles indicated that the temperature near the surface
remained warmer than 0.3°C.
Subfreezing temperatures occurred in the post-cold frontal
environment after 0600 on 6 December. Precipitation (see Fig. 2.3) started at 0900,5 December
and ended at about 0000,6 December. Additional small amounts of precipitation were observed
near 1500 on 6 December. The temperature was warmer than 0°C for almost all of the time that
precipitation occurred (except between 1100 and 1200) suggesting that melting snow or rain on
the reflector may have caused the relative maxima in vapor and liquid seen in the USBR data
(Fig. 2.4B). In addition the radiometer operator’s log also indicated that wet snow was blowing
against the lower part of the mylar window from 1200 to 2400, 5 December. The accumulated
snow was cleaned at 1800 December 5 when a decrease in both vapor and liquid values is evident
in Fig. 2.4B. As shown by Jacobson et al., (1986), wet reflectors cause increases in measured
brightness temperatures in both channels with the resulting values of vapor and water increasing
simultaneously. The increase in the vapor is much larger than the increase in liquid. The same
authors found that even a layer as small as 0.15 mm on the reflector can cause large errors in
the measurement o f the brightness temperature. This problem of contamination of the data by
melting snow on the reflector was not apparent in the DRI data (Fig. 2.4A). This is an indication
that the rotating reflector on DRI’s radiometer successfully shed the melting particles and
prevented the reflector from getting wet.
Table 2.1 presents a statistical summary of the USBR and DRI radiometer 20-min
averaged data and Fig. 2.7 shows scatter plots of the same data. It shows that the correlation
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
30
coefficient between liquid values, 0.83, was better than for the vapor values, 0.683 for the data
used to plot Fig 2.4. The mean liquid values for both radiometers were almost the same, with
DRI’s mean liquid value being slightly greater. This is also apparent on the scatter plot (Fig.
2.1A) where the majority of points reside above the 1:1 ratio line. As previously mentioned the
difference in mean values may have been caused by an differences in tipping curve calibration.
Table 2.1.Statistical summary of USBR and D RI radiometer measured integrated water vapor and
liquid depth at Kingvale, California from 0000 Dec. 05 to 1630 Dec. 06 1986.
QUANTITY
RADIO
METE
R
MEAN
STD. DEV.
LIQUID (mm)
USBR
0.119
0.100
DRI
0.134
0.111
USBR
0.798
0.256
DRI
0.792
0.122
VAPOR (cm)
CORRELATION
COEFFICIENT
0.825
0.92*
0.683
0.8?
* For data after 0900 Dec. 06,1986 when temperatures were below 0°C and no precipitation was
falling. Number of data points for this comparison are 34.
The lower correlation coefficient for vapor values was likely due to the greater variability
in the USBR vapor. Although the mean values were almost the same for both instruments, the
standard deviation for the USBR vapor was almost twice as great as that of DRI vapor. The
larger spread in USBR vapor values was caused by the relative peaks (marked in Fig. 2.4) which
were due to contamination from melting snow on the reflector or accumulations of snow on the
mylar window. These extreme values are also evident in Fig. 2.7A as outliers (USBR values 2
1.2 cm). For values < 1 cm, the DRI vapor values tended to be consistently larger, as was the
case with liquid.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
15:37 to 15:50 05 Dec. 1986
4
vapor
liquid
3
ar
0
16:09 to 16:22 05 Dec. 1986
— vap o r
— liquid
Vapor
Liquid
12:05
_ 4
12:32 ^
8
-
r o
>
13:29
120
180
240
300
360
4>—Azimuth Angle (Peg)
Figure 2.8 Integrated cloud liquid water and vapor in wet conditions by DRI (top two frames)
and USBR (from Jacobson et al, 1986) radiometers (bottom four frames).
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
32
00:00 November 20 to 09:00 December 06 1906
1.4
1.2
precipitable
water
1.6
1
,8
Rawinsonde
,6
4
.2
0
1.2
1.4
1.5
1.4
1.5
cm
integrated vapor depth
00:00 November 26 to 09:00 December 06 1906
1
1
.8
-
Rawinsonde
precipitadle
water
1
1.2
cm
integrated vapor depth
Figure 2.9 Scatter plot o f rawinsonde and radiometer measured precipitable water vapor for DRI
(top) and USBR radiometer.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
33
The USBR radiometer was not operated in a scan mode during this storm, but Jacobson
et al., (1987) have shown that, for a radiometer similar to the USBR design, a thin film of water
due to melting snow on the reflector causes the recorded brightness temperatures to vary
sinusoidally. They calculated theoretically and measured a periodicity of tt associated with the
water layer. The errors introduced into the vapor values were much larger than errors in liquid
values.
In Fig. 2.8 values of vapor and liquid are shown for two periods when the DRI
radiometer conducted a 360° azimuth scan at a 25° elevation angle. At the time of the scan, the
precipitation rate was more than 2.5 mm/hr and the temperature was +1°C. The time of the
scans matches the time of the first suspected melting snow contribution to the USBR data in Fig.
2.4B. The vapor traces do not show any sinusoidal periodicity that resembles those reported by
Jacobson et al.,; hence contamination by melting snow on the DRI reflector did not appear to
be a problem.
2.5
DISCUSSION OF RESULTS AND CONCLUSIONS
In an earlier study, Heggli et al., (1986), compared the performance of two microwave
radiometers, the USBR radiometer and one from the NOAA Wave Propagation Laboratory.
They reported a 0.99 correlation coefficient, an average absolute difference of 0.02 mm and a
root mean square difference of 0.02 mm for the liquid water values measured by the two
instruments. Measurements less than 0.05 mm of liquid were not included in their comparison
and periods when melting occurred or the reflector was wet were removed from their data set
prior to analysis. The correlation coefficient they found was much higher than the 0.83 reported
«
here. Tne lower correlation (.83) found in the current study is thought to be the result of
contamination by liquid or melting particles and by differences in averaging intervals of the raw
data. The fact that values o f liquid water less than 0.05 mm are included in the present work
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
34
might also have contributed to the difference. If some points, suspected to be contaminated, are
removed from the USBR liquid data (points at 1640-1720, 0020-0040,2240-2300 and 2320-2340)
the correlation coefficient between DRI and USBR radiometers increases to 0.87.
Heggli et al. also reported a correlation of 0.79 for the vapor when liquid was present.
This compares to 0.68 in the present case. This lower correlation in vapor values for the current
study is obviously due to the outlying points in Fig. 2.7B, again thought to be caused by a wet
reflector or mylar window. The correlation without the outliers improves to 0.71. Note also that
for the time after 06 Dec. 1986, when there was little or no precipitation falling and temperatures
were well below zero, the liquid and vapor channel correlations were 0.92 and 0.87 respectively.
This shows that a much better correlation between the radiometers can be achieved during "good"
weather conditions.
Heggli et al. compared the USBR radiometer vapor measurements with precipitable
water vapor measured by rawinsonde at Kingvale, California. A correlation coefficient of 0.94
was reported. Data from rawinsonde launches at Kingvale, from 0000 November 20 to 0900
December 6 are compared to the DRI and USBR radiometer vapor values (30-min averages after
the rawinsonde launch) in Fig. 2.9. It shows a 0.97 correlation coefficient with the DRI vapor
data and a 0.80 correlation with USBR radiometer vapor.
The data show that the DRI
radiometer overestimated integrated vapor by about 0.1 cm on the average, when compared to
the rawinsonde values. It is important to note that rawinsonde precipitable water values are by
no means absolute. There are errors associated with the relative humidity measurement. Also,
for comparison with radiometers the drift of the instrument as it rises results in it sampling a
somewhat different volume of the atmosphere. Nevertheless, the two types of instruments
generally show a good agreement.
Prior to using two radiometers for detailed water budget studies at two different locations
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
35
on a mountain barrier, the instruments were compared when they were located side by side. The
storm period selected for comparison was not specificallt chosen for an inter-comparison
experiment. One radiometer (DRI’s) was, in fact, being field tested after its recent construction.
The data systems of the two instruments were not aligned for consistent averaging periods. In
addition the radiometers were operator-attended for only a part of the sampling time so
continuous monitoring of weather conditions was not conducted. In contrast to more carefully
designed inter-comparisons, the study reported here probably offers a more realistic assessment
of radiometer operation in weather conditions that often contaminate the brightness temperature
measurements and when data are collected with instruments unattended. In fact, the situation
described, where ambient temperatures at the site were often > 0°C during precipitation periods,
provided an opportunity to compare the effectiveness of the two radiometers in obtaining reliable
measurements under much less than optimum circumstances.
2.6 SUMMARY
We have compared data from two radiometers with similar receivers but different
reflector designs. One (DRTs) used a spinning reflector to remove any snow or water drops that
fall on it while the other (USBR’s) was fixed. The USBR-DRI radiometer temporal plot showed
a similar trend between vapor and liquid throughout the comparison period data except at points
of suspected contamination.
The correlation between liquid water depths was 0.83 which
increased to 0.87 after removing some of the suspected values. The vapor correlation coefficient
was 0.68 and increased slightly to 0.71 after removing some of the suspected contaminated points.
The standard deviation in the vapor values was almost twice as much as that of the liquid.
The radiometers were also compared to rawinsonde launches made at the same location.
Correlations coefficient of 0.80 and 0.97 were found for the USBR and DRI instruments,
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
36
respectively. This correlation coefficient of 0.80 is less than reported by Heggli et al., for the
same instrument (0.94) which could be due to the wet reflector problem.
It is important to point out once more that this study was not set up as a careful inter­
comparison of instruments as was done by Heggli et al. However, the data collected show that
improvements can be achieved by modest changes in the overall design of the radiometers.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
37
PART III
STABLE WATER ISOTOPES
3.1
INTRODUCTION
In this chapter, stable oxygen isotopes of water are used to explain observed differences
in ice crystal habits of snow that fell upwind and downwind of the Sierra Nevada crest. The
degree of Riming of the ice crystals, frontal structure and convective activity of the storm will be
explained in relation to the stable oxygen isotopic ratios of the precipitation.
The relationships of the stable oxygen isotopes of water to climatic change and
geographic factors have been studied since the early 1950’s.
Dansgaard (1953,64) reported
observations showing that the stable oxygen isotopic composition of precipitation was related to
mean annual temperature, altitude, latitude and precipitation amount. H e also showed that this
type of isotopic data could be used to determine when a synoptic front passed a sampling area.
The isotopic composition of precipitation from storms depends on a variety of factors such as
the temperature (Picciotto et al., 1960, Warburton and DeFelice, 1986) and vertical dimension
of the clouds from which the precipitation falls, the altitude of the station (Dansgaard, 1964;
Hodsworth et al., 1991: Smith e t al., 1979) and the phase (Dansgaard, 1964) and type
(Gedzelman, 1982, 1989: Demoz, 1989) of the precipitation.
Additionally in 1986, it was first
shown by Warburton and DeFelice that the stable oxygen isotopic composition of precipitation
is related to the microphysical processes within the clouds in which the precipitation has formed.
Fujiyoshi and Wakahama (1986), Demoz et al., (1991), and Sugunoto e t al., (1988,1989) have
also made recent studies of the isotopic composition of snowfall and its short term variability.
The stable isotopic composition of precipitation observed at a specific site depends on many
%
factors connected with the history of the air mass.
The trajectory of the mass defines the
moisture source and hence the initial isotopic composition (Rindsberger et al., 1990).
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
The
38
source of moisture for the western United States is primarily the northeastern Pacific Ocean, and
the storm systems during winter move eastward from the Pacific into the Sierra Nevada which
is the first major topographic barrier they encounter. Approximately 90% of the precipitation
from these storms in the Sierra, falls on the upwind side of this barrier (Smith et al.,1979). This
distribution depends on the direction of motion of the low pressure system from the Pacific. An
approach from the southwest brings warm, moist air, while an approach from the northwest
introduces cold somewhat drier air from the Gulf of Alaska. The Sierra Nevada extensively
modifies these flows particularly in the lower levels. Hence the distribution of the precipitation,
as well as its isotopic composition can be affected. Some of the topographic effects on the storms
may be very difficult to determine from synoptic or mesoscale analysis, particularly in rough
terrain like the Sierra Nevada mountains. Fortunately, some evidence of the meteorological
processes can be obtained from direct sampling and joint analysis of crystal habits and isotopic
content of the precipitation. In our analysis of a storm in the Sierra Nevada area, we have found
variations in crystal habit and isotopic content that help elucidate the meteorology by identifying
different periods of precipitation type within the storm.
As moist air moves eastward from the pacific, the 180 / 160 ratio of the precipitation that
falls out progressively decreases. The reason behind this preferential condensation of one species
over the other lies in the difference of the isotopic fractionation factor. The fractionation factor
(defined as the ratio of 180 / 160 in the liquid or solid phase to 180 / 1<s0 in the vapor phase) is
practically the same as the ratio of the corresponding vapor pressures of the species involved (in
this case lsO and 1£0 ).
Due to their lower vapor pressures, the liquid or solid phases, at
equilibrium, are enriched in the heavier isotope. As a result of the fractionation processes,
precipitation falling from a storm near the coast shows a higher percentage of lsO than
precipitation further inland. As the storm continues to move inland, the vapor is progressively
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
39
depleted of the heavier isotope and hence lower and lower values are found.
Several other factors control the stable isotopic ratio at a sampling station. Temperature
at which the precipitation particles form is very important since it affects the fractionation factor
(example; fractionation factor at -10°C and -25°C differ by -25 % o). Therefore, the colder the
temperature of precipitation formation the lower the isotopic ratio becomes. For the Sierra
Nevada, an empirical equation relating temperature and 5180 was found by Warburton and
Defelice to be; 5180 = 0 .9 T - 3.4. The type of precipitation has an additional effect, because the
fractionation factor for vapor/ice is larger than for vapor/water (example; fractionation factor for
vapor-to-ice is about 20% o larger than for vapor-to-water at 0°C ). Once snow flakes are formed,
their isotopic composition does not change while falling because the rate of isotopic exchange
between solid and vapor is very low, the diffusivity of lsO in ice being 1 0 '15 m2/sec at -10°C
according to Sugimoto et al. 1987. The SlsO content of the cloud water is conserved when
freezing to ice, also from similar considerations (see Demoz, 1989). Accretion of supercooled
cloud water on to snow flakes/crystals is the major process that can greatly modify the isotopic
content of ice crystals from formation to fall out. We will expand on this point further in the
next sections. The temperature changes and air mass differences that are encountered as
atmospheric fronts pass over a sampling station, also greatly influence the S180 as we will see
more clearly in the next section. The convective activity, amount of rime on the snow crystals,
cloud base and depth and elevation of the sampling site are all factors that affect the isotopic
ratios of the precipitation and are discussed in depth in the next section. An extended discussion
and related equations of stable oxygen isotope theory are given by Demoz (1 9 8 9 ).
A recent report by Heggli and Rauber (1988) describes the characteristics and evolution
of supercooled water in wintertime storms over the central Sierra Nevada.
One of the case
studies presented in this paper is one of the 63 storms discussed by these authors. It was
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
40
classified as a developing storm embedded in a strong westerly or south westerly flow and
occurred on 26-27 March, 1985. In this paper, we present a study of the stable oxygen isotopic
composition of ice-phase winter precipitation over the Sierra Nevada. The air masses present
in the storm are identified using the isotopic signatures of the precipitation reaching the ground.
A comparison of upwind and downwind measured ratios of stable isotopes of snowfall is
performed and used to explain the observation of larger proportions of colder origin snow
crystals at the downwind station. The effects of convective activity of the storm and riming on
the stable isotope content are also explained. To the authors best knowledge, the use of stable
isotopes in winter storm analysis over short time scales, as done here, have not been done before.
3.2. EXPERIMENTAL DESIGN AND INSTRUMENTATION.
The locations of the sites where snowfall was collected and of other equipment sites, are
shown in Figure 1.1.
The region above 2100 m elevation is hatched.
The snow sampling
occurred at two stations located on the upwind and downwind sides of the Sierra Nevada crest
during the winter of 1984-85. Kingvale (KGV) is 8 miles upwind of the crest and Hobart Mills
(HM, station S29 in Table 1.1) is about 8 miles downwind of the crest. KGV is at an elevation
of 1859 m MSL and HM is at 1824 m MSL.
All snow samples were collected time-sequentially, and to ensure sufficient sized samples for
analysis, were between 5 and 15 minutes duration depending on snowfall rate. Measurements
included precipitation rate, snow crystal habit and degrees of riming. Many micro-photographs
were made of the ice crystals during the snow sample collection periods.
Rawinsondes were launched from KGV and Sheridan (SHR), California at 3 hourly intervals.
A C-band radar was in operation at SHR. Data from these instruments, synoptic charts and
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
satellite maps were assembled as part of the Sierra Cooperative Pilot Project.
The complete
instrumentation network used in the program for the 1984-85 season is discussed by Reynolds
and Dennis (1986).
The supercooled liquid water referred to in this paper was derived from brightness
temperature measurements made with a vertically pointing dual-channel microwave radiometer
operating at KGV with receiver frequencies of 20.6 GHz and 31.65 GHz. The integrated liquid
water and water vapor depths through the atmosphere were computed as described in Part II.
The stable oxygen isotope analyses of the snow samples were made at the DRI laboratories.
Oxygen isotopic composition is represented as per mill (%o) deviation from a Standard Mean
Ocean Water (SMOW) value as defined by Craig (1961) and is denoted by SlsO. Measurement
precision for S180 is better than ±0.2 %o. The equation defining S180 is,
swo -
R i . , ~ , R* x 1000
where R^, and R,, are the isotopic ratios (180 / 160 ) of the sample and the standard respectively.
Demoz (1989) and DeFelice (1986) present a detailed discussion of measurements techniques
and standards.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
42
334
;940
Figure 3.1 500 mb height analysis (solid) and temperature (°C dashed) at 0000, 27 March 1985.
300(7
30 0
350,
400
450
500
550
600
650
f
650
700
750
800
850
900
950
1000
Figure 3.2 Temperature sounding at KGV, 1500, 26 March 1985.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
43
3.3
THE STORM OF 26-27 MARCH 1985.: - A CASE STUDY.
The main synoptic and mesoscale features of this storm have been presented in detail by
Reynolds and Kuciauskas (1988), Huggins et al. (1985) and Heggli and Rauber (1988). A brief
summary of the storm is given below. The reader is referred to the above reports for an
extensive and thorough discussion of the mesoscale structure of the storm.
The 500 mb chart, Figure 3.1, for this day showed a trough approaching the project area
and Fig. 3.2 shows a sounding taken at KGV around 1500 UTC on March 26. The precipitation
on this day occurred as a series of intense bands as seen by the radar reflectivity over KGV.
Figure 3.3 shows a plot of (A) the radar return over KGV, (B) the precipitation rate and (C) the
S180 record at KGV. Each horizontal step in B and C represents the length of time of collection
of that snow sample.
3.3.1
All times are in UTC.
RESULTS FROM KINGVALE.
The precipitation sampling at KGV started at 1600 on March 26 and ended at 0500 March
27. As shown in Figure 3.3, the S180 values were quite low at the beginning of the precipitation
period (-23.8 %o at 1700 on March 26).
In the same period, cold habit snow crystals were
predominant (see Table 3.2).
The 5lsO values became more positive with time for the next four hours and the precipitation
rate increased from 1 mm/hr to 2.5 mm/hr between 1700 and 1800 hours and remained steady
at that rate for the next 3 hours. During this same period a warm front lay across the area. It
was at its highest elevation (2.5 km) at 1600 hrs becoming progressively lower to 1.7 km at 2100
hrs.
The equivalent potential temperatures overalyed on the SlsO values are shown in Figure
3.4. Even though data is not available, it is expected that the cloud base during the warm frontal
passage period have decreased. The radar plots also show that echo top height was increasing.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
44
□
£
^
10-19 dBz
20-24 dBz
25-29 dBz
5-
X
£
E
lQ_
CL
0
-1 0
o
o
O
-1 5
-2 0
00
r»
<o
-2 5
-3 0
T
t— I— r
i
17
20
23
r
T
02
i— r
05
Time of Day (UTC)
Figure 3.3 Temporal analysis of the 26-27 March 1985 storm as it crossed the central Sierra
Nevada. Time-height radar reflectivity profile over KGV (A). Minimum detectable signal over
KGV is 8 dBz. Precipitation rate (B) and 5“0 measurements (C) at KGV.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
45
r 9
8
7
-1 0
6
e
E
5
*
1—
4
o>
S
CO
*
X
o
© -1 5
CO
3
306
-2 0
304
2
1
290
-2 5
17
20
23
Time (UTC)
Figure 3.4 Frontal positions and time-height cross section of equivalent potential temperature
(K) from 1600 U T 26 March 1985 to 0500 UT 27 March 1985 overlaid on 5lsO values for the
same duration.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
46
WIND DIRECTION
200
240
280
0 3 /2 6 /8 5
320
360
15 UT
4
cn
10
x
o
8
UJ
X
< — DIRECTION
6
4
- SPEED
2
0
10 20 30 40 50 60 70 80 90 100
WIND SPEED ( M /S )
B
Radiometer
Liquid (mm) & i.oo
Vapor (cm)
Liquid (mm) Solid, Vapor (cm) Dashed
______
' l ' -- V-1'
at Kingvale
' ' ---
0.8 0
|
1600
"I
-*—
r
2000
0000
0400
TIME (UTC)
Figure 3.5 Wind speed, wind direction (A) an d radiometer measured integrated cloud liquid
water and precipitable water vapor (B) for 26 March 1985.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
47
After 2130 hrs the 5lsO values decreased slowly again by 2 %o until approximately 0130 hrs 27
March. This four hour period we have chosen to refer to as a transition period between the
initial onset of the upper level cold front and the occurrence of the low surface level cold front
as shown in Figure 3.4. The small drop in the S180 value at 2115 may have signalled the onset
of the upper level cold front. The radiometer continued to record supercooled liquid water of
around 0.1 mm throughout this entire period (Fig. 3.5). The steady low values of integrated
liquid water depth suggest that convective activity was not producing measurable liquid water
prior to 0200 March 27.
Between 0200 and 0500 hrs 27 March, the SwO values showed a steady rise from -15 %o to -12
%o.
The liquid water values were slightly higher, convective cells had developed and the
precipitation rate decreased. After 0400 March 27, the radiometer measured an increase in
integrated liquid water depth with rapid changes indicative of convective and/or with orographic
lifting of the convectively unstable air mass behind the cold front.
3.3.2
RESULTS FROM HOBART MILLS.
The precipitation sampling at the down-wind station, HM, started at 1600 March 26 and
ended at 0200 March 27. Results are plotted in Figure 3.6. Several interesting features are
observed by comparing Figure 3.6 with Figure 3.3.
The lowest S180 value at KGV for this day occurred between 1700 and 1715 on March
26. At HM, which is about 16 miles east of KGV, the lowest Sl80 value was observed one hour
later between 1800 and 1815 March 26. Between 2100 and 2115, SlsO showed another relative
minimum at HM which also was observed at approximately the same time at KGV (see # in Fig.
3.3 and Fig. 3.6).
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
Loc: Hobart Mills
Date: 8 5 0 3 2 6 -2 7
Q-
-10
o
o
-1 5
-20
!<-WARM
<o
FRONT
PERIOD—
>
-2 5
-3 0
16
18
20
22
00
02
TIME (UTC)
Figure 3.6 Precipitation rate and S180 measurements at HM on 26-27 March 1985.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
The 5180 values at HM also became progressively more positive starting at 1800 and
ending at 2230 March 26 similar to the results for KGV during the warm frontal passage period.
The 5180 values began to decrease slightly again at 2230 at the time of the upper level trough
intrusion as it did at KGV. The S180 values at HM were lower than those measured at KGV for
the early phases of the storm from 1600 hrs to 2100 hrs after which the values are essentially the
same for the transition period referred to earlier. The physical explanation for this are given in
section 3.4. A statistical summary of 5180 values is presented in Table 3.1.
As indicated earlier, the surface cold front reached KGV after 0100 March 27 (see Fig.
3.4). Unfortunately, the short term detailed sampling of snowfall at HM stopped at 0000 March
27, the last sample being collected over a two hour period from 0000 March 26 to 0200 March
27. This did not allow signatures of the warm front-cold front transition structure to be clearly
detected at HM.
Precipitation rate measurements at HM show a more or less steady increase throughout
the sampling time. Neither radiometer derived liquid water, nor radar observations were made
at HM.
Table 3.1 Summary of SlsO (%o)measurements at KGV and HM.
DATE
850326
LOCATION
DATA
POINTS
MEAN
MIN.
MAX.
a s 18o
KGV
62
-16.8
-23.8
-11.8
12.0
HM
40
-21.7
-29.1
-15.0
14.0
22
4.9
5.3
3.2
2.0
DIFFEREN
CE
A5lsO = | MAX. - MIN. | : RANGE OF VARIATION AT A STATION
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
50
Table 3.2 Summary of snow crystal and riming observations at KGV and HM.
KINGVALE
HOBART MILLS
TIME
MAIN HABIT
light
1800-1930
plates, columns and
rosettes
needles, frozen drops and
some dendrites
light to
no rime
1930-2100
dendritic fragments,
needles and columns
2000-2130
columns and
columns with plates
rimed
2100-2230
columns and needles
2130-2300
needles, hexagonal plates
and combination
of needles
light to
moderate
2230-0000
needles, dendritic
fragments and rimed
plates
2300-0045
plates and
BBP
light to
moderate
0045-0200
needles and frozen drops
moderate to
heavy
0200-0300
small size needles
heavy at 0200
and light to
moderate after
TIME
MAIN HABIT
RIMING
1600-1700
dendrites
and BBP
light to
no rime
1700-1830
BBP and assemblage
of plates
1830-2000
RIMING
some
riming
moderate
to heavy
BBP stands for Broad Branched plates.
3.4
DISCUSSION
Dansgaard (1953) has shown that the SlsO values observed in precipitation are related
to the warm frontal surface height. The greater the height of the warm front surface from the
ground the more negative is the 5lsO of the precipitation. Gedzelman and Lawrence (1982)
presented this fact, but only in a schematic form. Figure 3.4 of this paper also supports this
conclusion during the warm frontal period until about 2200. Note that the effect of the elevated
front can be seen starting 2000 and was strong after 2200 as its influence on the 5lsO shows. But,
from the precipitation data it would seem that the warm frontal precipitation and its effect were
dominant until 2200.
For the 26-27 March case presented here , the 5,80 values observed in snowfall at the
high altitude site directly upwind of the Sierra Nevada crest, increased from -23.8%o at 1700 when
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
51
the warm frontal surface was about 2.5 km high to -13%o when this surface had lowered to 1.7 km.
The steady increase in S180 appeared to be interrupted briefly by significantly lower values around
2100 March 26, which coincided with the arrival of the upper level cold front. As indicated in
Figures 3.2,3.4 and 3.6, the isotopic composition of the precipitation which fell from the air mass
between 2200 March 26 and 0100 March 27 was different from that observed before and after
this time. The 5lsO decreased from -14%o to -18%o. Within this "transition" period, the narrow
warm sector between the warm and cold fronts, the air mass was influenced by the elevated front.
Heavy riming of crystals and graupel were observed during most of the transition region. The
correlation of 5180 and precipitation rate was positive in the warm frontal region (1700 to 2300),
which is can be explained by the decrease in cloud base associated with warm fronts (i.e. crystal
origins at wanner temperatures).
The 5lsO value at KGV increased sharply from -15% o at 0100 March 27 to -12% o at 0345
March 2 7 except between 0200 and 0215 when it showed a value of -21% o (Fig. 3.4). This
coincided with the passage of the cold frontal surface. The frontal surface extended from the
ground at 0145 to about 6 km high at 0345 March 27 and associated with it was a sharp transition
in echo top height during this period. During this time, the radar data indicated there was strong
convective activity. As a result, low level moisture (from warmer regions) was being transported
up to form the liquid water. This is one reason the 5lsO values sharply increased.
The effect of strong convection on cloud and precipitation microphysics and isotopic
composition of the snowfall was discussed by Demoz et al. (1991), Demoz (1989), Gedzelman
et al. (1982,1988) and Warburton and DeFelice (1986). Demoz et al. and Gedzelman et al.
reported that 5180 values are higher from convective than stratiform precipitation because the
moisture source, for convective systems, are lower level (warmer) air parcels. The 8lsO values
between 0345 and 0400 were the highest observed for the entire period of the storm and were
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
52
produced from this convective part of the cloud system.
The SI80 measurements at HM on the downwind side of the Sierra Nevada reflect all the
features associated with frontal structures seen on the windward side at KGV. The main events,
the warm frontal passage, the lowest 5180 and the onset of the passage of the elevated cold front
(indicated by # ), can be seen on the 5lsO plot for HM (Fig. 3.6). This suggests that the storm
structure was preserved as it crossed the Sierra Nevada. The time of the observation of the
lowest 5lsO value at KGV and HM differ by about one hour suggesting a speed of about 7 m/s
for the lower part of the storm to cover the distance between the stations (warm frontal motion).
Note, one must remember that a frontal surface is a simplified representation of a more diffuse
boundary between cold and warm air and we assume the motion of a cold front is primarily due
to horizontal advection rather than wave-like propagation.
The 5180 values at HM are lower than that o f KGV (Table 3.1). The difference in
elevation between these two stations is 35 m with KGV being at higher elevation. However, this
difference in elevation is too small to explain the observed differences in 5180 values between the
stations. In an earlier study, Smith et al. (1979) reported on the areal distribution of deuterium
in this region. Deuterium concentrations are about 10 times higher and linearly related to values
of 5180 . In their report, Smith et al. (1979) determined the rate of change of deuterium with
altitude for 11 storms. The smallest rate of change of deuterium for every 1000 meters was -20% o
and the highest was -45% o. For a storm with strong, zonal, westerly winds (similar to this case)
the value they reported was -28% o for every 1000 m. This would account for less than -l% o of the
difference in 5ieO observed for the 35 meters altitude difference between KGV and HM.
Table 3.1 shows a mean 5lsO at KGV (-16.8%o) which is 4.9%o more positive than at HM
(-21.7%o). Both the minimum and maximum values at KGV were also more positive than the
minimum and maximum values at HM by 5.3%o and 3.2%o respectively. The range of 5lsO variation
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
53
was 12% o at KGV and 14%o at HM, a difference of 2.0% o. These statistical data indicate that the
snow that fell at HM originated at higher, colder temperatures than the precipitation at KGV.
The minimum 5lsO values were observed near the beginning of the storm sampling time when
echo/cloud tops were about (or higher than) 5 km at KGV. This is also related to lower cloud
base that is associated with the passage of a warm front over KGV (or any station). The time
of observation for the minimum S180 value at HM was one hour later than at KGV. The lifting
of the front as it crosses the mountain barrier accompanied by increase in cloud base on the
down wind side may also contribute to the lower values observed at HM.
Additionally, as a storm moves eastward (inland from the coast), lsO concentration is
depleted further as more and more of the heavy isotope is removed in the precipitation. This
results in relatively lower S180 values. However, this general trend (i.e. lower 518with distance
from coast) is not always true as suggested by Smith et al. (1979). Even so, this effect would be
very small and is assumed negligible between KGV and HM.
Finally, Table 3.2 is a summary of the dominant habits and degree of riming observed at
KGV and HM stations. The data at KGV are divided into eight different somewhat distinct ice
crystal observation periods. These periods are also used to group the crystal observations made
at HM. The crystals with the coldest temperature of origin are observed at HM. Rosettes, which
are produced at temperatures of -20°C, or colder, were abundant prior to 2000 at HM while at
KGV, dendrites and broad branched plates were dominant. The echo top at KGV at this time
was about 6 km high and at -23°C. Since cloud top is probably higher than this, it aould appear
that the crystals that fell at HM at this time must have originated near this cloud top region.
Hence, since the temperature and elevation at which the snow crystals formed as well as the
temperature and elevation of the collection of rime deposits on the snow crystals (location of the
supercooled liquid water) appear to be the main factors that determine the value of S180 , it must
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
54
be that the precipitation that fell at HM captured their water substance from higher and colder
parts of the cloud system than at KGV. A similar result was reported from aircraft investigations
of winter storms over the Rocky mountains of Colorado by Uttal et al. (1988).
Precipitation at KGV and HM were also collected on 4 March, 1985. The precipitation
observations were made at the two sites simultaneously. Figure 3.7 shows the changes in
isotopicaly derived "temperatures" at the two sites.
The temperature (and hence the 8180
content) increases with time at both sites, but the values at HM are always colder (lower SlsO)
than those at KGV. These differences indicate that liquid water capture at the downwind site
was occurring higher in the cloud system than at the upwind site, i.e. at colder temperatures.
During the winters of 1984-85 and 1985-86, isotopic analyses were performed on 405 snow
samples collected at KGV.
The seasonal results are shown in Figures 3.8.
frequency of occurrence of 5lsO values (in steps of 2 %o).
They show the
The "temperature" scale on the
abscissa has assumed the linear relationship given by the relation
S180 = 0.9T -3.4. The results indicate that 85% of the snow analyzed captured most of its water
substance at temperatures warmer than -15°C, with a maximum around -11°C.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
55
ISOTOPE DERIVED
“ TEMPERATURE" (°C )
-3 0
-2 0
Downwind of Crest
-1 0
Upwind
o f Crest
1600
1700
1800
1900
2000
2100
TIME (UTC)
Figure 3.7 Isotope derived "temperature" according to the equation SlsO = .9T -3.4 of
precipitation that fell on sites KGV (dashed) and HM on 4 March 1985.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
56
EQUIVALENT TEMPERATURE (°C)
55
50
-2 4
•
•
i
i
I
«
—20
i
i
I
|
-1 5
I
l
l
-1 0
1
“i
-6
i— i—
j— i—
-2
i— i— r
1 9 8 5 /8 6
1 9 8 4 /8 5
45
ALL DATA AT KGV
40
35
0
1
30
g
an
25
^
20
15
10
r--J
5
0
-3 0
-2 5
-2 0
-1 5
-1 0
STABLE OXYGEN ISOTOPE RATIO ( 0 /
-5
)
Figure 3.8 Frequency analysis of the S“0 collected at KGV showing the equivalent temperature
for 1984/85 and 1985/86 seasons.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
57
3 .5
SU M M A RY AND C O N C L U SIO N
Data collected from a storm of type A l (i.e. developing storm embedded in a strong
westerly flow) that passed the Central Sierra Nevada on 26-27 March 1985, are presented here.
Sampling of the precipitation that fell in the form of snow was. done at two stations, of almost
the same elevation, located on the upwind and downwind sides of the mountain barrier. The data
show that the S180 values from the precipitation are closely related to the structure of the frontal
system and its modification by the mountain barrier. The large, short-term temporal variations
of the S180 values at both stations have been discussed in relation to radar echo top data, frontal
height and ice crystal types of the snow which fell. The following is a summary of the findings
of this analysis.
First, it was possible to separate the 12 hour observing period at KGV into three
somewhat separate mesoscale structure periods. There was a warm frontal period followed by
a period affected by an upper level front, referred as "transition period”, then a period affected
by the cold front.
Second, the £lsO values showed different characteristics within the three periods. During
the warm frontal influence, the 5180 showed the largest change. From a minimum value of -24% o
(-29% o at HM), the values increased to about -14% o in about five hours (2%o per hour). Between
2000 and 0100, the "transition period" when the elevated front was observed, the 5180 values
decreased by about 6% o from -14% o to -20% o (1 .2 %oper hour). The absence of convective activity
during most o f these two periods suggests the primary contribution to changing the 518values was
from the frontal structure. Starting about 0100 and after, mainly the cold frontal period, the 5lsO
values show again a steady increase from -21.5% o to -12% o (4.5% o per hour). Recall that the storm
high showed convective activity and lower cloud tops during this cold frontal period. Hence the
increase in 5180 was a combination of all these factors. Note also that the highest rate of SlsO
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
58
change occurred during the cold frontal period, 4.5%o per hour, and the lowest during the
transition period, 1.2%o per hour. A similar trend of the 5180 values was observed at the site
downwind of the crest, HM. These repeated characteristics, mainly in the warm frontal period,
of the 5180 suggests that the mesoscale storm structure for this period was mostly preserved as
it crossed the mountain barrier.
Third, the S180 values for the snow that fell at KGV and HM during the same time
intervals differed by about 5%o, that at HM being more negative.
This suggests that the
precipitation that reached HM had a larger component of ice from colder regions of the cloud
mass (or a smaller component from the warmer regions). As shown in Table 3.2, this is
supported by ground ice crystal observations at the two sites. Note that the crystals at HM tend
to be of colder origin than those falling at KGV at corresponding times. This might be explained
if one considers the location where the ice crystals that fell in the windward site originated.
Crystals that originate higher in the cloud travel relatively further horizontally than those
originating at lower levels. Also, because it has been found (Reynolds and Kuciauskas, 1988 and
others) that most of the liquid water in clouds over this region in winter is located in the lower,
warmer levels of the clouds, the overall growth rates due to riming, diffusion and aggregation of
precipitation particles to precipitable sizes are faster at lower altitudes than growth rates at higher
altitudes. This is especially true on the upwind side of the barrier (see Huggins and Rodi, 1985).
It also leads to longer residence times in the clouds for higher elevation of origin crystals and
probably greater horizontal transport. It is therefore expected that a larger percentage of the
precipitation in the form of snow comes from higher cloud levels at the lee station which, in turn,
would lead to lower S180 values at HM.
Finally, the 8180 values could be greatly modified if riming or highly convective instability
exist. The degree of riming, ice crystal origin, convective instability and the location of the
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
59
supercooled liquid water are all important factors to consider when interpreting the short term
5180 values at a given location. In this work, an attempt have been made to show how the 5180
values at a station in the central Sierra Nevada are related to the frontal structure of the winter
storms. The short term variations of 8lsO in a station not only are caused by the mesoscale
structure and geographic location of the station, but also by the microscale processes above the
sampling site. This work has demonstrated the use of SwO measurements from upwind and
downwind station in studies of the frontal structure of the storm in a mountain area.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
60
PA R T IV
STUDY O F W IN T E R ST O R M S
U PW IN D AND D O W NW IND O F T H E C R E ST O F T H E SIE R R A NEVADA
4.1
IN T R O D U C T IO N
The influence of mountains on precipitation enhancement, mainly on the upwind side of
a mountain barrier, is well documented in the literature. The reviews by Smith (1979) and
Cotton and Anthes (1989) present a good discussion of the theoretical and numerical side of the
problem as well as a good list of references. This theory of smooth adiabatic lifting due to
orography was the main idea in the conceptual model put forward by Ludlam (1955) for purposes
of cloud seeding. Ludlam’s hypothesis and the discovery of artificial nucleants to increase ice
crystal concentration in supercooled liquid water (SLW) clouds by Schaefer (1946) and Vonnegut
(1947) served as a basis for many cloud seeding operations since the early 1960’s. A review of
these programs is presented by Elliot (1986) and Reynolds and Dennis (1986). One such project
was the Sierra Cooperative Pilot Project (SCPP) which was reviewed thoroughly by Reynolds and
Dennis (1986).
During the 1986/87 season, an extensive field investigation of the winter storms that
affected the central Sierra Nevada mountains was done as part of SCPP. The data was
documented by Hemmer et al., (1987). Several investigators have reported on the evolution of
SLW and mesoscale characteristics of the storms affecting the Sierra Nevada. These include
Reynolds and Kuciauskas, 1988, Reynolds and Dennis, 1986, Huggins and Rodi, 1985, Huggins
et al., 1989, Heggli and Rauber, 1988, Heggli et al., 1983, Heggli and Reynolds 1985, Marwitz,
1980,1985, 1987 and the references within. Most of the above studies were concerned mainly
on the characteristics of winter storms upwind of the main mountain crest of the Sierra Nevada.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
61
Huggins et al., (1990) addressed the downwind side by presenting for the first time a comparison
of radiometric measurements of cloud liquid water made at Kingvale (KGV) and Truckee (TRK).
Kingvale is located slightly west of the Sierra Nevada crest and TRK is about 8 miles east
(downwind) in a broad valley (see Fig. 1.1 in Part I). In their study, Huggins et al., did not
present the microphysics of the storms.
The amount of SLW over a station is a result of several liquid water production and
depletion processes. It is primarily produced by cooling of the air as it rises (adiabatic ascent).
In this study area, the air is lifted over the Sierra Nevada and the vapor condenses to form liquid
water and ice.
The relative amounts of the liquid and ice are controlled by the rate of
depositional growth, accretion and cloud droplet coalescence competing with the rate of
condensation of vapor to water droplets, entrainment of dry air and removal by precipitation.
In the lee side of the Sierra, evaporation and sublimation occur. The radiometer then measures
SLW when the production is greater than the depletion processes.
The SLW is removed as
precipitation by riming processes and converted to ice mass. The removal depends on a number
of meteorological factors.
The compressional warming associated with the downslope winds and the wind strength
itself is highly related to how much precipitation fell out of the storm upwind. Durran and
Klemp (1983) and Klemp and Wilhemson (1978a&b) modeled a severe downslope storm over
Boulder, Colorado. They found that the addition of moisture decreased the downslope wind
from 45 to 25 m/s. They also found the lee side warming, often attributed to the release of latent
heat on the wind ward side and dry adiabatic descent on the lee side, to be cooler by several
degrees for precipitating clouds. They concluded that the most important factor for the warming
is the amplitude of the mountain wave which is larger in the dry case. Since all the case studies
to be presented are precipitating clouds, the warming (and hence the SLW evaporation) involved
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
62
may not be very large.
The development stage, as well as the mesoscale structure involved, is also an important
factor in comparing and analyzing the storms. Heggli and Rauber (1988) classified 63 storms that
affected the central Sierra Nevada into two main and five subdivisions to look at the SLW upwind
of the crest. The case studies presented below are chosen from these groups of storms analyzed
by Heggli and Rauber. A detailed analysis of the storm on December 18,1986 and brief account
of storms on December 19, December 22, and January 3 as case studies are presented.
This chapter of the dissertation attempts to develop a more complete picture of these
particular storms through a study of the SLW and the precipitation microphysics. First, the SLW
measurements at KGV and TRK are compared and discussed including the temporal evolution
of the SLW, frequency of occurence and relative changes between KGV and TRK in the different
storms. This is also used to identify which storms had efficient precipitation forming processes.
To aid in this analysis, surface wind measurements at Slide Mountain (SM) weather station and
precipitation amounts across the barrier for each storm are presented. Secondly, an analysis of
the ground observed snow crystals is used to learn more about the relative importance and
contributions of riming, fragmentation and aggregation into the snowpack mainly on the
downwind side. The same analysis is also used to learn more about ice forming mechanisms in
the winter clouds on the downwind side of the Sierra.
The findings of this and the previous stable isotopes section, are intended to aid in
developing a more complete conceptual picture of the main regions of interest in mature winter
orographic clouds over the Sierra Nevada.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
63
4.2
A CASE STUDY, 18-19 DECEMBER 1986: - A CASE STUDY.
4.2.1
SYNOPTIC STRUCTURE AND WINDS.
A short wave approached the west coast on 18 December 1986. At 1200 UTC, the low
pressure was located off Washington state coast while a high pressure was over Idaho. The
surface cold front at 1200 was positioned as shown in Fig. 4.2.1. The relative location of the high
and low pressure centers (Fig. 4.2.1) caused the flow to split over the central Sierra Nevada.
Heggli and Rauber (1988) classified this storm as predominantly zonal with split flow in
the middle troposphere associated with a dissipating storm. This classification can be clearly
understood from the series of 500 mb plots presented in the Fig. A4.1 appendix. The plots are
a 24 hour interval analysis of the 500 mb height contours showing the slow movement of the low
eastward, the split flow and the dissipating nature of the storm. Notice how the low center on
the west coast moves east slowly and weakens as the ridge intensifies and moves northward.
As the low center moved east, the wind direction over the sampling area was mostly from
the south west ( about 240 degrees) and wind speed showed a marked increase (25 kt was the
maximum recorded speed). The direction became predominantly from 250 degree (at SM) when
the low center was over Washington state and stayed the same even after the frontal passage.
The front passed Lincoln (LNC), California after 1800 December 18, Kingvale (KGV), California
about 0000 December 19 and Slide Mountain weather station (SM) between 0100 and 0300 on
19 December. The slow and dissipating nature of the front did not allow the observation of a
marked change in temperature values but could be identified from precipitation rate
measurements in Fig. 4.2.11. This slow eastward movement was due, mainly to the ridge located
ahead (Fig. 4.2.1) causing the flow to split at the experimental site, and the topography of the
Sierra Nevada mountains.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
64
1024
1026
1016
Figure 4.2.1 Surface weather map (top) and 500 mb (bottom) layer heights at 1200 UTC, December
18,1986.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
SLIDE MOUNTAIN WEATHER STATION
A d issipating, sp lit (low s to rm over th e S ie rra Nevada.
SUDE MT.
1986
18 DEC.
DIR.
<—
II— > 19 DEC.
360
NO DATA
SPEED
->
40
< -
270
t/i
o
4.7*
CALM
z
jj I
o
UJ
45X
180
LU
r,
90
04
TIME
= t= J Z X = H ]
1-5
4-4
(0 • )
( 2 ■ ) <11 • ) (41 a ) (1 5 ■ ) (2 5 • )
18/19 DECEMBER 1986
1030/22 TO 1030/23
7-10 11-15 17-21 27-95
WIND SPEED SCALE (KNOTS)
NOTE - WIND DIRECTION IS THE
HRECTION WIND IS SLOWING FROU
1 knot = 0.447 m/s
Figure 4.2.2 Measurements of wind speed and direction at Slide Mountain weather station (A)
and wind rose analysis (B) for December 18,1986.
P'
Ln
66
The wind measurements at SM are summarized in Fig. 4.2.2. Slide mountain weather
station is located east of TRK on a mountain peak at 2955 m MSL. Part A of the Figure shows
a temporal analysis and part B is the statistical summary for the wind speed and direction at SM
weather station. Starting 0000 December 19 ( when the front was over KGV), the data shows,
a change in wind speed to lower values (see Fig. 4.2.3). This change could have been caused by
the upper part of the front. That is, even though the surface front was west of SM, the
interaction of the physical barrier with the front causes the upper part of the frontal surface to
be ahead (east) of the lower level (Smith, 1982). As a result, the wind speed changes at SM
showed frontal signs for an extended time before the surface front reached there.
The statistical summary of the wind measurements at SM shows 43% of the time the
wind was blowing from the west-south west, 23% from south west (about 225°) and 16% from
west (270°). The remaining 18% had a direction south of south west (160-225°). The highest
speed recorded was 25 Knots (11.175 m/sec) and 41% of the time the speed was between 11-16
Knots (5-7 m/s). The cross barrier nature of the wind (=* 250° for 43% of the time) and the
relatively strong wind speeds is a good precondition for orographic lifting and liquid water
development.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
67
18 December
18 December
15
i „
12
i
09
i
06
i
03
i_____ ...
00
21
18
15
Figure 4.23. Time versus height cross section from KGV soundings; (A) Relative humidity
(solid), temperature (dashed) and (B) equivalent potential temperature contours. A full wind
barb is 10 m/s. Shaded areas are where 30Jdz < 0. Time increases from right to left.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
68
4.2.2
RADIOMETER MEASURED LIQUID WATER.
In Figure 4.2.4 (a and b), temporal plots of 15 minute averages of integrated cloud liquid
water and vapor depths measured by the USBR and DRI dual channel radiometers at KGV (A)
and TRK (B) respectively are shown. The "Z" and "S" on top of the plots stand for zenith and
scan respectively. All the liquid water at KGV and TRK was present at temperatures below 0°C
and hence supercooled. In terms of liquid water concentration, using one gram of liquid water
to be equivalent to 1 cm3, a one km thick cloud with uniform liquid content of 0.1 g/m3yields an
integrated liquid water depth value of 0.1 mm.
The radiometer observations show SLW values as early as 1000 Dec. 18 about 12 hrs
before frontal passage. The SLW values were generally above 0.1 mm at KGV while at TRK the
values were greater than 0.1 mm only after 1600. The SLW values show a sharp decrease at 2300
Dec. 19 at both stations, KGV and TRK, near the time the cold front passed KGV. As most of
the storms over this area show (Reynolds and Kuciauskas, 1988; Heggli and Rauber, 1987), SLW
('
is greatest in the leading edge of the front where precipitation tends to be minimum (and vertical
velocity would be greatest) and behind the cold front where cloud tops subside and convective
instability is relatively large. This was the case with the SLW being the greatest before 0000 Dec.
19 and after 0800, Dec. 19. Between 0000 and 0800, Dec. 19, the SLW was generally less than
0.2 mm at KGV. SLW was present throughout the storm duration with a peak value of 0.75 mm
at KGV and 0.4 mm at TRK. The continued presence of SLW and low precipitation rate from
the shallow clouds, were good indicators of inefficient precipitation processes. The maximum
value of SLW was recorded in pre-frontal clouds both at KGV and TRK. The radar data and
cloud structure in the upwind side are discussed in detail by Deshler et al. (1990).
The SLW values at KGV were consistently higher than those at TRK. This should be
expected because of the effect of topography, i.e. the condensation supply rate at KGV is higher
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
69
due to its location near the crest on the upwind side while the air over TRK is generally
subsiding. Evaporation due to compressional heating over TRK of the SLW was manifested in
larger vapor values and lower liquid water values (see Fig. 4.2.4).
In Figure 4.2.4B, the DRI radiometer operated in a scan mode between 1700 and 2320.
Within this period of time IS independent scans were made at an elevation angle of 15°. These
scans and an average of the measurements at each angle for all the scans is given in Fig. A3 in
the appendix. At IS degree elevation, the scan path is well above the highest peak in the barrier.
From the scans, it can be seen that the maximum values of the SLW (longest arrows) shift from
scan to scan as the clouds moved eastward. This observed change in magnitude with time, for
all the vectors in each scan, suggests that there was no contamination from a fixed target. The
largest vectors tend to be in an orientation of about 330° and 120°. The averaged maximum value
of SLW was recorded at 120° and 180°. The effect o f Lake Tahoe, located between 120° and
180°, on the SLW is not known. From the average plot SLW was minimum between 120° and
180°. In each of the scans made, SLW north of TRK was always minimum. This was above the
valley north of TRK (shown in Fig. 1.1 in Part I) probably due to evaporation as the air sank and
warmed by compression.
Figure 4.2.S is an azimuth/time contour plot of SLW of all the scans made. At the
beginning, the data show a value of 0.S mm at 200° and 310° which is the general location of the
Sierra Nevada barrier. The maximum value at 200° started shifting to 120° at 1900. It coincided
with the start of the steady increase in wind speed at SM from 10 knots to a peak value of 25
knots by 2000. The second maxima at 310° also shifted to 330°. The average wind direction at
SM was from 250°. The shift toward the east of south east and north of the SLW maxima was
mostly due to the effect of the high pressure center and 250° wind. This was a result of the split
nature of the flow. The scan measurements at TRK were all made in pre-frontal clouds.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
861218/19:1030-1030
1030 (GMT) 18 DEC 1986 to 1030 (GMT) 19 DEC 1986
f-2
- S
1.00
1.20
microphysics sampling
Liquid
V apor
0.80
a<-*>
0.80 -1
<
0.60
9
O
o<0
•v
£ 0.40
0.40
3 0.20
0.00
10
12
14
16
18
20
22
0
TIME (GMT)
o
3
0.00
2
4
6
8
10
861218/19:1030-1030
1030 (GMT) 18 DEC 1986 to 1030 (GMT) 19 DEC 1986
rz
.lip n , nnn i—11—i il| 11
0.50
Liquid
microphysics sampling
-s
-1.50
V apor
0.40
*
9
1.00
<
0.30
9
o
©
9
•V
£ 0.20
0.50
o*
a o .io
o.oo
9
3
o.oo
TIME (GMT)
Figure 4.2.4. Radiometer measured integrated cloud liquid water and precipitable water
at KGV (A) and TRK (B) for December 18,1986.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
TIME/AZIMUTH PLOT FOR DEC. 18, 1986
1659 (gm t) 18 dec 1986 to 2319 (gm t) 18 dec 1986
Threshold =0.01 m m
Contour Interval=0.05 m m
360r320^ -v 2 8 0 -
toX)
^ 240-
200
^
-
160-
.5 120
-
N
<
8040f e
17
18
19
20
21
22
23
TIME (g m t)
Figure 4.2.5. Spatial distribution of radiometer measured integrated cloud liquid water at TRK
on December 18,1986.
0
72
STATIONS AROUND TRUCKEE
DECEMBER 18, 1986
8945
8225
1890
a
2195
»»*&Z Z Z v
PRECIP. (MM*100)
TOTAL PRECIPITATION
DECEMBER 18, 1986
B
2945
19*5 1890
^
1783 1691
l
1
T
I
NW6 S99 S13 S l l
I
S21
I
I
SB* S86
I
I
I
1573
I
"I"
S29 S33 SN7 SN9 SND
SITE ID
PRECIP. (UMMOO)
^ 1 SITE ELE. (MSL)
Figure 4.2.6. Total precipitation (mm) accumlated on December 18, 1986 at stations around TRK
(A) and across the barrier (B).
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
73
The clouds around 2300 were all of frontal band. The maximum SLW recorded west of
TRK was due to clouds over the crest unaffected by compressional warming and evaporation.
As the cloud band moved east, compressional warming and evaporation as well as the absence
of any orographic lift because of the flat terrain North of TRK resulted in a persistent SLW
minimum. A re-development of SLW over the Carson mountains on the east side of TRK was
hot that large, possibly due to the wind flowing nearly parallel to the Carson range mountains.
It could also be due to undetected SLW, from shallow clouds, below the radiometer beam. But
the amount of precipitation was very low which is another indicator of small redevelopment.
Figure 4.2.6A shows total precipitation recorded at stations around TRK. Stations SN1
and SNH are north of TRK on a flat valley while S84 and SN7 are located on west and east of
TRK respectively. It shows very little precipitation values north of TRK due to the evaporation
associated with the topography.
To the south, S86, SN4 and SN5 show a decrease in
precipitation but not as much as stations to the north. In Fig. 4.2.6B the total precipitation for
this day is presented for a number of stations, starting from station NW6 located about 40 km
west, to station SND approximately 40 km east of the crest. The station elevation in meters and
total precipitation in millimeters is indicated in the graph for each station. It shows a very slow
descent in elevation has a very large effect in decreasing the precipitation amount. It is also clear
from the figure, re-development of clouds by the orographic lift due to the Carson Range
mountains (station SN7 and SN9) did not produce large precipitation values as stations before
the main crest (for same duration of time) which are even at lower elevations. This suggests that
smooth orographic lifting starting well ahead of the mountain barrier may be the main reason
for large precipitation values observed upstream. The absence o f such lift causes a decrease in
precipitation over stations SN7 and SN9. Marwitz(1980) suggested that low level blocking of
winds upstream acts to decrease the effective height of the Sierra Nevada barrier and decreases
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
74
the orographic precipitation. But Grossman and Durran (1984) reported the effect to be one of
positive contribution to the precipitation well ahead of the crest. More recently, Peterson et al.
(1990) found that blocking enhances the precipitation through the entire barrier. In this case
study, the sharp decrease in precipitation slightly east of the main crest, the relatively low change
in precipitation west of the main crest, the not so-large redevelopment of precipitation over the
Carson Range seems to support the hypothesis put forward by Peterson et al.,(1990). Note that,
even though we point out this from precipitation results only, the various theories need to be
addressed carefully with analysis of the flow fields, stability and other cloud parameters.
A frequency analysis of the vapor and liquid values measured at KGV and TRK is
presented in Fig. 4.2.7. The data used for the frequency analysis is the 15 minute average plotted
in Fig. 4.2.4. There were 61 samples at TRK and 96 at KGV. The histograms show, for SLW
at KGV, 20% of the time the value was between 0.16 and 0.21 mm and values as high as 0.8 mm
were observed. For TRK, values were between 0.16 and 0.21 mm only 4% of the time and the
maximum was 0.41 mm. Note also the sharp decrease of the distribution curve at TRK versus
that at KGV. When comparing the vapor values, larger maximum values were measured at TRK
than at KGV. The vapor was 0.8 about 19% of the time at KGV as compared to 8% at TRK.
The value at TRK was 0.95 mm 13% o f the time. Again the distribution (spread of the data) at
TRK was larger than at KGV, contrary to the SLW. A summary of the data at TRK and KGV
is given in Table 4.2.1.
The decrease in SLW and increase in vapor at TRK is consistent with evaporation of
cloud liquid water or sublimation o f ice in the descending air. As mentioned above, the crystals
that fell on the lee side were rimed. This contributes another removal mechanism for SLW and
hence further lowers the liquid water value on the lee side since condensation is hot occurring.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Dec. 18, 1986 at Kingvale
Dec. 18, 1986 at Truckee
LIQUID
Dec. 16, 1966 at Kingvale
-I_____ i_____ L
20
-
LIQUID
Dec. 16, 1966 at Truckee
16 -
VAPOR
VAPOR
12
Q
3)
U
C
Qi
L.
U.
“.4i"
1
I
Figure 4.2.7. Frequency analysis of the radiometer measured integrated cloud liquid (mm) and
precipitable water vapor (cm) at TRK and KGV on December 18, 1986.
i
1.4
76
Table 4.2.1 Summary of radiometer measurements on 861218.
VAPOR
LIQUID
SITE
4.2.3
MEAN
MAXIMUM
MAXIMUM
MEAN
KGV
0.23
0.75
0.80
1.02
TRK
0.10
0.41
0.91
1.33
PRECIPITATION MICROPHYSICS.
Snow crystals were collected at TD on a 2.5 cm diameter petri dish exposed horizontally
from 5 to 45 seconds depending on snow fall intensity. Photographs were taken of the crystals
and their melted drops to provide mass calculations of each snowflake (Mitchell et al., 1989)..
The crystal habits were classified according to Mogono and Lee (1966). The classification chart
is given in the appendix.
This storm was seeded from aircraft with Agl between 1841-1857. The location of the
ground microphysics observation point (KGV) was the target of the seeding experiment. Deshler
et al. (1990) presented a detailed analysis of the effects of seeding. They were unable to
determine "for certain" seeding effects from ground measured observations. Also, the seeding
window (about 16 minutes) is small compared to the storm duration. Note that precipitation
rate, SLW and even the snow crystal observations are made on 15-minute averages. Hence, all
conclusions and discussions made here assumed natural processes.
Ground observations of ice crystals at KGV and TD are presented below.
A)
RESULTS FROM TAHOE DONNER.
Table 4.2.2 is a summary of observation of the main ice crystal habit types, riming and
aggregation available for this case study. Sampling at this station was done from 1825 to 2100.
The clouds at this time were pre-frontal with SLW amounts between 0.2 mm and 0.4 mm. Many
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
hexagonal plates and few needles fell before 1843 and were unrimed. Between 1843 and 1920
the crystal habits changed into few stellars, needles and mostly rimed dendrites. The SLW
showed a peak at this time (Fig. 4.2.4). Dendritic crystals were dominant from 1920 to the end
of the sampling time with some plate, column and graupel also observed. The degree of riming
progressively increased from unrimed to heavy, and the snowfall rate was light during the
sampling time. Most of the riming was observed on the dendritic crystals and the R3b’s (graupellike snow of lump type) were of dendritic base. Aggregation was observed throughout the
sampling time with "few" up to 1940 and after 2030, "some" around 1950 and 2010 to 2020 and
"many" around 2000. Surface temperature changed from -1.8°C to 0°C, with -0.8°C recorded at
1825 and 0°C at 1950 when many large dendritic aggregates of light to moderate riming were
falling.
B)
RESULTS FROM KINGVALE.
Ice crystal sampling at KGV started at 1700 and ended at 2330 (started an hour before
and stayed two hours after the sampling stopped at TD). This is presented in a summary form
in Table 4.2.3. Pre-ffontal clouds with high SLW (up to 0.75 mm) were observed at this time
(Fig. 4.2.4). The Table shows the crystal habit type, riming and aggregation information for every
15 minute interval. For example, at 1700 to 1715, light to moderately rimed needles, heavily
rimed columns, moderate to heavy rimed Broad Branched Plates (BBP’s), light to heavy rimed
dendritic crystals and lump graupel fell. It is evident that dendrites, needles and long hollow
c o lu m n s were present from 1700 to 2000 with riming of the columns and dendritic crystals
progressively increasing. The degree of riming for needles decreased from light to moderate to
light to no rime. The BBP’s were mostly heavily rimed. From 1800 to 2000, aggregation was
observed mostly of dendritic crystals and BBP’s with the latter falling mainly between 1800 and
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
78
1830. Information on how many crystals were aggregates was not available. Ground temperature
at KGV was constant near 0°C throughout the sampling time.
Table 4.2.2 Ground microphysics summary for the storm of 18 December 1986 at Tahoe Donner.
TIM
E
N la
N le
pld
pie
cle
R3b
1825
1834
uM
RIMI
NG
AGG.
COMMENTS
U
Ap
T = -0.8°c, light
snow
u
Af
light snow
light snow
1843
1855
V
V
V
V
M
AF
1905
V
V
V
V
M
AF
VM
M
AF
M-H
AF
1920
y/
1930
1940'
V
1950'
V
V
L-M
very light snow
light snow
L-M
As
T=0.0°C
2000*
V
L-M
a m
large aggregates
2010
V
L-M
As
very light snow
2020
L-M
As
light to moderate
2030
L-M
Af
very light snow
2040
V
L-MH
stopped snowing
' Seeding window at KGV; M- many, F - few and s - some aggregation; "A" - aggregation.
M - Moderate rime, L • light rime and U - unrimed; V indicates crystal was observed.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
Table 4.2.3 Ground microphysics summary for the storm of 18 December 1986 at Kingvale.
TIME
N la
Clf
LM
Pla
Pic
MH
Pld
Pie
R4b
LH
1715
LM
LM
1745
LM
LM
1815
LM
LM
LM
MH
MH'
MH
LM
MH
LM
1845
1915
LM'
1945
LM
LH
LH'
2015
UL
LH
MH
2045
2115
UM'
UL
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
80
4.2.4 DISCUSSION OF RESULS.
The synoptic analysis of the storm showed that the front was moving very slowly (about 12
hours to move from the coast to KGV) and weakening. The crystals at KGV (collected between 1700
and 2000) fell from pre-cold-frontal clouds.
In Table 4.2.3 a decrease in degree of riming from light-to-moderate to light was observed
for the needles, contrary to the trend for other crystal types. This suggests that the needles originated
at low altitudes, probably lower than the bulk of the supercooled liquid water (SLW). From the
crystal habits given in Table 4.2.3, temperature of origin varied from -3°C for the needles to -17°C for
the dendrites. This temperature range was located below 5 km and relative humidity over KGV from
ground up to this level was above 70%. The temperature at 4 km at 1800 was -1S°C.
The graupel observed throughout nearly the entire sampling period indicates the presence of
high SLW and convective activity; both of which were confirmed by radiometer measurement and
soundings (discussed in the next section). Warm habit crystal dominance throughout the sampling
time implied that cloud top temperatures were warmer than -20 °C (Ludlam, 1955). This was
confirmed from 1 minute averaged data of the K^-band radar given in Fig.4.2.8 and soundings at 1800
(Fig. 4.2.1). The Ka-band radar echo top versus time shows a similar picture of the cloud as that
shown (see Fig. 4.2.9) by Deshler et al. (1990). At 1800, echo top height at KGV was about 3.5 km
(1.5 km AGL) and varied from 0.5 - 4.5 km AGL over the next 3 hrs. Echo top height around 2000
was highest suggesting the clouds on the far left side of the Figure 4.2.9 at 1700 were over KGV at
this time.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
6000
KV 8 6 1 2 1 8
50 0 0
o
<
E
4000
CL
1o— 3 0 0 0
o
x
o
UJ
o
<
QC
2000
1000
2100
2000
1900
1800
1700
TIME UT
Figure 4.2.8. K,-band radar echo top height versus time at KGV on December 18,1986.
-I ' 1l
I" 1i
IB 0EC 1986
I700UTC
□ 0 -9
■ 10-14
□ 15-19
-»o-c \ V ' ’
w
10
20
30 -
40
30
60
70
SO
30
100
110
120
130
140
DISTANCE FROM SHERIDAN RADAR (Km)
Figure 4.2.9. Cross section of the clouds over the barrier showing radar echo and temperature
structure on December 18,1986 at 1700 UTC (from Deshler et al, 1990).
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
82
In these pre-cold frontal clouds, the crystals observed were needles, hollow columns, BBP’s
and dendritic crystals with very few hexagonal plates and stellars. Aircraft measurements reported by
Deshler et al., over KGV showed small ice particles (0.2-0.3 mm) and few large dendrites between
1635-1730. Aggregation at ground station at this time was not observed. Cloud top temperature was
about -10°c over KGV and clouds were shallow. By 1815, clouds became deeper and dendritic
aggregates and irregular shapes were observed. Regions near cloud top were the primary source for
these dendritic crystals. Only dendritic crystals and some BBP’s aggregated. It is also clear that when
lightly rimed dendrites fell, (or in the absence of dendrites), crystal aggregation was absent. This
observation is in line with the hypothesis that aggregation requires relativefy large tentacled crystal
structure falling through a cloud at a favorable temperature (greater than -4°C, according to Furukawa
et al., 1987). Also the "filter effect" of the dendritic branches as discussed by Lew et al.,(1986)
increases the collection efficiency of dendrites. Dendritic crystals originate at -12°C to -17°C, they
aggregate way below their embryonic origin which by that time they had become large enough to
aggregate effectively.
A trend of increased fragmentation and decrease in aggregation with increased riming was
observed (see Fig. 4.2.10B below). The correlation of riming and fragmentation could be the result
of collision between the ice hydrometeors of different velocity (Vardiman, 1978) that resulted from
the difference in riming. The negative correlation between riming and aggregation could be that as
the dendritic crystals get rimed, hence filling the space between the branches, the "filter effect" is
reduced. This reduction in collection efficiency leads to a decrease in aggregated crystals. Dendritic
aggregates of size 4-6 mm were observed.
The crystals that fell at TD were very similar to that of KGV with fewer needles. From 1825
to 1843 many unrimed dendrites (few needles at about 1834) fell with few of those as aggregates.
After 1855 most of the crystals that fell were dendritic with needles, with stellars and column observed
occasionally. The dominance of dendritic crystals on the downwind side leads to the conclusion that
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
83
regions near cloud top were the primary source of ice particles that fell over TD. The same reason
may be why very few needles were observed.
A summary of the habit observation made at TD is given in Fig. 4.2.10 A&B. Part A of the
figure shows the different crystals that fell in percentage by number. In this figure, the heavily rimed
crystals were mostly of dendritic base. Most of the dendrites were light to moderately rimed and were
aggregated. This indicates SLW located below the -15°C level for riming and relatively warm (> -4°C)
conditions for aggregation were met before the dendritic crystals reached the ground. Longer in-cloud
residence times (because of relatively higher origin in the cloud) for dendrites falling on the downwind
side than upwind side of the Mountain barrier and higher growth rate around the -12°C might also
have increased their chances for further aggregation.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
84
N o . PERCENTAGE OF CRYSTAL
HEAVY RIMED, WARM, PLANAR Sc MISC.
120
100
BO
60
40
20
0
18:45
19:15
■ I HEAVY RIME
^
WARM
19:45
I
PLANAR
20:15
IS3SS51 MISCELLANEOUS
DECEMBER 18. 1088
N o . PERCENTAGE
RIME, AGGREGATES & FRAGMENTS
AGGREAGTES
FRAGMENTS '
USD RIMED
DECEMBER 18. 1888
Figure 4.2.10 Relative numbers (in percent) of observed ice crystals on December 18,1986 at TD.
Heavy rimed, warm origin, planar and other type of crystals (A) and rime, aggregate and fragments
(B).
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
85
Observations over the Park Range mountains of Colorado (Uttal et. al., 1987) and the Sierra
Nevada (Huggins et. al., 1985) have shown high ice particle concentration of smaller size to be found
on or slightly downwind (Choularton and Perry, 1986) of the crest.
That is, maximum ice
supersaturation occurs just upwind of the crest where the updraft is largest. However, due to the wind
drift, the maximum is shifted, at lower altitudes, to slightly downwind of the crest. This also increases
the chance for aggregation on the downwind side.
Not as many crystals were highly rimed at TD compared to KGV. This may be due to
depletion of the SLW (due to evaporation) which showed a peak of 0.65 mm at KGV and 0.40 mm
at TRK and was consistently lower. It is also true that more heavily rimed particles have highest fall
speeds and fallout first; i.e. furthest upwind.
Crystals caught in the higher wind speeds at higher altitudes move quasi-horizontally before
growing to very large sizes and fall out. Eventually, these crystals have crossed the crest by the time
they have grown to larger sizes. Measurements and modeling have shown (Meyers and Cotton, 1988)
that the maximum in pristine ice crystal concentration is found slightly downwind of the crest and
lower in altitude than the location of the SLW maximum. Also, on the downwind side, ice saturation
decreases more slowly than water saturation. This enables the crystals crossing the summit to
continue growing for certain distance downwind while water droplets evaporate. These conditions
Combined with relatively warm environment lead to the observed higher aggregation and less riming
on the lee side. Snowfall intensity throughout the sampling time was light.
The observation of the dendritic crystals indicate that cloud were not very cold and the
smaller number of needles and columns falling at TD than at KGV may be due to higher cloud base
at the lee side. The higher cloud base is also, again, due to compressional warming and loss of liquid
water by precipitation. In addition, needles and columns form at lower altitudes, possibly due to ice
multiplication and are more likely to fall out on the upwind side. Conditions favorable for this
process are observed often in the upwind side. However, data used to plot Fig. 4.2.10 (calculated by
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
86
colleagues Zhang and Mitchell and in preparation for publication in JAM by Demoz et al) shows that
about 30%-40% of the total snow that fell at TD was associated with warm habit crystals and this
accounted for a 2.4% mass percentage while 85% of the mass was due to rimed dendritic crystals. The
remaining 12.6% was contributed by stellars graupel and fragments. This suggests that the SLW was
located at elevations below where dendritic crystals originate and grow to sizes capable of capturing
the SLW drops. This will lead to temperatures of -10°C or lower to be the location of SLW which
is usually located below about 1 km above the peak of the crest. This is in agreement to the findings
of Warburton and DeFelice (1986) from stable isotopic study. Note also that this conclusion is
reached in Part III from the stable isotopes frequency distribution of the 1984/85 and 1985/86 winter
seasons. These findings indicate that warm habit crystals, despite their large numbers, contribute little
to the total snowfall mass in the Sierra Nevada.
From Fig 4.2.10B, fragmentation was observed throughout the sampling time. Almost 40%
of the crystals were fragments mainly of dendritic origin. A feature shown by the figure is that, within
the sampling time, no significant change in aggregation, riming and fragmentation was observed. This
may be a result of the slow movement of the storm that may have led to the same environmental
conditions of crystal growth for extended periods of time. The changes observed show (though small)
increases of rime and fragmentation were associated with the decrease in aggregation.
In Fig. 4.2.4A, the ground sampling period is indicated. It shows a trend of decreasing SLW
from 0.4 mm at 1815 to a minimum of 0.1 mm at 1915 and increased afterwards to 0.35 mm at 2115.
Between 1830 and 1915 the plot shows a relative maxima of SLW up to about 0.6 mm. During this
time, Table 4.2.3 shows that most of the dendrites were rimed and aggregated and graupel was also
observed. Note that at this time, only the planar crystals were aggregated and the needles were
unrimed. At TRK (Fig. 4.2.4 and Table 4.2.2), Slw during the sampling time (1815 -2100) was
generally greater than 0.20 mm with the highest value of 0.40 mm at about 1930. During this time,
the ground observation shows that the majority of the crystals were dendrites with moderate to heavy
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
riming and graupel observed at 1930. The table also shows that this time aggregation varied from few
to some and many (around 2000) and many of the aggregates were dendrites. Note that again not as
many needles were observed as at KGV. The general trend of SLW at KGV and TRK was the same
but KGV values were much higher.
The precipitation rates were below 2 mm/hr during the sampling time at TD. Precipitation
rate (mm/hr) measurements for KGV and Donner grade (a station close to TD) are given in Fig.
4.2.11A&B.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
88
8
SITE S13
0C
X
\
2
2
front
z
o
4
a.
ou
cc
CL
0
i
12
1
r
T
i
1------1------ r
B
SITE S82
cc
X
\
2
2
Z
O
front
■«h-»
la
4
Ll I
CC.
CL
■■■'
i . . , iiiH inxrbi1inxTtT£i,
M
f a .H lc v J ]
HE
4
oc
X
&TE SN6
3
z
0
2
1
a.
a
VC
1
Q.
leel
0
12
13
14 15 16 17
18 19 20 21
22 23
0
1
2
3
4
5
6
TIME (UTC)
Figure 4.2.11. Precipitation rate measurements cm December 18,1986 at KGV (A), Donner
Grade (B) and Tahoe Meadows (C). Note the position of the front as it moves east. TRK is
located east of S82 and west of SN6.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
89
4.3
THE STORM OF 19/20 DECEMBER 1986: - A CASE STUDY
Cutoff lows or large amplitude short-waves near 40°N systems have mainly south
southwesterly flow that shift to northerly and easterly after frontal passage. The low center
movement eastward is usually slow (15 m/s according to Heggli and Rauber, 1988). The 19
December storm is associated with a large amplitude shortwave and exhibits the above mentioned
characteristics. The 500 mb and surface weather maps are given in the appendix.
The temperature, relative humidity, winds and equivalent potential temperature contours
are in Fig. 4.3.1. These plots indicate a frontal passage at KGV between 0000 and 0300,
December 20. The frontal passage could be seen from the SLW and precipitation measurement
plots in Fig. 4.3.3 and Fig. 4.3.6 respectively. The frontal cloud band was placed over KGV at
about 0100 from the precipitation plot in Fig. 4.3.6 and produced maximum value of about 3
mm/hr. During this frontal precipitation, the SLW values dropped to very low values (Fig. 4.3.3).
The winds changed more east after the frontal passage at lower altitudes and northerly at high
altitudes and decreased markedly in speed. A drop of the moisture level is also evident from Fig.
4.3.1 starting about 0900, December 20. Precipitation stopped at about 0400 after frontal passage
at KGV at most of the stations west of the crest but continued on the eastern side until 0700
December 20.
A plot of the winds at SM is given in Fig. 4.3.2. It shows a wind direction o f 250° nearly
throughout the entire sampling period including the frontal passage while the speed decreased
starting 2000 until slightly after the front passed with the exception between 2300 and 0000
(possibly up to 0200) marking the frontal passage at SM.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
90
19 December*
1
15
L
12
I
09
I
06
.
■
03
,
00
r
■
21
18
15
Figure 4.3.1. Time versus height cross section from KGV soundings; (A) Relative humidity
(solid), temperature (dashed) and (B) equivalent potential temperature contours (B). Shaded
areas are where dQJdz < 0. A full wind barb is 10 m/s. Time increases from right to left.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CUTOFF LOW OR LARGE AMPLITUDE WAVE NEAR 4 0
SLIDE MT. DECEMBER 1 9 /2 0 , 1986
N
SLIDE MOUNTAIN
360
50
SPEED
DIRECTION
40
270
(/)
LU
O
30
180 &
LlJ
UJ
CL
</> 20
Q
LU
{£
Q
90
12 / 1 9 /8 6
12 / 2 0 /8 6
DECEMBER 19/20. 1991
t-J
20
04
00
TIME
08
M
M l
l U I I I M I S 'N '
(0•) (>•) <»••)<«•)(»»•)(»»•)
WIND . S P U D 5C M £ (KNOTS)
NOTE - WWO OMECDON S THE
DMCCnOM WIND £ SLOWMO FROM
Figure 4.3.2. Measurements of wind speed and direction at Slide Mountain weather station (A) and
wind rose analysis (B) for December 19,1986.
\©
92
The wind direction was ideal for cross-barrier flow and SLW development. This crossbarrier wind direction and relatively weak synoptic system lead to extended periods of SLW over
the area. The effect of the re-development of clouds over the Carson Range was evident from
the precipitation plot made across the barrier in Fig. 4.3.5A. Relatively high values collected at
station SN7, SN9 and SND located east of TRK compared to stations over the valley (S29, S33)
was mainly due to orographic effect of the Carson Range. Lee side effects, or a precipitation
"shadow" is apparent north of TRK (stations S29, SN1) where total precipitation was greatly
reduced.
Mean SLW at TRK for the day was 0.06 mm while at KGV the mean value was 0.16 mm.
The vapor value also showed an increase of 0.25 cm from a mean value of 0.70 cm at KGV to
0.95 cm at TRK. Vapor maximum increased from a value of 0.91 cm at KGV to a value of 1.16
cm over TRK while SLW maximum at KGV was 0.48 mm to 0.18 mm at TRK.
The SLW measurements for this case are given in Fig. 4.3.3A and Fig. 4.3.3B for KGV
and TRK respectively. It shows an abrupt decrease at 2300 to less than 0.1 mm at TRK and less
than 0.2 mm at KGV. The maximum SLW was found in pre-frontal clouds just before 2300 and
was 0.48 mm at KGV (0.18 mm at TRK). This drop occurred about one hour earlier than the
passage of the front as seen in Fig. 4.3.1 and Fig. 4.3.6. The frontal band was noted from radar
(not shown here) at about 0245 according to Hemmer et al., (1987). SLW was measurable for
several hours following frontal passage. A summary of the SLW is given in Table 4.3.1 and a plot
of the precipitation at KGV, Donner Grade and Tahoe Meadows is given in Fig. 4.3.6. As
mentioned earlier, the extended precipitation observed east of the main crest might be due to
shallow cloud development over the Carson Range for several hours after frontal passage. Note
also that the low center movement was slowed down due to the ridge east of the sampling area.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
861219/20:1030-0730
1030 (GMT) 19 DEC 1986 to 0730 (GMT) 20 DEC 1986
rz
-s
1.00
0.60
microphysics sampling
L iquid-------V apor ---------
0.80
s
s
0.40
0.60 ®
f it
O
a
u
m
0.40 *2.
0.20
*n
•N
9er
3
0.20
0.00
10
0.00
12
14
16
18
20
22
TIME (GMT)
0
2
4
6
8
861219/20:1030-0730
1030 (GMT) 19 DEC 1986 to 0730 (GMT) 20 DEC 1986
rz
-s
- 1.20
0.30
liq u id
V apor
microphysics sampling
J
g
g
0.80
0.20
u
2m
*
„
•*t
0.40
0.10
9
or
9
3
o .o o iy ^ n n ^ h lllli.
10
12
14
16
18
,
,!
20
22
TIME (GMT)
0.00
0
2
4
6
8
integrated cloud liquid water and precipitable water
a t KGV (A) and TRK (B) for December 19,1986.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
TRUCKEE: DECEMBER 19, 1986
1703 (gm t) 19 dec 1986 to 2200 (gm t) 19 dec 1986
Threshold =0.05 m m
Contour Interval=0.05 m m
360
0.1
(Deg)
320
280
240
e
°-t ....... 0.1.........O.i---- "
.-•0.1 —\
..—0.10.1 'O
v
o.i
.
-
0.1
...
- -0 .I
0.1--.
—
0.1
- -
0.1
' o.i
-
0.1
* 0.1......
200
Azimuth
- 0 .|
— 0.1 —
— 0.1
—
0.1
0.1
160
o
o
120
"o
80-
0.1
0.1
0.1
0.1
0.1 —
o '
' fti —
'■
40
0.1
—
0.1
-
22
TIME (g m t)
Figure 43.4. Spatial distribution of radiometer measured integrated cloud liquid water at TRK
on December 19, 1986.
95
The radiometer scan measurements at TRK are given in Fig. 4.3.4 in time/azimuth plot
form. SLW maximum between 160°-200° and 320° from 1700 to about 2100 was observed. These
maximum values were mainly due to shallow clouds over the main barrier to the west. Note also
that the plot does not show much SLW from clouds over the Carson Range to the east except
at 2030 to 2100. During the time of the scan, precipitation over the stations located to the east
of TRK was not observed.
The SLW value distribution at both stations is plotted in Fig. 4.3.7. For TRK, it shows
more than 50% (30% for KGV) of the time, the value was less than 0.05 mm, usually considered
a threshold. The maximum SLW at TRK was half as much of that of KGV and falls off rapidly.
A distribution of the instantaneous value measurements of the radiometers together with
precipitation rate measurements at selected stations is given in the Appendix.
4.3.1
PRECIPITATION AND MICROPHYSICS
Precipitation sampling was done from 2350 to 0315 at TD and from 2115 to 2317 at
KGV. The ground sampling period at KGV coincided with the pre-frontal SLW maximum
observed in Fig. 4.3.3 and precipitation rate of 1-2 mm/hr while at TD, very little SLW (threshold
values, 0.05 mm) and precipitation of less than 1 mm/hr were observed.
The ground
microphysics observations are presented in Table 4.3.2 and 4.3.3 respectively. The tables show
that the crystals at TD were unrimed and of colder temperature origin while at KGV mostly
needles were observed. This reduction in riming is considered to be due to the effect of
evaporation of the SLW downwind of the main crest. Snowfall intensity at both stations was
generally light. As observed in the 18 December case study, aggregation was higher when
dendrites fell and so was riming. In this case study, the crystals observed were unrimed and colder
origin at TD which is expected by looking at the SLW from TRK (Fig. 4.3.3). That is, the storm
was greatly affected by evaporation on the lee side. This was also indicated from the spatial SLW
distribution in Fig. 4.3.4. At KGV, on the other hand many warm habit and rimed crystals were
observed.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
96
STATIONS AROUND TRUCKEE
2604
PRECIP. (MM*100> H i SITE ELE. (MSL)
DECEMBER IS . 1988
TOTAL PRECIPITATION
2945
2087
1573
NWS S99 S I3 S ll
S21 SB4 S86 S29 S33 SN7 SN9 SND
SITE ID
PRECIP. (MM* 100)
SITE ELE. (MSL)
DECEMBER 19. 1088
Figure 43 3 . Total precipitation (mm) accumlated on December 19,1986 at stations around TRK
(A) and across the barrier (B).
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
97
cc
X
N
2
2
4
front
...te r* ..
3
to 2
LJ
OS
7to 1
5 o
<*—
•
6
8
10
12
14
16
18
2D
22
0
2
4
6
8
10
12
TIME (UTC)
1
1
1
------- 1------- 1 ---- 1
1
T------- 1------- 1------- 1------- 1------- 1-------
front
....... I ____ i_____L ___ L .....*
II1 . n imn . iAmni..
6
8
10
i
12
..111 inniOiiii
tin m mi
nil
14
16
i iiji
n i n ..ifi ihnn.M
. iliiiurninrrm
r t i h t r n*i
18
20
22
0
2
4
■__ ii__.__
, i _ .■_ ii
*
8
10
■ t
12
TIME (UTC)
<o
w
TIME (UTC)
Figure 4 3 * . Precipitation rate measurements on December 19, 1966 at KGV (A), Donner
Grade (B) and Tahoe Meadows (C). Note the position o f the front as it moves east TRK k
located east of S82 and west of SN6.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
KINGVALE PRECIPITABLE LIQUID HATER: 19 DEC. 1906
KINGVALE SUPERCOOLED LIQUID HATER: 19 DEC 19B6
1
1
J _ _ ------------------------------ ------------------------------------ ----------------
28
32 -
LIQUID
24 -
VAPOR
28 24 -
20
M
-
Z
20
-
16 12
-
84-
0
m.
mk
16 -
a3
ui
„ •*!
i-KP V
12
-
Ms:
I®
S®.
IP III?
si®
3
J~
SUPERCOOLED LIQUID HATER IN HH
PRECIPITABLE LIQUID HATER IN CM
TRUCKEE SUPERCOOLED LIQUID HATER: 19 DEC. 19B6
TRUCKEE PRECIPITABLE LIQUIO HATER: 19 DEC. 19B6
_L_
- I ----------------- 1___________ I___________ I___________ I___________ I___________ L.
28
LIQUID
VAPOR
24
20
-
16 -
12 84-
0.2
.4
SUPERCOOLEO LIQUID HATER IN HH
3------ 1------ T
i
PRECIPITAbi t LIQUID HATER IN CM
Figure 4.3.7. Frequency analysis of the radiometer measured integrated cloud liquid (mm) and
precipitable water vapor (cm) at TRK and KGV on December 19,1986.
$
99
Table 4.3.1 Summary of radiometer measurements on 861219-20.
861219-20
VAPOR
LIQUID
SITE
MEAN
MAXIMUM
MEAN
MAXIMUM
KGV
0.16
0.48
0.70
0.909
TRK
0.06
0.18
0.9S
1.160
Table 4.3.2 Ground cloud-microphysics observations at TRK on 861219-20.
TIME
HABIT
RIME
AGG.
COMMENTS
2350
dendrites, Nle, Nla
L-M
few
light snowfall
0040
mostly dendritic
L-M
0050
Pie, Cle, Capped columns, bullets, rosettes
U-L
many
moderate
0115
bullets, rosettes, hollow columns
U
some
moderate to intense
0130
rosettes, side planes, Pld & small Pie
U
few
light(T=-2.5°C)
light
very light
0145
0200
Clf, Pla and mostly side planes
U
none
barely snowing
0300
almost all dendrites, plates Pic, P2c, P2d
U
some
light-moderate
stop sampling T=-3°C
0315
Table 4.3.3 Ground cloud-microphysics observations at KGV on 861219-20.
TIME
HABIT
2115
frozen drops
2123
Nla, dendritic fragments
2129
same
2131
2132
RIME
AGG.
COMMENTS
light snow
L-M
few
small size
Nla, few drops
M-H
few
light
N la
L-M
light
sun visible
2144
2251
plates, dendrites
L-M
few
light
2303
fragments
L-M
some
light
2310
Nla, graupel
M-H
some
2313
Nla, granpel, dendritic fragments and frozen drops
M-H
some
2317
same
very light
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
100
4.4
THE STORM OF 22/23 DECEMBER 1986: - A CASE STUDY
The general synoptic scale structure of the storm and its classification is similar to the
storm of 18 December 1986 discussed above. It was classified by Heggli and Rauber (1988) as
split flow in the middle troposphere associated with a dissipating cyclonic storm. The cause for
this split and dissipating nature was the high pressure center located over eastern Nevada. The
500 mb heights and the surface weather map are given in the appendix. A discussion of the
storm system was given by Heggli and Rauber (1988) and Huggins e t al., (1990) also present the
radar data at KGV.
The surface front passed KGV at 1700-1800 December 22 and was notable as a single,
narrow, frontal precipitation band (precipitation rate values are plotted in Fig. 4.4.6). An
orographic cloud remained after the frontal passage over the Sierra for several hours and
produced some precipitation. Unlike the 18 December storm, precipitation did not occur at
KGV, prior to the frontal passage.
Figure 4.4.1 shows humidity, temperature, equivalent potential temperature and wind
measurements over KGV. At KGV, the passage could be placed (from Fig. 4.4.1) between 1800
and 2100 (see also Fig. 4.4.6). The soundings (relative humidity) indicate a double layer of clouds
(seeder-feeder clouds) over KGV.
The wind measurements at SM on this day are plotted in Fig. 4.4.2. It shows a constant
direction but decreasing speed until 0300, 23 December. A t 0300, the wind direction changed
from westerly to northerly and northeasterly accompanied with a slight increase in speed. This
change in speed and direction was associated with the intensification of the high pressure center
located in eastern Nevada and weakening of the storm completely. The passage of the front at
SM, between 1700 and 1800 was accompanied by a drop in wind speed but no change in
direction.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
101
22 December
11
10
Wo
3M;
9
JC/3 8
X
3 7
313-^.
,3/f
£C5 6
M
« 5
wt
4
•3/0
.aw.
3
2
06
09
03
00
21
IS
15
22 December
t.
V
w
«.
§6
a»
—-a o i c ------ L
^SO
%
_____ v-J*
,v
7 0 - '^ —
V
5Q:
.70.
Figure 4.4.1. Time versus height cross section from KGV soundings; (A) Relative humidity
(solid), temperature (dashed) and (B) equivalent potential temperature contours. Shaded areas
are where dQJdz < 0. A full wind barb is 10 m/s. Time increases from right to left.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
SLIDE
MT.
22 DEC
1986
SLIDE MOUNTAIN WEATHER STATION
N
< — , — > 23 DEC
360
50
dir.
speed
40
270
»•
oLU
Q■
to
i—
O
z 30
1C
Q
asf
180 o
LU
LU
UJ
CL
¥
o
o
z
(/>
a
z
90
10
2 2 /2 3 DECEMBER 1986
0530/22 TO 1030/23
1 -1
0
« -•
3 - 1 —□
7-1 0 1 1 -1 0 11-SI 71-1
( 7 • ) <• • ) <io . X u - I B • ) « l •>
8
12
16
20
00
TIME (UTC)
8
WHO SPEED SCALE (KNOTS)
NOTE - WHO DIRECTION IS THE
DIRECTION VANO IS BL0WNC FROU
Figure 4.4.2. Measurements of wind speed and direction at Slide Mountain weather station (A) and
wind rose analysis (B) for December 22,1986.
1 knot = 0.447 m /s
103
061222/861223:1330-1030
1330 (GMT) 22 DEC 1986 to 1030 (GMT) 23 DEC 1986
fZ
-s
1.50
Liquid ----------
yap0r
.
.
-1.50
.
microphysics ssnipliii^
S
S
1.00
JS
w
O
.
0>
a
1.00 ^
■c
^ 0.50
0.50
S
O’
3
0.00
23
1
TIME (GMT)
0.00
861222/23:1330-1030
1330 (GMT) 22 DEC 1986 to 1030 (GMT) 23 DEC 1986
r-Z
-s
-1.50
0.40
Liquid
microphysics ru sampling
V apor
0.30 H
fHf
1.00
•*
0.20
0.50
5- 0.10
°-00i SITW
o.oo
23
1
TIME (GUT)
Figure 4.4.3. Radiometer measured integrated cloud liquid water and precipitable water vapor
at KGV (A) and TRK (B) for December 22,1986.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
Radiometer SLW was recorded starting 1300 throughout the lifetime of the storm (Fig.
4.4.3A and B). It was at its greatest at KGV, up to 1 mm, prior to 1700 (time of frontal passage).
The clouds at this time were pre-frontal clouds of high base. Cloud base was found from the
soundings at KGV to be about 1 km above ground in the beginning and lowered later. The sharp
SLW drop at 1700 to less than 0.1 mm was associated with the passage of the front and frontal
band precipitation at KGV. At this time, the SLW at TRK, was also considerably less than 0.1
mm. It was at its maximum after 2100. It has been suggested by Huggins et al.,(1989) that a
decrease in the intensity of seeder clouds on the upwind side allowed more SLW to advect over
the crest leading to the observed increase at T R K The time/azimuth plot of the SLW in Fig.
4.4.4 shows, from TRK, the maxima between 1900 and 0000 to be in the easterly quadrant (80°
to 120°). Since the wind speeds at SM weather station were from about 260”, this suggests that
clouds from south of KGV accompanied by redevelopment of SLW over the Carson Range were
responsible for much of the SLW recorded after 2100 from T R K This re-development of clouds
over the Carson Mountains was also evident from the total precipitation plot made cross-barrier
and around TRK given in Figs. 4.4.5A and 4.4.5B respectively. SLW values, which were about
1 mm at KGV and less than 0.3 mm at TRK on the average, and heavy riming of crystals
observed (discussed in the next section) suggest that the SLW was efficiently incorporated into
the precipitation during much of the storm duration. This was particularly true between 1700 and
2000 at KGV when SLW was low and precipitation was maximum. A summary of the liquid and
vapor measurements for KGV and TRK are given in Table 4.4.1.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
TRUCKEE: DECEMBER 2 2 /2 3 , 1986
1700 (gm t) 22 dec 1986 to 0100 (gm t) 23 dec 1986
Threshold =0.05 m m
Contour Interval=0.05 mm
360r
.a
120
20
21
22
TIME (g m t)
Figure 4.4.4. Spatial distribution of radiometer measured integrated cloud liquid water at TRK
on December 22, 1986.
8
106
STATIONS AROUND TRUCKEE
2945
2604
2225
I 890
1945
i
SB6
1931
l7 fl3
2195
1390
‘— 1----- — 1-- -—1-- -— 1----— 1----—r
S84
S29
SN1
SNH
SN7
SN6
SN5
SN4
SITE ID
I PRECIP. (MMMOO)
SITE ELE. (MSL)
DECEMBER 22. 1868
TOTAL PRECIPITATION
B
2945
2087.
1945 l a a o
1783
1573
T
I ‘ I ‘ I
NW6 S09
S13
S ll
I '--!
S21
‘—I-1—I ‘—I-c—1--‘—I--- —|-
SB4
SB6 S20
S33 SN7 SN9
SND
SITE ID
PRECIP.
(MMMOQ}
SITE ELE. (MSL)
DECEMBER 22. 1986
Figure 4.4J . Total precipitation (mm) accumlated on December 22,1986 at stations around TRK
(A) and across the barrier (B).
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
107
12
1
OS
!
!
!
!
!
i
|— |— j— i— |— j— j— j— j— j— i— i— i— i— j— r
X
jfronti
S 8
a
a
£ 4
i
rO
J
12
L
14
i* lil hT
16
18
HI
20
in nim-jrl I hi 11 i nim irrflh
22
0
2
j
4
j
6
j
j
j
8
j
10
TIME (UTC)
cc
X
— J— ,— J— ,— J— p
12
—j— !— r~ j
* » j • ’ ’! ' ! ! ! ! !
"j
8 ~ 4 " 4 ... j-— —ff r o i
5
a
8
£
Il
C
ao
| e ft
4
1_L i
12
14
<n
A
i irri.fi
16
18
l l f L i r Twniwfli[H i i i o.i
20
22
0
2
i i i i i i i
4
6
8
10
TIME (UTC)
:—
cc
n:
s
{ ———
'!— — —
3
—— —
j
__ _ __ , .... ....
i
!'■'!—T“ 1 j
front
!_ _ _ i
12
14
—
16
d
18
im i
20
22
0
2
4
.,..1. _[...L
6
8
i
10
TIME (UTC)
Figure 4.4.6. Precipitation rate measurements on December 22, 1966 at KGV (A), Donner
Grade (B) and Tahoe Meadows (C). Note the position of the front as it moves east. TRK is
located east of S82 and west of SN6.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
108
4.4.1
PRECIPITATION AND MICROPHYSICS
The ground microphysics observations at KGV started at 1730 and ended at 0020. At
TD, the observations were made from 2145 to 0100. As mentioned earlier, the crystals fell from
post-frontal clouds at TD and frontal as well as post-frontal clouds at KGV. The crystal types,
riming and aggregation information are given in Table 4.4.2 and Table 4.4.3 for both stations.
The observations show that the crystals that fell at KGV consisted mainly of needles, dendrites,
capped columns, plates and frozen drops. After 2000, lots of frozen drops with needle extension
(referred here as "Hatchets") were observed. Prior to 1958, many of the crystals observed were
aggregates and rimed dendrites while the needles were unrimed or very lightly rimed. This
crystals fell from frontal band of clouds when precipitation was heaviest. Echo top temperatures
at this time were -20°C or colder (Huggins et al. 1989). From 1958 to the end of the sampling
time, a large amount of frozen drops and "Hatchets" were observed. During this time aggregation
decreased slightly and riming was moderate to heavy most o f the observation period. The SLW
at this time show a slight increase which was due to orographic clouds with embedded convection.
This observation of heavy riming and lots of needles at KGV is also an indicator that a secondary
ice-multiplication processes was taking place. The observation of graupel also indicates that
convective activity and "pockets" of high SLW were present as shown using radar data by Huggins
et el. (1989).
At TD, observations showed a large number of needles, graupel and dendritic crystals
with moderate to heavy riming before 2300 and unrimed needles from 2315 to 0100. As can be
seen from Fig. 4.4.4, most of the SLW measured before 2300 at TRK was due to orographic
clouds west, over the Sierra Crest, and east, over the Carson Range mountains. Snowfall was
light and aggregation was observed throughout the sampling time. A summarized plot of the
observations is given in Fig. 4.4.7. The warm habit crystals (Nib, N le, N2a) made up 27%,
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
heavily rimed dendritic or planar crystals 71% (R3b, P7a) and 2% of the snowfall mass was
contributed by 14 and R2b type crystals. The SLW during the snow sampling time period was
less than 0.25 mm at TRK (except for a single data between 2130 and 2145) while at KGV values
were between 0.1 mm and 0.35 mm. In this case study, the decrease in SLW was accompanied
with heavy riming in both KGV and TD. Hence, the SLW reduction must be due to efficient
precipitation producing processes. This is similar to the conclusion suggested by Huggins et al.
(1989).
Table 4.4.1. Summary of radiometer measurements on 861222-23.
LIQUID
VAPOR
SITE
MEAN
MAXIMUM
MEAN
MAXIMUM
KGV
0.20
1.03
030
1.27
TRK
0.12
031
1.16
139
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
110
N o . PERCENTAGE
PLANAR, WARM & RIME
too
80 SO
40
20
-
0
21:45
22:00
r— — i-----— i------ — t
— r~
22:14 22:21 22:30 23:00 23:30
IPLANAR
HI
RIMED
1:00
EZH WARM
DECEMBER 22. 1006
N o . PERCENTAGE
AGGREGATE, RIME & FRAGMENTS
B
T
21:45
i
22:08
T i
22:14
'
i
22:21
I AGGREGATE
'
i
i
22:30
23:00
* I ‘ I
23:30
FRAGMENTS
1:00
(HI RIMED
DECEMBER 22. IBM
Figure 4.4.7 Relative numbers (in percent) of observed ice crystals on December 22, 1986 at TD.
Heavy rimed, warm origin, planar and other type of crystals (A) and rime, aggregate and fragments
(B).
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
KINGVALE PRECIPITABLE WATER: 22 DEC. 1906
18 -
KINGVALE SUPERCOOLED LIQUID WATER VALUES 22 DEC. 1906
10
-
l iq u id
15 -
15 -
12
12
-
VAPOR
-
z
Z
3
a
UJ
TRUCKEE PRECIPITABLE LIQUID WATER: 22 OEC. 1986
TRUCKEE SUPERCOOLED LIQUID WATER: 22 DEC. 19B6
18 -
18 -
LIQUID
1512
1.2
PRECIPITABLE WATER CH
SUPERCOOLEO LIQUID HATER IN HH
VAPOR
15 12
-
-
PH
6
3
0
SUPERCOOLED LIQUID WATER IN HH
.4
6
.8
1
1.2
PRECIPITABLE LIQUID WATER IN CH
Figure 4.4.8. Frequency analysis of the radiometer measured integrated cloud liquid (mm) and
precipitable water vapor (cm) at TRK and KGV on December 22,1986.
112
Table 4.4.2. Ground cload-micropfaysics observations at TD on 86122-23.
TIME
HABIT
RIME
2145
N le, dendritic
M-H
2206
R3b, N le
2214
mostly N le
M-H
2221
same
M
2230
small P ie & R3b, some N la & N le
2245
2300
AGG.
COMMENTS
many
T—0.7°C
L-H
some
light snow
most N la, some small R3b
L-H
few
light
mostly small R3b
M-H
few
light
2315
no snow
2330
all N la
U
many
barely snowing
0100
all N la
U-L
few
stopped
Table 4.43. Ground cloud-microphysics observations at KGV on 861222-23.
TIME
N la
F la
1730-1812
UL
U*
1812-1823
V*
MH*
MH*
1823-1840
V
MH*
MH*
1840-1903
yf
y/
MH*
MH
1903-1914
yf
yf
MH*
y/
V
LM
1942-1958
L*
yf
1958-2101
MH*
1914-1942
yf
V
2101-2130
MH*
2130-2156
LM*
2156-2223
P ic
P ie
C2a
P7b
C Pla
St
R4b
LM*
FD
HT
M
yf
M
M
yf
yf
MH
LM
MH
yf
yf
yf
MH*
yf
yf
yf
MH*
yf
yf
2223-2251
MH
yf
yf
2251-2310
V
yf
yf
2337-2341
MH*
MH
2341-0020
MH
MH
LMH
yf
yf
* indicates aggregation was observed; HT stands for "hatchets"
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
113
4.5
THE STORM OF 03/04 JANUARY 1987: - A CASE STUDY
This sttirm was classified as a developing storm in strong southwesterly flow. This
classification is the same as that of the 26/27 March, 1985 case study presented in part III earlier.
The soundings for this day given in Fig. 4.5.1 reveal a deep cloud system at KGV. At KGV,
frontal passage was between 0000-0300, January 4. The frontal passage was clearly visible from
the moisture as well as the temperature (and equivalent temperature) plots at both stations.
Precipitation at KGV fell from pre-frontal clouds starting from 1200 up until the frontal passage
at 0300. Radar data (from Hemmer et al., 1987, presented in the appendix) showed a sharp edge
when precipitation rate was at its highest (9 mm/hr) just before the front. The precipitation on
this day fell from a series of bands the largest of which coincided with the front.
Winds at SM are plotted in Fig. 4.5.2. The plots show a 250° approach and very high
speeds prior to the front. The sharp drop in speed starting at 0300 signaled the passage of the
front. As the surface cold front approached, wind became weaker and more northerly. Part B
of this figure shows, 70% of the time the winds were from 250° direction and are good for cross
barrier flow and SLW development.
Radiometer data at KGV was not available until 0400 January 04. SLW at 0400 over
KGV was less than 0.1 mm and progressively decreased to zero. Most of the data at TRK was
also not available before 0000 January 04. At TRK the radiometer was operated in scan mode
just before frontal passage time. The scans (Fig. 4.5.4) made at the time show the clouds
approaching the area mainly from the southwest. When converted to zenith, the values were 0.35
mm at 2200 (scan # 1) and a drop to less than 0.1 mm (scan # 2) at 2300 and were between 0.1
mm and 0.15 mm at 0100 (scan # 3).
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
114
03 Januaiy
U
10
9
co 8
x
5 7
H
X 6
oM
g 5
V O
301,
V»*
_3«Y_
Jca
4
3
2
03 Januaiy
li
10
9
v
70
<r
4
3
SO
-3 d =
V
SO
702
-70.
J 10 ^
J N
,u>ya
7/ S
•'I/Stow
Figure 4.5.1. Time versus height cross section from KGV soundings; (A) Relative humidity
(solid), temperature (dashed) and (B) equivalent potential temperature contours. Shaded areas
are where d6y& < 0. A full wind barb is 10 m/s. Time increases from right to left.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
SLIDE MOUNTAIN
DEVELOPING STORM, IN STRONG SOUTHWESTERLY FLOW.
SLIDE MT. JANUARY 0 3 /0 4 , 1987
50
360
SPEED
DIRECTION
40
270
o
i
O
UJ
o
v-y"
30
z
o
180 1o—
o
Lul
UJ
UJ
gc
CL
(O
.01
CALU
20
o
o
Z
5
90
JANUARY 03/04-, 1991
0 1 /0 3 /6 7
0 1 /0 4 /8 7
»■) 0 »
0 1 a) ( l a )
(11
a ) ( n a)
WIND SPEED SCALE (KNOTS)
20
00
TIME
04
08
NOTE - WIN0 DETECTION IS THE
OUtECIUH WIND IS StDWMO FROM
Figure 4.5.2. Measurements of wind speed and direction at Slide Mountain weather station (A) and
wind rose analysis (B) for January 03,1987.
0000 (gm t)
4 jan 1987 to 1200 (gmt)
4 jan 1987
0.15
1.20
liquid
Vajpor
a
Water
a
0.10
0.80
Vapor
Depth
<u
(0
0.40
-t-i
(cm )
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
TRUCKEE: JANUARY 0 3 /0 4 , 1987
P
<T
S
0.00
0
1
2
3
4
5
6
7
TIME (gmt)
8
9
10
11
Figure 4.5.3. Radiometer measured integrated cloud liquid water and precipitable water vapor
at TRK, January 03, 1987.
0.00
12
H-*
&
117
SCAN
H t :8 7 l) 1 0 3
W
SCAN # 2 ;8 7 0 1 0 3
SCAN # 3 :8 7 0 1 0 3 /0 4
w— i— .— ■
j
»
» • ■—
0.02 m m
Figure 4.5.4. Radiometer scan measurements made over TRK on January 03, 1987. Scan #1:
2130 - 2200, #2: 2300 - 2330, #3: 0000 - 0030 UTC.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
e
118
The measurements imply that high SLW was observed prior to the frontal passage as did
the other case studies. Note in Fig. 4.5.3 that the SLW falls off to less than 0.05 mm (threshold
values) after frontal passage. The SLW measured at TRK after the frontal passage was mainly
from high clouds that were visible on the satellite pictures (not shown here) for the day. A
distribution of the instantaneous SLW values for the available data at TRK is given in the
appendix. It shows more than 60% of the time it was less than 0.1 mm.
4.5.1
PRECIPITATION AND MICROPHYSICS
The precipitation measured at stations around T R K does not show high variability with
topography. This is mainly due to the speed of the system as it moved across the area which is
different from the case studies presented earlier. The precipitation cross barrier and from
stations around TRK is given in Fig. 4.5.5.
Ice crystal observations were made at Tahoe Donner on this day from 2100 to 0430. The
data is presented in Table 4.5.1 and it shows an interesting shift from warm to cold habit crystal
dominance as the front passed the area. Prior to 0100, there were high percentages of warm
habit and planar crystals. After 0100, when SLW was generally below 0.1 mm and the front was
very close (probably the upper part of the front was over the sampling area) the crystal habit
changed into some planar and primarily cold habit origin (Fig. 4.5.7). Many of the crystals at this
time were side planes. The switch in crystal type was due to the strong temperature gradient
across the frontal boundary. The surface temperature during the sampling time changed from 1.2°C to -5.2°C.
Figure 4.5.7, presents the number percentage of rimed, aggregates and fragment (A) and
planar warm and cold crystals (B) observed on this day. Note that the observation of rimed and
warm habit crystals was strongly correlated. Both riming and warm habit crystal concentration
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
119
decreased slowly and were not observed after 0115 when the crystals were mainly cold habit. By
this time SLW was below 0.1 mm as noted above and decreasing. As surface temperature
became about -5°C colder origin crystals, more aggregates but less fragments and no riming was
observed. From Table 4.5.1 and Fig. 4.5.3, it is clear that the unrimed crystal observations were
made after frontal passage and when SLW was below threshold (0.05 mm). Note that more
fragmentation was observed when more rimed particles were observed.
This case study clearly shows that warm temperatures are not necessary for aggregation
to take place. It also shows the crystal type change across a cold frontal boundary from mainly
warm habit to mainly cold habit origin. The temperature change associated with the cold frontal
passage over the station was large. Like the other case studies, for the lee side, the SLW drops
off dramatically to very low values after frontal passage. In this case very low values (0.05 mm)
were observed. This is the reverse of what is often observed at the upwind site.
The habits of falling ice crystals varied dramatically from storm to storm and within a
single storm before and after frontal passage. These habits are a reflection of the temperature
and moisture conditions within the cloud where the ice crystals grew. For example, in the earlier
case studies, the ice crystals were heavily rimed, but relatively few aggregated. In this case study,
the crystals were lightly rimed but many aggregated; a direct reflection of the SLW available in
the storm.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
120
STATIONS AROUND TRUCKEE
2945
1763
PRECIP. (MM* 100)
H H SITE ELE. (MSL)
JANUARY 3. I960
TOTAL PRECIPITATION
2945
2067
1856<£^7| f p l l
1945 1890 17B3
iinii i S f i i
NW6 S99
S13
S ll
S21
SB4
SB6 S29
S33
SN7 SN9 SND
SITE ID
■ 1 PRECIP.
(MMMOO)
H H SITE ELE. (MSL)
JANUARY 3. 1986
Figure 4.5.5. Total precipitation (mm) accumlated on January 03, 1987 at stations around TRK (A)
and across the barrier (B).
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
121
az
x
12
7~!
!
!
!
i
I
i— i— !— r
0UJ
!£
4
^431^311213132433111321363
I
1
rO
<n
o
J
I
1
8
I
I
_al
L_
10
12
14
16
18
20
22
2
I
l_
4
6
TIME (UTC)
az
x
12
1
!
I
I
!
!
!
!
!
!
I
I
I
I
I
i— i— i— i— i— r
8
UJ
az
a.
4
i
N
cn
CD
J
'6
I n In
8
10
lo _
J
12
14
16
18
I
20
I
22
I
.
I
j
j
j
j
0
TIME (UTC)
az 12
x
t
i
i
I
J
I
8
I
L
10
i
i
i
i
i
i
i
i
i
i
i
i
i
i— i— i— r
8
o
UJ
az
CL
I
<13
z
<0
rmlfTT
J ___ 1
14
12
16
18
20
1
22 ..
TrltrfHilrTltl
i
o
TIME (UTC)
Figure 4.5.6. Precipitation rate measurements on January 03,1987 at KGV (A), Donner Grade (B)
and Tahoe Meadows (Q .
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
122
No. PERCENTAGE
PLANAR, WARM Sc COLD
2 1:1S 2 1 :45 2 2 :15 22:45 23:15 23:45 0:15 0:45
1:15
WARM
PLANAR
1:45 2:15 2:45 3:15
I
I COLD
JANUARY 03. 1907
No. PERCENTAGE
b
RIMED, AGGREGATES, FRAGMENTS
100
I
I
1
%Ug^jffs§g I
I
I
2 1 :1 5 2 1 :4 5 2 2 :1 5 2 2 :4 5 2 3 :1 5 2 3 :4 5 0:15 0:45 1:15 1:45 2:15 2:45 3:15
I RIMED
I AGGREGATES
CZ1 FRAGMENTS
JANUARY 03. 1987
Figure 4.5.7. Relative numbers (in percent) o f observed ice crystals on January 03,1987 at TD.
Heavy rimed, warm origin, planar and other type of crystals (A) and rime, aggregate and
fragments (B).
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
Table 4.5.1. Ice crystal habit for the storm of January 03, 1987 at Tahoe Donner.
TIME
HABIT
RIME
AGG.
COMMENTS
2100
dendritic, Nle, N la, some
gnropel
M-H
many
T= -1.2°C
2115
dendritic, N la and graupel
M-H
some
moderate to heavy rate
2130
same
M-H
few
T= -1.50C
2145
dendritic, some N le and few
graupel
L-H
few
heavy snow fall
2200
P ie, N le, N la, dendrites and
graupel
L-H
some
same
2215
same
L-H
many
T=-1.5°C, laige agg.
2230
same with few side planes
L-H
some
T=-1.5°C, heavy rate
0005
Pic, some graupel
L-H
many
T=-1.7°C, heavy rate
0015
Pic, seme graupel
L-H
many
T=-1.7°C, heavy rate
0030
Cpla, N le, P ic
L
many
T=-2.0°C, heavy rate
0045
SI, some N le
L
many
T=-2.0°C, heavy rate
0100
N le, SI
L
many
T=-2.0°C, heavy rate
0115
N le,
L-H
many ' T=-2.0°C, heavy rate
0130
SI, some Plb, some R4a
L-M
many
T=-2.1°C, heavy rate
0145
Pic, many SI
L
many
T=-2.6°C, heavy rate
0155
Cp3d
L
many
T=-2.6°C, heavy rate
0210
Cp3d, some Cpla
L
many
T=-3.0°C, heavy rate
0215
SI, N le
U
many
T=-3.9°C, heavy rate
0230
Cp3d, P la some
U
many
T=-4°C, heavy rate
0245
SI, some P ic
U
some
T=-4°C, heavy rate
0300
SI, some P7a, some C le
U
some
T=-4.2°C, heavy rate
0315
SI, some Pla
U
few
T=-4-5°C, moderate
0330
SI, few P2f
U
many
T=-4.7°C, heavy rate
0345
SI, few Pla
U
many
T=-5.0°C, moderate
0400
SI, few Pla
U
some
T=-5.0°C, moderate
0415
SI, some Cp2a, few C lc
U
some
T=-5.0°C, light
0430
some SI, few Cle, few Pla
L
some
T=-5.2°C, very light
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
124
4.6
SUMMARY AND CONCLUSION.
In this part of the dissertation, case studies of storms over the Sierra Nevada are
presented. These case studies are discussed mainly with respect to their SLW and ground
microphysics on both the upwind and downwind sides of the main Sierra Nevada barrier. Surface
wind observation at SM, soundings and precipitation amounts across the barrier are also used in
the analysis to augment the conclusions reached. Each storm has been used to show different
facts that are observed as storms cross a major mountain barrier. The results of this study and
discussions will mainly focus on the lee side observations. The case studies were selected from
the 1986/87 winter season inventory of the SCPP. The results are summarized below.
Four case study days; 18,19/20,22/23 December 1986 and 03 January 1987 were selected
for analysis. These storms were part of the 63 storms classified by Heggli and Rauber (1988),
hereafter referred to as HR. Two of these (18 and 22/23 December) storms were classified as
zonal showing a split flow in the middle troposphere associated with a dissipating cyclonic storms
(type A3); one (19 December storm) was meridional with a cuttoff low o r large amplitude
shortwave near 40°N (type B l); the fourth ( 03 January) was a zonal developing storm in a strong
southwesterly flow (type A l).
The storms of 18,19 and 22 December (type A3 and B l) showed good cross-barrier wind
direction (225°-250°). Also the movement of the low system in general eastward was very slow
due to a high pressure center located east of the sampling area. This slow movement was
responsible for the long periods of SLW observation. Frontal boundaries for these storms show
very weak temperature gradients.
The 03 January case, type A l storm, was very different from the type A3 and B l. The
frontal boundary in the soundings was easily identifiable from the strong temperature gradients.
At SM, the frontal passage was clearly visible from the change in wind speed and direction. The
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
125
low center moved east Caster than the other case studies and a good alignment of the winds for
cross-barrier SLW development were observed.
4.6.2
PRECIPITATION RATE.
Precipitation was recorded in the 18 December case from both pre- and post-frontal
clouds. Soundings indicate that both the 18 and 22 December (A3 type) storms had a seederfeeder type of clouds over KGV. For the 22 December case, a narrow band of precipitation fell
at the passage o f the front. Precipitation generally decreased behind the front where generally
subsidence and convective instability is large. For the 03 January case (A l type), no precipitation
was recorded after frontal passage at KGV. The precipitation for this case fell from a series of
bands. The largest of these bands coincided with the frontal band and resulted in the highest
precipitation rate for the day (9 mm/hr). Precipitation in the 19 December (type B l) case study
stopped behind the front in stations upwind of the crest but continued even after frontal passage
in the downwind side which was mainly due to the stagnation o f the low center over the area.
In all the storms, the effect o f topography was visible in the cumulative precipitation amount
plotted cross-barrier and around TRK.
4.6.3
RADIOMETER MEASURED SLW.
Temporal SLW o f the storms was measured at KGV and TRK. The measurements were
presented for each case study and discussed with respect to location of the cold front, wind
direction at SM and most importantly the ground microphysics measurements with respect to
frontal passage. The results at TRK show that the variation of SLW (temporal) is similar to that
observed at KGV where the highest values of SLW were usually found prior to frontal passage.
After the cold front, the SLW values fall sharply. It is important to note that this trend was
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
126
opposite that of the precipitation recorded. In all the storms, wind direction was from southwest
leading to a cross-barrier alignment with the Sierra Nevada and thus good SLW development.
The scan measurements made at TRK showed SLW values to be highest over the Sierra Nevada
and the Carson Range mountains. This was mainly due to SLW persisting over the main Sierra
Crest without total evaporation followed by re-development of clouds over the Carson Range.
The storm of 22 December showed the largest effect of re-development o f SLW over the Carson
Range mountains.
In Table 4.6.1 and 4.6.2, a summary of the SLW and Precipitable water vapor values
observed at KGV and TRK for all the storms is presented. Table 4.6.1 shows that for SLW
values of a 0.1 mm, the hours of observation at TRK were always less than at KGV. For 19
December and 22 December, the duration of SLW ( 2 0.1 mm) observed at TRK was 2 and 3
hrs compared to 15 and 16 hrs at KGV respectively. This seems to indicate that evaporation of
the SLW between KGV and TRK was very high for both case studies. But when the ground
microphysics and precipitation observation for these two cases are closely examined, the reason
for the reduction of SLW becomes apparent. For 19 December case study, the crystals at KGV
and TRK were unrimed and the cross-barrier precipitation was very low. Thus most o f the SLW
that was observed at KGV was most probably converted to vapor by evaporation on the lee side.
On the other hand, for 22 December case study, the crystals observed at both KGV and TRK
were heavily rimed (graupel) and also high precipitation amounts were observed in cross-barrier
stations.
This indicates that the SLW was efficiently being converted into precipitation.
Therefore, the reason for the highly reduced values and hours of observation of SLW at TRK
than at KGV for the two case studies was different. For 22 December, SLW value of 1 mm at
KGV was reduced to 0.3 mm (peak values) at TRK and high precipitation fell throughout the
stations. Also from Table 4.6.2, the 22 December case study shows the largest decrease in SLW
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
127
in crossing the Barrier with minimum of vapor increase. This is a direct confirmation that the
SLW did not get converted into vapor by evaporation but rather was efficiently removed from
the cloud as precipitation. The conclusion by H R that the continuous presence of SLW in
dissipating storms with low precipitation suggests inefficient precipitation forming processes from
measurements at a single station (KGV) does not agree with observations on the 22 December
case study. But the low precipitation rates, shallow clouds and high SLW observed on the 18
December case study at KGV and TRK suggest inefficient cloud processes in agreement with
HR.
In Table 4.6.2, the maximum and mean values of SLW and vapor recorded at both
stations in pre- and post-front clouds and the relative change in percent are presented. The SLW
maximum was always found to be in the pre-frontal clouds for all the storms sampled. A drop
in the maxima just after the frontal passage, observation of large vapor and small liquid values
at TRK and vice-versa at KGV are consistently observed in all the cases. The maximum SLW
measured at both stations and the percent change of both vapor and liquid are given in the
following table 4.6.2. From the table, the largest decrease in SLW in crossing the barrier with
relatively minimum amount o f vapor increase was seen in the storm of 22/23 December 1986.
This indicates that the SLW was removed efficiently across the crest.
The stage of evolution of the storms was found to control the SLW evolution over TRK.
This conclusion is similar to that of H R made at KGV. The dissipating storms showed SLW
throughout the lifetime of the storm over TRK and KGV where the greatest values were found
in the pre-frontal clouds. This conclusion was also found by HR for KGV records. The
observation of highest SLW in pre-frontal clouds was also true for the cuttoff low case (19
December). For developing storms, the highest SLW values were found in the post-frontal
region over KGV (e.g. 26 March case study in Part III and HR results). But over TRK, the 03
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
128
January case study which is a developing storm, shows below 0.05 mm values of SLW in the post
frontal region. The available data shows that a drop of SLW was observed just before the
passage of the cold front from about 0.1 mm to 0.05 mm. This does not fit the conclusions of
H R from observations in the upwind side of the crest.
Table 4.6.1. Summary of the duration(hrs) of radiometer measurements at TRK and KGV.
DATE
a 0.1 mm
a THRESHOLD (0.05 mm)
KGV
TRK
KGV
TRK
18 D EC (A3)
20
6
20+
23
19 D EC (B l)
15
2
21
22
22 D EC (A3)
16
3
19
9
03 JAN. (A l)
NO DATA ON THIS DAY
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
129
Table 4.6.2. Summary of the mean and maximum radiometer values at TRK and KGV.
PRE-FRONTAL
STORM
18 Dec.
19 Dec
22 Dec.
%
CHANGE
MEAN
KGV
TRK
LIQ.
0.23
0.13
VAP.
081
0.98
# PTS.
54
38
LIQ.
0.20
0.06
VAP.
0.71
0.94
# PTS.
56
40
LIQ.
0.53
0.10
VAP.
087
1.09
# PTS.
13
31
MAXIMUM
%
CHANGE
KGV
TRK
43.48
0.48
0.18
62.50
-20.99
0.95
1.16
-22.11
54
38
70.00
0.75
0.41
4583
-3289
1.02
183
-3089
56
40
81.13
1.03
081
69.90
-25.29
0.99
189
-40.00
13
31
POST-FRONTAL
18 Dec.
19 Dec.
22 Dec.
LIQ.
0.22
0.04
•18.00
084
0.12
77.78
VAP.
0.70
0.77
-10.00
086
086
0.00
# PTS.
40
22
40
22
LIQ.
0.08
0.06
25.00
0.19
0.11
42.11
VAP.
0.69
084
•21.74
085
0.91
-7.06
# PTS.
31
20
31
20
LIQ.
0.15
085
VAP.
0.78
1.27
# PTS.
71
4.63 GROUND CLOUD-MICROPHYSICS.
The ground ice crystal observations reflected the structural and synoptic difference
between the storms. The crystal habit varied dramatically from storm to storm and within a
single storm before and after frontal passage.
These habits are believed to reflect the
temperature and moisture conditions within the cloud where ice crystals grew. The A3 type
storms (18 and 22 December case study days) showed very s im ila r ice crystal habits and rim in g
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
130
tendencies even though the observations were done on different stages of the storm development.
In both cases, the ice crystals were heavily rimed, an indication of substantial SLW in the storms,
but relatively few aggregated. The crystals fell from a pre-cold frontal clouds for 18 December
case and post-frontal for 22 December case. In the 03 January case, the crystals were lightly
rimed but many aggregated and in 19 Dec. case unrimed crystals were observed due to large loss
o f SLW by evaporation. These different ground observations were a result of the different
synoptic nature o f the storms.
The crystals that fell on 18, 22 and 19 December cases were mostly originating at
temperatures less than -17°C with -12°c to -17°C being the dominant group. Soundings showed
that cloud top was the source of these crystals. As presented in Table 4.6.3, 85, 71 and 55
percent of the crystals observed were dendritic for 18,22 December and 03 January case studies
respectively. The habits were observed to change during storms, with larger amounts of warm
habits more prevalent at the beginning of a storm and larger amounts of colder temperature after
frontal passage. The frontal passage was also accompanied by changes in degree of riming. For
the 19 December case, most of the crystals were unrimed and cold habit at TD and moderately
rimed and warm habit at KGV. Recall also that the maximum observed SLW at TRK was only
0.18 mm and this case was the most affected by lee side evaporation. In the other case studies,
most o f the aggregation, riming and fragmentation was observed when crystals o f origin between 12°C and -1TC fell near the beginning of the sampling time, a pre-frontal region. Aggregation,
fragmentation and riming was mostly observed when dendritic crystals fell in all the case studies.
At TD, crystals tend to be more rimed when snowfall was light and at these times the
radiometer measured relatively higher amounts of SLW. When graupel was observed, the
radiometer indicated high convective type SLW. Heavy riming cases were usually associated with
low precipitation rates and high SLW which may reflect a relatively small number concentration
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
131
o f ice-crystals within the supercooled cloud. This does not agree with observations made on the
upwind side of the Sierra Nevada and those made by Feng and Grant (1982). Perhaps this is
because TD is downwind of the barrier. Observations similar to ours were reported in the
Cascades by Hobbs (1975). Aggregation of crystals is an important mechanism for particle
growth to precipitation size. The contribution of aggregation to the total snowfall mass was
significant. Contribution of riming, aggregation and fragmentation to the total snow fall is plotted
in Fig. 4.6.1 below (from a paper in preparation for publication by Demoz et al.). In this study,
the primary components of crystals that aggregated were planar dendrites and radiating
assemblages of plates. Simultaneous riming and aggregation were observed most of the time
during the sampling periods. The amount of riming associated with aggregation could be divided
into two categories. Heavy riming often accompanied by graupel where precipitation rates and
amount of aggregation were diminished and situations with a higher percentage of aggregates
characterized by moderate to light riming. Light riming did not appear to inhibit aggregation.
On 03 January, observation o f aggregation throughout the sampling period with surface
temperatures below -5°C indicated a "warm layer" at the crystal surface may not be necessary for
aggregation to occur. This confirms the arguments by Rauber (1987) and the findings o f Mitchell
(1988). More aggregation was always observed when precipitation was higher which was most
notable in the 03 January case. Sampling on this day was done only at TD. The timing was such
that the collected crystals represent the pre- and post-frontal clouds. A clear change of crystal
habit from warm to cold habit across the front was observed. On this day more aggregation,
heavy snowfall rate, less riming and less fragmentation showed a positive correlation as
mentioned earlier.
Riming was observed when SLW values were above threshold value (0.05 mm). The
degree of riming also increased as the SLW increased and was observed when planar crystals,
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
132
principally dendrites, fell. Cold crystals (columns and bullets) exhibited little riming (0 3 January
case). This is thought to be due to their compact shapes, which gives rise to a smaller collision
area and lower collision efficiency. In addition, these crystals usually were observed after frontal
passage when very low amounts of SLW (< 0 .0 5 mm) were observed.
An increase in fragmented crystals was usually accompanied by a decrease in aggregation
and increase in riming. At TD, generally less needle crystal concentration, higher aggregation
and less riming than KGV was observed. Cold habit crystals also appear more often at TD than
at KGV. This is clearly shown, especially in the 19 Dec.case study. In most of these cases,
especially the 22 December, needles where observed when riming was heavy.
In all of the storms, high mass percentage of the crystals that fell at TD comes from
rimed dendritic crystals. Lower amounts of needle crystals at TD than at KGV was observed.
This is due to higher cloud base at TD than at KGV. Crystal aggregation at TD was also higher
than at KGV and also favorable conditions for secondary ice multiplication processes oust more
at KGV than at TD. The table (Table 4.6.3) gives a summary of the contribution of each crystal
type by mass percentage at TD.
It is important to point out that the same data set was used in a study by Mitchell et al.,
(1 9 9 0 ). The work by Mitchell et al only dealt with relationships of m axim um crystal dimension
to mass and not in relation to storm structure, SLW, comparison to KGV measurements or any
other way discussed here. Therefore it can not be used as a basis for comparison. This kind of
work, that is comparing lee side observations of radiometer to the upwind side and at same time
using the microphysics to explain the measurements was not done before. This kind of study is
important to weather modification projects to understand which storms are naturally efficient and
which are not. This kind of knowledge will concentrate efforts to those storms which are
inefficient.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
133
Table 4.6.3. Summary of ice crystal observed by mass percentage at TD.
DATE
MASS (%)
CRYSTAL TYPE
18 D EC
22 D EC
WARM HABIT
2.4
HEAVILY RIMED DENDRITIC
8S.0
OTHERS
12.6
WARM HABIT
27.0
HEAVILY RIMED DENDRITIC
71.0
03 JAN.
OTHERS
2.0
WARM HABIT
9.0
DENDRITIC/PLANAR
55.0
SIDE PLANES
34.0
COLD HABIT
2.0
ICO.
75.
50.
25.
0
CD
£
S
loo.
CD
0.100
0.010
0.001
1
uj
I
0.100
0.010
0.001
I
I
I
t I I f I
.000
f—TTTTTJ
I— I
I
I IT T 'I
1 .000
r-i it i
1 1 1 1' ' 1
0.001
0.010
0.100
1.000
CRYSTAL MASS ( m g )
Figure 4.6.1. Cumulative mass percentage for overall storm as a function of crystal mass: total snowfall (1),
riming (2), aggregation (3) and fragmentation (4) for 03 January (A), 22 December (B) and 18 December (C).
(From Demoz et al. in preparation for publication)
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
134
PARTY
SUMMARY, CONCLUSION AND RECOMMENDATION FOR FUTURE WORK.
In the earlier chapters, an investigation of winter storms was presented using dual­
channel radiometer measurements of liquid and vapor, the stable oxygen isotopes of the
precipitation and ground measurements of precipitation rates and ice crystal habits. The storm
sampling was mainly done from two stations across the Sierra Nevada crest line. The first part
of this dissertation dealt with comparing the two radiometers used. The remaining chapters were
concerned with describing the structure and evolution of the winter storms and their
precipitation, as observed at an upwind and a downwind station. A summary of the main
conclusions reached are presented below.
I) RADIOMETER/RAWINSONDE - RADIOMETER COMPARISON.
Prior to using the radiometer measurements at two separate stations, an inter-comparison
of the USBR and D RI radiometers located side by side was necessary. Data from a storm on
5-6 December 1986, that was sampled by both radiometers located side by side at KGV, were
used. The DRI radiometer was being field tested at this time after its recent construction and
the instruments were not aligned to have the same averaging periods. Ground temperatures were
warmer than 0°C leading to wet snow conditions. In short, the experiment was not a carefully
designed inter-comparison but did provide an assessment of radiometer operation in typical
winter storm conditions. The situation described, where ambient temperatures at the site were
wanner than 0°C during precipitation periods, provided an opportunity to compare the
effectiveness of the two radiometers in obtaining "correct" measurements under much less than
optimum circumstances.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
135
Temporal records of the data showed periods during which melting snow or rain affected
the response of the USBR (fixed reflector) radiometer but did not appear to affect the DRI
(moving reflector) unit. The absence/presence of these "contaminated" periods for the USBR
unit, was mainly explained by the difference in the design of the radiometers.
Periods of
"contamination" lead to larger variability in the USBR radiometer data and lower correlation
coefficients between the data Grom the two radiometers. Correlation coefficients of 0.83 for the
liquid and 0.68 for the vapor values were found for radiometer-radiometer comparisons. When
some of the points suspected of "contamination" were removed the correlations improved to 0.87
and 0.71 for liquid and vapor values respectively.
The radiometers vapor channel outputs were compared to calculated rawinsonde
precipitable water values and 0.97 and 0.80 correlations were obtained for D RI and USBR
radiometers respectively. It should be noted that these comparisons were done for warm storm
conditions and may represent the worst scenario for the storms sampled and analyzed here in
1986/87. The main problem was "contamination" by wet snow on the USBR radiometer reflector.
As shown by Heggli et al. (1987) much better correlations are obtained when ground
temperatures are below 0°C.
All of the storms from which data was used in this thesis, were cold and the radiometers
were located at sites above the snow line in the Sierra Nevada. Hence the problems discussed
above are believed to haye had a minimum effect, enabling 80% correlation or better between
SLW and rawinsonde derived precipitable water to be achieved very easily. The data used in this
study were also checked routinely for wet weather or any other kind of "contamination".
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
136
II) STABLE OXYGEN ISOTOPES.
Observation of ice-crystal morphology of precipitation in the form of snow falling from
winter storms in the Sierra Nevada at KGV (upwind) and HM (downwind) were used, together
with stable oxygen isotope (S“0 ) measurements of the precipitation, to understand further the
storm structures. The observation sites were located at almost identical elevations on the upwind
and downwind sides of the crest. Short term variations of 5180 values observed during the 26
March 1985 storm were analyzed and compared with the soundings at KGV, radar reflectivity
over KGV, precipitation rate and radiometric measurements. A summary of the findings is as
follows.
A)
It was possible to separate this particular storm into three somewhat separate
mesoscale structure periods; a warm frontal period, followed by a transition period and then a
cold frontal period. These three different periods were characterized by distinct SlsO rates of
change and values.
B) The largest change with time in the data occurred during the warm frontal influence.
It increased from a minimum value of -24% o (-29% o at HM) to about -14% o in about five hours ( 2% o
change per hour). The main contribution to these changes came from the decrease in height of
the warm front with time. In the second stage, referred to as the "transition period", 5lsO values
decreased by 6% o from -14% o to -20% o ( -1.6% o per hour). This transition period was strongly
influenced by an elevated cold front (cold surge) which produced a brief minimum in the SlsO
values. In the cold frontal period, a steady increase from -21.5% o to -12% o (4.5% o pere hour) was
observed.
During this period the radiometer and radar indicated orographic lifting of the
convective type generating substantial supercooled liquid water (SLW) at low elevation producing
heavily rimed ice crystals with higher 5I80 values. It was also noted that the highest rate of
change of S180 occurred immediately following the cold frontal passage when the SLW reached
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
137
its peak value. The lowest rate of change occurred during the transition period.
C) Similar trends in the SlsO values were observed at HM as those at KGV. These
repeated characteristics at both stations suggest that the mesoscale structure of the clouds for this
period was well preserved as they crossed the mountain barrier. The data also show that
principal events, such as the elevated cold surge, had a more pronounced effect at HM than at
KGV. This pronounced change in 5180 at HM is believed to be due to the interaction o f the
topography and the frontal surface boundaries which are lifted as they crossed the crest.
D) Because the 5lsO values for the snow that fell at KGV and HM during the same time
intervals in the first half o f the storm differed by about 5%o, those a t HM being m ore negative,
it is concluded that the precipitation at HM contains a larger component o f ice from colder
regions o f the cloud mass. Precipitating crystals that originate higher in the cloud travel relatively
further horizontally than those originating at lower levels (mainly due to the lesser availability
o f SLW at higher levels,lower growth rate and higher winds). Also, it has been found (e.g.
W arburton and DeFelice, 1986, Reynolds and Kuciauskas, 1988) that most of the SLW in winter
clouds over this region is located and being captured. in the precipitation in the lower, warmer
levels o f the clouds.
The overall growth rates by riming, diffusion and aggregation of
precipitation particles to precipitable sizes, are faster at lower altitudes than at higher altitudes.
E ) Ice crystals at HM w ere o f colder origin than those falling a t KGV during the pre­
cold front period. The estimated tem perature difference for this period of ice phase precipitation
production varied between 4 and 7 centigrade degrees.
F) An analysis o f the 1984-85 and 1985-86 w inter seasons 6lsO values at the upwind site
(KGV) revealed that the ice phase w ater capture for snow peaked at -11 to -13°C ("equivalent
tem perature” derived from the isotope values).
G ) During the passage of the surface cold front and after, the differences in 5lsO at the
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
138
two sites were quite small probably because, during cold frontal passage, the effects from
convective processes (which are higher during this tim e) play a larger role in the precipitation
production process.
H)
Finally, the degree of riming, ice crystal origin, convective instability and location of
SLW in cloud systems are all im portant factors to consider when studying the stable isotopic
composition o f precipitation at a specific location. The isotopic composition o f the ice phase
precipitation is affected not only by the geographic location of the precipitating clouds, but by the
mesoscale structure within which the microscale processes producing the precipitation are
occurring.
H ence measuring the stable isotopic composition o f precipitation can yield
information about the main mesoscale and microscale processes occurring in a w inter storm.
I l l ) R A D IO M E T R IC L IQ U ID AND V A PO R O B SER V A T IO N S.
As described earlier, Heggli and Rauber (1988) classified a total o f 63 winter storms that
affected the central Sierra from 1983/84 through 1986/87 field seasons. The same data shows that
m ore than 60% o f the storms w ere zonal, 29% were meridional with 11% o f the storms being
unclassifiable. The table below shows the percentage o f storm types within the 1983/84 to
1986/87 field seasons.
Table 5.1 Statistics of storms affecting the Sierra Nevada.
#
%
DEVELOPING
OCCLUDED
DISSIPATING
14
12
12
22
19
19
TOTAL
38
60
CUT-OFF
NORTHERLY
7
11
11
18
TOTAL
18
29
7
11
STORM TYPE
ZONAL
MERIDIONAL
NEITHER
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
139
The case studies chosen for discussion here represent both zonal and meridional type
storms and thereby are representative of the majority of the storms that affect this area. In all
the storms presented here, wind direction at SM had a cross barrier component throughout the
observation periods which was ideal for the development and sustained presence of SLW.
Supercooled liquid water was present in all the case studies presented throughout most
o f the life times of the storms. Peak values o f SLW were always found near the frontal
boundary; in the leading edge of the front w here precipitation tends to be minimal. Only one
case, 26 March 1985 in Part III showed high SLW behind the cold front where subsidence
occurred and convective instability was large. A high degree of riming o f crystals at the ground
was observed during these peak SLW values. The mean value of SLW was greater in the prefrontal period for all the case studies presented here while peak values depended on the stage
o f evolution o f the storm. Above threshold (0.05 mm) values of SLW were observed, depending
on the storm , anywhere from 9 to 23 hours over TRK and for very much longer periods at KGV.
In the 26-27 March case o f a zonal and developing type of storm, post frontal regions and
convective cells at the end of the storm showed the largest SLW. This was similar to the findings
of Heggli and Rauber (1988) and Heggli et al.(1983). A similar observation was reported by
Sassen et al.(1990) and Long (1986) over the Tushar mountains of Utah. The occluded zonal
type storms have greatest SLW in the immediate post frontal region and in long convective
periods after the cold front has passed. In the zonal storm of the dissipating type (18 and 22
December case studies), SLW was present throughout the storm life time and was largest in the
pre-frontal stratus clouds.
In the 22 December case, SLW at TRK (downwind site) was very low compared to the
values observed at KGV. This storm, as suggested by Huggins et al. (1989), showed different
characteristics of SLW. The presence of very low SLW at TRK but almost the same vapor values
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
140
at both stations, suggests an efficient precipitation process at both sites.
The case of 19 December 1986, a meridional type, showed SLW influenced mainly by the
trajectory o f flow with peak values in the pre-frontal region.
The precipitation and SLW measurements show an inverse relationship as was observed
in many storms over the Park Range mountains in Colorado (Rauber et al, 1986), with high SLW
associated with shallow clouds.
The observation of high SLW values at leading edges o f bands near the front is
considered to be caused by enhanced convergence and relatively higher vertical velocities ahead
o f the front. These conditions and wind direction w ere not expected to change very much
between KGV and TRK. This would explain close connections between the largest SLW regions
within every storm that passed over both stations.
A frequency analysis of SLW measurements at KGV and TRK showed that values of
liquid were consistently larger over KGV than over TRK (vice versa for vapor values). This is
mainly due to the effect of compressional warming over the downwind valley. It is also noted
that the presence of SLW was always for shorter periods of tim e over TRK, perhaps due to the
effect o f evaporation of SLW on the downslope side o f the mountains.
Scan measurements of SLW over TRK show the eastward movement o f the clouds
towards the Carson Range mountains. The SLW that was advected eastward past KGV and
survived the forced descent from the main crest and the re-development o f liquid by orographic
lifting over the Carson Range appeared as peaks on the azimuth/time plots from TRK. To the
north of TRK, w here there is no orographic effect present, the SLW was consistently very low
as the air sank and warmed by compression.
Finally, a comparison of the increase in vapor and decrease in SLW at TRK with respect
to values at KGV reveals which storms had more efficient precipitation removing processes. In
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
141
the 22 December case the SLW showed a relative decrease o f more than 0.4 mm (from 0.53 at
KGV to 0.1 at TRK) while the vapor increased by only 0.13 cm (from 0.87 to 1.09). When data
from all the storms were compared, the 22 December case study was found to have the largest
decrease in SLW with relatively low o f vapor between KGV and TRK suggesting that it was the
most efficient storm. This storm was suggested to have an inefficient precipitation processes by
Heggli and Rauber (1988) from SLW measurements over KGV only. This is a very good way
o f dem onstrating the benefit and necessity o f sampling at an upwind and downwind stations if
correct evaluation o f efficiency o f precipitation processes is to be made.
IV ) G R O U N D BA SED M IC R O P H Y S IC S .
The passage o f a cold front across a sampling station was accompanied by changes in ice
crystal habits and degree o f riming. Observations show that warm habit crystals, mainly columns
and needles, fall in large numbers prior to cold frontal passage. Side planes and bullets increase
after the passage o f the surface cold front. A large decrease in the num ber of rim ed crystals after
cold front passage was a common observation in the storms which were studied.
Aggregation played a significant role in increasing the total mass of snowfall for example
on 03 January 1987. The major crystals that w ere observed to aggregate often have their origin
in the dendritic growth region (-10°C to -20°C). The aggregates observed contained centrally
located large crystals suggesting that aggregation for most part requires a large crystal capable
o f falling through a cloud containing sm aller crystals. For the case study o f 03 January 1987 (a
deep frontal system), aggregation was observed even when surface tem peratures w ere well below
-6°C. This suggests that "warm layer" on the surface o f ice crystals may not be necessary to form
aggregates.
During the shallow orographic cases (18,19,22 December) aggregation and precipitation
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
142
w ere observed to decrease when there was heavy riming. The decrease o f aggregation during
heavy riming conditions may be that the dendritic branches become filled with rime water reducing
the chance for entanglement.
The collection efficiency for dendritic crystals is enhanced by the "filter" effect of the
branches. It is also found that SLW is frequently found at elevations lower than the -10°C level.
The dendrites start riming below about the -13°C level. In fact, crystal sizes o f 0.5 - 0.8 mm, at
tem perature of -12°C, are often observed from aircraft. This size is much larger than needed for
the onset o f riming (0.1 - 0.3 mm according to Reinking, 1979). Therefore, since a high
proportion o f the crystals that fall in this region are dendritic, the contribution from riming to
the snowfall is very large. The location of SLW at low levels also allows warm habit crystals to
be rimed. Cold habit crystals, on the contrary are observed with very little riming mainly due to
their location with respect to the cold front. They fell after cold frontal passage when very low
SLW was observed. In studies that w ere done in the upwind side, SLW increases after frontal
passage where clouds are often shallow and hence riming increases. T he report by Feng and
G rant (1982) is an example of such an observation. But in this study, SLW was minimum after
cold frontal passage over TRK (a downwind station), and riming was less. Observations similar
to the present study w ere made by Hobbs(1975) in a downwind station over the Cascade
mountains.
The qualitative correlation of riming and fragmentation can be explained by large
branches of rimed dendrites breaking off from the main crystals. This process is enhanced also
by the increase in fall speed as the crystals get m ore and m ore rimed.
During one case study (the deep frontal system 03 January 1987), high concentrations of
warm habit crystals w ere accompanied by heavy riming. This may be an indication of a secondary
ice formation mechanism, such as the Hallett-M ossop o r so called ice-splintering process. H allett
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
143
and Mossop (1974) suggest that a relative velocity of colliding crystals and droplets of greater
than 0.7 m/s and a tem perature o f -5°C as appropriate conditions for the mechanism.
In Fig. 5.1 below, a conceptual picture of the clouds from a mature storm over this
region, as documented by the present and many prior studies, is presented. The cloud structure
and'processes depend as mentioned earlier on the position o f the fronts and can not be simplified
by a single figure. But, the model will help visualize some of the discussions and conclusions
reached in this study. It presents the general terrain profile in this region showing the main crest
and the Carson Range peak with distance from Sheridan, California on the x-axis and
tem perature scale (in °C) on the left and altitude (in km) on the right side y-axis. In presenting
the conceptual model picture, soundings, radar and satellite observations, ground and aircraft
measurements o f ice crystal habits and sizes are used. The cloud cover could vary from solid
overcast to scattered depending on the location and stage o f development o f the storm.
Two main tem perature regions are shown in the cloud. These regions are the diffusional
and the riming growth regions for dendritic crystals. Several crystal types have their origin in
these tem perature regions. Dendrites (-13°C to -17°C), radiating assemblages of dendrites (-15°C
to -22°C), crystals with broad branches (-13°C to -1TC), stellars (-13°C to -17°C) and assemblages
of plates (-18°C to -22°C) are some of the crystals that originate in this tem perature range. The
dendritic types are by far the main crystals observed to aggregate, rim e and fragment in high
proportions in the downwind region. These crystals originate at -13°C to -20°C tem perature
range. The main mechanism of growth for crystals in this region is assumed to be diffusional.
As they fall out of this region, they grow mainty by accretion until they reach the 0°C level and
start melting.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
-2 0
Ice Crystal (Dendrite) Nucleation Region
0.5 - 0.8mm D endrites'
Diffusional Growth
Region for Dendrites
-1 3 -
Air Sinking
1 - 4mm Dendrites
o
O
- (0
u
&
Ed
<
I-
Altitude km (MSL)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Cloud Top
II
Riming Growth
Region for
Dendrites
iegion
for SLW
Cloud Base
— I-------- 1---------1--------1-------- 1-------- 1-----
90
100
110
120
130
140
Distance from Sheridan (km)
Figure 5.1
A conceptual model of a m ature winter orographic cloud system over the Sierra
Nevada. See text for an explanation o f the different regions.
150
145
The figure shows regions where different microphysical characteristics are observed. A
combination o f different microphysical processes may act at a region at different times depending
on the stage o f the cloud development. In the following, only the main processes thought to be
im portant for the development of the observed characteristics will be mentioned.
D otted regions are regions where high concentrations of pristine crystals are observed
o r expected. This was shown by the modeling work o f Meyers and Cotton (1988) but no
explanation was given. The concentration is probably higher on the immediate downwind side
o f the crest because o f wind drift and the vertical velocity maximum region. Several results could
be mentioned which may be contributing to the observed high concentration. Time dependence
o f the ice forming processes lead to the increased ice crystal concentration as the cloud traverses
eastward. The cooling associated with cloud top lifting and evaporation as the cloud crosses the
barrier may activate tem perature dependent ice forming mechanisms that further increase the
concentration. Riming followed by splintering, mechanical breakup o f fragile crystals due to
enhanced turbulence near the barrier, and enhanced development o f ice in the strong uplift zones
preceding the crest, are also part o f the processes that may act in concert, o r independently,
resulting in the observed high concentration over the immediate downwind side o f the crest.
In the region represented by positively tilted lines, heavy precipitations are usually
observed. Observations made in SCPP show high radar echo return and graupel are centered
over this region. Aircraft measurements have shown up to 4 mm size dendrites at about the
-13°C level (top of this region) where aggregates begin to form. In the lower portions of this
region (at about the -5°C level) conditions are usually favorable for the Hallett-M ossop ice
multiplication processes. Heavy riming is observed. The condensation supply rate caused by
rising air is at its highest before it is cut-off by the barrier. Crystals at Kingvale are m ore rimed
than at Tahoe Donner. This is mainly because o f updraft just near the peak and m oisture input
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
on the windward side, for KGV, and the liquid losses due to compression in the lee side as well
as m oisture source cut-off to the lee side by the barrier, for TD. The rate of precipitation
increased as riming o f the particles increased over KGV. This was not shown to be true for
stations east of the main crest. In fact the riming was heaviest when the precipitation rate was
low. The riming mechanism is at its highest in the lowest kilom eter o r so above ground. This
region is labeled also as the main region of supercooled liquid w ater in Figure 5.1. This follows
the work of D eshler et al. (1990), Huggins and Rodi (1985), Heggli et al. (1983) and our ground
observations presented in the case studies here.
In Fig. 5.1, the vertical velocity maxima regions are indicated by negatively tilted lines.
Vertical velocities o f 0.5 m/s are usually observed in these regions. Observations and modeling
have confirmed these to be regions of high vertical velocity. These strong uplift regions also help
enhance the development o f pristine ice crystals which are then deposited downwind.
Aggregates, mostly of dendritic base, with some needles, are observed in the region
represented by vertical lines. Many aggregates o f dendritic crystals are observed across the
barrier.
As the figure shows, the dendritic crystals reach sizes favorable for riming and
aggregation even before they fall out of their diffusional growth layer.
The chance for
aggregation (for those crystals falling east o f the barrier) is enhanced even m ore due to the high
concentration of pristine crystals located in the path of the falling dendrites (heavy arrows).
Ground observations show that aggregation is m ore im portant in the precipitation development
than rim ing over the downwind stations.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
147
RECOMMENDATIONS FOR FUTURE WORK
(i) The microwave radiom eter is increasingly being used due to its importance in weather
modification and w ater budget studies. The use of multiple radiom eters for w ater and vapor flux
studies is shown to be im portant. But the use of such data depends on how the instruments
compare with each other particularly in their measurement of the integrated liquid water.
Comparison and performance analysis of the radiometers, when co-located is necessary in order
to validate measurements made when the instruments are separated. In this work such a
comparison was presented of the USBR and DRI radiom eters. Even though the comparison
shows good results, a well planned and careful inter-comparison of the instruments in different
w eather conditions has not yet been done. The results o f such an experiment are very important
to show that the instruments measure the same quantity over a long period o f tim e without large
errors. This kind o f comparison may also help prove convincingly the benefit o f the spinning dish
design.
(U) W e have shown the importance of stable isotope studies in identifying mesoscale and
microscale cloud/storm processes. The use o f stable w ater isotopes in weather, modification
studies was started very recently by Dr. Joseph W arburton o f DRI. Much of the work being done
was from data collected as part of the SCPP project. All the data sets available do not cover
entire duration o f storm periods, thus some o r all signatures of storm events are lost. A study
that is well planned to include upwind and downwind stations and entire storm durations would
be very helpful in understanding the precipitation chemistry and changes in the stable isotopes
o f the precipitation with different storm types and within a storm relative to the frontal surfaces.
A modeling o f stable isotope changes from soundings upwind to support the observations made
would also advance our understanding o f the frontal modification by a major barrier like the
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
148
Sierra Nevada. This kind of modeling that includes the downwind side o f the barrier when
coupled with the observations routinely made by DRI scientists would undoubtedly benefit the
quest for insight into ice-phase winter cloud microphysics and the mesoscale influence.
(ill)' As mentioned earlier, a large base of observational data exists on storms over the Sierra
Nevada as part of the SCPP and other projects. But large scale modeling of the microphysical
and frontal interaction have not yet been made. A simulation o f the barrier winds (Parish, 1982,
Waight, 1984 and Rasmusen et al., 1988) and a simple param eterization of the air flow (Elliot,
1981,1986 and Elliot and Rhea 1984) over the Sierra have been studied from the modeling view
point. A diagnostic method o f targeting during airborne seeding experiments o f the SCPP was
also described by Rauber et al. (1988). A part from these models, most of the SCPP efforts
concentrated on observations. To successfully conduct seeding operations, modelling on all scales
from microphysical to synoptic should be conducted. The upwind and downwind supercooled
liquid w ater measurements by the radiom eters and other observations would benefit from having
model-outputs as a basis for their analyses.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
149
REFEREN CE
Anthes, R.A., and T.T. W arner, 1978: Development of Hydrodynamic M odels Suitable for Air
Pollution and O ther Mesometreological Studies. Mon. Wea. Rev., 106,104S -1078.
Browning, K.A. and G.A. Monk, 1982: A Simple Model for the Synoptic Analysis o f Cold
Fronts. Quart. J. Roy. Meteor. Soc., 101, 893 - 900.
Choularton, T.W. and SJ . Perry, 1986: A Model o f the Orographic Enhancement o f Snowfall by
the Seeder-Feeder Mechanism. Quart. J. R. Meteor. Soc., 112, 335 - 345.
Clark, T.L., 1977: A Small Scale Dynamic Model Using a Terrain Following Coordinate
Transformation. J. Comput. Phys., 24,186 - 215.
Craig, H., 1961: Isotopic variations in meteoric waters. Science, 133,1702-1703.
Dansgaard, W., 1953: The abundance o f o18in atmospheric w ater and w ater vapor. Tellus, 5,461469.
Dansgaard, W., 1964: Stable isotopes in precipitation. Tellus, 16:436-468.
DeFelice, T.P., 1983: Location of liquid water in Sierra cloud systems by Physical techniques.
M.S. Thesis, University o f Nevada-Reno.
Demoz, B. B. 1989: A study of the ratio lsO/uO in relation to physical cloud characteristics in
precipitation from ice-phase winter storms over the Sierra Nevada. M.S. thesis,
University of Nevada, Reno
Demoz, B.B., J.A. W arburton and R.H. Stone 1991: The influence o f riming on the oxygen
isotopic composition of ice-phase precipitation. Atmos, res. 26,6, 463Demoz, B.B., J.A. W arburton and R.H. Stone 1988: The use of oxygen isotopes in studies of
ice-phase precipitation from w inter storms. 10* international cloud physics conference,
Bad Homburg, FRG. 15-20 August.
Deshler, T., and D.W. Reynolds, 1987: The Evolution of Artificially Nucleated Ice Crystals in
Shallow Orographic W inter Clouds: Two Case Studies. 11th Conf. on Weather
Modification, Edmonton, Amer. M eteor. Soc., 34 - 39.
D eshler, T., D.W. Reynolds and A.W. Huggins, 1990: Physical Response o f W inter Orographic
Clouds Over the Sierra Nevada to Airborne Seeding Using Dry Ice o r Silver Iodide. J.
AppL Meteor., 29, 288 - 330.
Dirks, R.A., 1973: Precipitation Efficiency o f Orographic Clouds. J. Rech. Atmos., 7,177 -184.
Elliot, R.D. and J.D . Rhea, 1984: Comparison of Observed and Predicted M odel Winds Over an
Orographic Barrier. Proc. 9* Weather Modif. Conf., May 21 - 23, Park City, Amer.
M eteor. Soc., 83 - 84.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
150
Elliot, R.D., and E.L. Hovind, 1964: The W ater Balance of Orographic Clouds. /. Appl. Meteor.,
235 - 239.
Elliott, R.D., 1986: Review of W intertime Orographic Cloud Seeding.
Precipitation
Enhancement - A Scientific Challenge, R.R. Braham, Jr., Ed., M eteor Monogr., 21,
Amer. M eteor. Soc., 87 -101.
Federer, B., N. Brichet, J. Jouzel, 1982: Stable isotopes in hailstones. Part I: The isotopic cloud
model. /. Atmos. ScL, 39,1324-1335.
Fujiyoshi, Y., G. Wakahama, 1986: Short-term variations o f oxygen isotopic composition of
falling snow particles. TeUus., 388,353-363.
Furukawa, Y., M., Yamamoto, and T. Kuroda, 1987: Ellipsometric Study o f The Transition Layer
on the Surface o f an Ice Crystal. /. Cryst. Growth, 82, 665 - 677.
Gedzelman, S.D., and J. R. Lawrence, 1982: The isotopic composition of (yclonic preciptation.
/. Appl. Meteor., 21,1385-1404.
Gedzelman, S.D., 1988: D euterium in w ater vapor above the atmospheric boundary layer. Tellus,
40b,134-147.
Gedzelman, S.D., J. M. Rosenbaum and J. R. Lawrence, 1989: The megalopolitan snowstorm of
11-12 February 1983: Isotopic composition of the snow. J. Atmos, set, 46,1637-1649.
Gordon, G.L., and J.D. Marwitz, 1986: Hydrometeor Evolution in Rainbands Over th e California
Valley. J. Atmos. ScL, 43,1087 - 1100.
Guirand, F.D., J. Howard and D.C. Hogg, 1979: A dual-channel microwave radiom eter for
m easurement o f precipitable w ater vapor and liquid. IEEE Transactions on Geoscience
Electronics, G E -17,129-136.
H allett, J. and J.W. Telford, 1981: A case study o f ice and supercooled w ater distribution in a
Sierra storm: Hypothesis and implication for ice phase evolution and seeding. Reprint
from 8th Conference on inadvertent and planned weather Mod., Reno, NV, AMS.
H allett, J., 1987: Faceted snow crystals. J. Opt. Soc. Am, 4, 581-588.
Heggli, M .F., L. Varidman, R.E. Stewart, A Huggins, 1983: Supercooled liquid w ater and ice
crystal distributions within Sierra Nevada w inter storms. /. Climate and A ppl Meteor., 22,
1875-1886.
Heggli, M.K., and R.M. Rauber, 1988: The characterstics and evaluation o f supercooled water
in wintertim e storms over the Sierra Nevada: A summary o f microwave radiom eter
measurements taken during the Sierra Cooperative Pilot Project. /. Appl Meteor., 21,9891015.
Heggli, M.F., R.M. Rauber and J.B. Snider, 1987: Field Evaluation of a Dual-Channel Microwave
Radiom eter Designed For M easurements of Integrated W ater Vapor and Cloud Liquid
W ater in the Atmosphere. J. Atmos. Oceanic TechnoL, 4, 204 - 213.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
151
Heggli, M., and D.W. Reynolds, 1985: Radiometric Observations of Supercooled Liquid W ater
Within a Split Front Over The Sierra Nevada. /. Climate Appl. Meteor., 24,1258 - 1261.
Heymsfield, A., 1982: A Comparative Study of The Rates of Development o f Potential Hail
Embryos in the High Plains Storms. /. Atmos. ScL, 39, 2867 - 2897.
Heymsfield, A., 1987: Aggregates as Embryos in Seeded Clouds. Precipitation Enhancement A Scientific Challenge. M eteor. Monogrs. No. 21 and No. 43, Amer. M eteor. Soc., 33 41.
Hill, G.E., 1986: Seedability o f W inter Orographic Clouds. Precipitation Enhancement - A
Scientific Challenge, R.R. Braham, Jr., Ed., M eteor. Monogr., 21, Amer. M eteor. Soc.,
127 -136.
Hobbs, P.V., 1978: Organization and Structure of Clouds and Precipitation on the Mesoscale and
Microscale in Cyclonic Storms. Rev. Geophys. Space Phys., 16, 741 - 755.
, R.A. Houze, Jr., and T J . Matejka 1975: The dynamical and microphysical structure of an
occluded frontal system and its modification by orography. J. Atmos. ScL, 32,1542-1562.
Huggins, A W .,R .J. M eitin and R.F. Reinking, 1990: Comparison of radiom etric measurements
of cloud liquid w ater near a mountain crest and over a downwind valley. Fifth conf. on
Mountain Meteo., Boulder, Colorado.
Huggins, A.W., J.O . Rhea, G.L. Hemmer, A.P. Kuciauskas, and C.J. Wilcox, R.D. Elliott and
D.A. Griffith, and J.A. Flueck, 1985: Interim Progress Report, Vol. I, for the period
Sept. 1,1984 through Aug. 31,1985. Prepared for Div. o f Atmos. Resources Research,
U.S. Bureau of Reclamation Under Contract 4-CR-81-03860. P.O. Box 25007, Denver,
CO 80225.
Huggins, A.W., and A.R. Rodi, 1985: Physical Response of Convective Clouds Over The Sierra
Nevada to Seeding With Dry Ice. J. Climate AppL Meteor., 24,1082 - 1098.
Huggins, A.W., R.W. Rauber, T. W. Lee, A P . Kuciauskas, G.L. Hemmer and C.J. wilcox, 1986:
SCPP Meteorological and Statistical Support, Intrim Progress Rep. 1985 - 1986.
[Available from Bureau o f Reclamation, Division of Atmospheric Resources Research,
D • 1220, Denver Federal Center, Denver, CO 80225.]
Interim Program R eport, SCPP M eteor, and Stat. Support. 1 September 1984 through 31 August
1985, Volumes I and II. Contract No. 4-CR-81-03860. United States Departm ent o f the
Interior, Bureau of Reclamation.
Jouzel, J., and L. Merlivat, 1984: Deuterium and oxygen-18 in precipitation: Modeling of the
isotopic effects during snow formation. J. Geophy. Res., 89,11749-11757.
Jouzel, J.„ G. L. Russell, R. J. Suozzo, R.D. Koster, J.W.C. W hite, and W.S. Broecker, 1987:
Simulations of the HDO and H2180 atmospheric cycles using the NASA GISS General
Circulation Model: The seasonal cycle for present-day conditions. J. Geophys.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
152
Res. ,92,14739-14760.
King, W.D., 1984: Seeding Stratiform Clouds: The Effects of Cloud Param eters on Aiming. /.
Climate Appl Meteor,:, 23,1265 - 1272.
Lew, J.K., D.C. Montague and H.R. Pruppacher, 1986: A Wind Tunnel Investigation on The
Riming o f Snowflakes. Part II: Natural and Synthetic Aggregates. J. Atmos. ScL, 43,2410
-2417.
Liou, Kuo-Nan, 1980: An Introduction to Atmospheric Radiation. Acad. Press, Inc., SanDiego,
California. Page 276.
Ludlam, F.M., 1955: Artificial Snowfall From M ountain Clouds. Tellus, 7, 277 - 290.
M artner, B.E., 1986: Microphysical Seeding Effects in Non Convective Clouds Over The Sierra
Nevada. 70* Conf. on weather Modification, Arlington, Amer. M eteor. Soc., 30 - 33.
Marwitz, J.D ., 1980: W inter storms over the San Juan mountains. Part I: Dynamical processes.
J, A ppl Meteorol 19, 913 - 926.
Marwitz, J.D.(1983): The kinematics of orographic airflow during Sierra storms. J. Atmos. ScL
40,1218 -1227.
Marwitz, J.D ., 1987: Deep Orographic Storms Over The Sierra Nevada. Part I: Therodynamic
and Kinematic Structure. J. atmos. ScL, 44,159 - 173.
Meyers, M.P. and W.R. Cotton, 1988: A Numerical Investigation o f An Orographic Precipitation
Event. Preprints o f 70* International Cloud Physics Conference, 15 - 20 August 1988, Bad
Homburg, FRG.
Mitchell, D. L., R. Zhang and R. L. Pitter, 1990: Mass dimensional relationships for ice particles
and the influence o f riming on snowfall rates. /. Appl Meteor., 2,153-163
Mitchell, D.L., 1988: Evolution of Snow - Size Spectra in Cyclonic Storms. Part I: Snow gGrowth
by Vapor Deposition and Aggregation. /. Atmos. ScL, 45, 3431 - 3451.
Mogono, C., and C.W. Lee, 1966: Meteorological Classification o f Natural Snow Crystals. J. Fac.
ScL Hokkaido Univ. Ser. 7, 2, 321 -362.
Nakaya, V., 1954: Snow Crystal, Natural and Artificial Harvard University Press.
Ono, A , 1969: The shape and riming properties of ice crystals in natural clouds. /. Atmos. ScL,
26,138-147.
Parish, T.R., 1982: Barrier Winds Along The Sierra Nevada M ountains. J. Appl Meteor., 21,925
-930.
Peterson, T.C., L.O. Grant, W.R. Cotton and D.C. Rogers (1991): The effect o f decoupled low-
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
153
level flow on w inter orographic clouds and precipitation in the Yampa River Valley. J.
Appl. MeteoroL, 30, 368 -386.
Picciotto, E., X. D e M aere, and L. Friedman, 1960: Isotopic composition and tem perature of
formation o f antarctic snows. NATURE,181, 857-859.
Prasad, N., 1986: Percipitation Development in Seeded Clouds over The Sierra Neada. Rep.
# A5152, D ept, o f Atmos. Sci., University o f Wayoming, Laramie, WY 82071. Contract
2 - 07 - 81 - V 0256, Bureau o f Reclamation.
Rangno, A.L., 1986: How good are our Conceptual Models o f Orographic Cloud Seeding.
Precipitation Enhancement - A Scientific Challenge, R.R . Braham, Jr., Ed. M eteor.
Monogr., 21, Amer. M eteor. Soc., 115 -126.
Rasmussen, R.M., P. Smolarkiewicz, W.D. Hall and T. Clark, 1988: Comparison o f Orographic
Cloud and Precipitation Development of Three D ifferent Geographic Locations Using
a 3 - D nested G rid Model. Preprints, K f1international cloud phusics conference, August
1988, Bad Homburg, FRG.
Rauber, R.M., and L.O. G rant, 1986: The Charactersitics and D istribution o f Cloud W ater over
the M ountains o f N orthern Colorado During W intertim e Storms. Part II: Spatial
Distribution and Microphysical Characterstics. J. Climate A ppl Meteor., 25, 489 - 504.
Rauber, R.M., R.D. Elliot, J.O . Rhea, A.W. Huggins and D.W. Reynolds, (1988): A Diagnostic
Technique for Targeting during Airborne Seeding Experiments in W intertime Storms
over the Sierra Nevada. J. A ppl Meteor., 27, 811 - 828.
Rauber, R.M., 1987: Characterstics o f Cloud Ice and Precipitation During W intertime Storms
Over The Mountains o f N orthern Colorado. J. Climate and A ppl Meteor., 26,488 - 524.
Reinking, R.F. and R.J. M eitin, 1989: Recent progress and needs in obtaining physical evidence
for weather modification potentials and effects. J. Wea. modif, 21,85-93.
Reinking, R.F., 1979: The Onset and Early Growth o f Snow Crystals by Accretion o f Droplets.
J. Atmos. ScL, 36, 870 - 881.
Reynolds, D.W., and A.P. Kuciauskas, 1988: Rem ote and in site observations of Sierra Nevada
winter mountain clouds: Relationships between structure, precipitation and liquid water.
J. A ppl Meteor., 27,140-156.
Reynolds, D.W., A.S. Dennis, 1986: A review o f the Sierra Cooperative Pilot Project. Bull
Amer. Meteor. Soc., 67, 513-523.
Rindsberger, M. and Sh. Jaffe, 1989: Oxygen isotopic variation o f falling snow particles with time
during the lifetime of a convective cloud: observation and modelling. Tellus, 41b, 511523.
Rindsberger, M., Sh. Jaffe, Sh. Rahamim and J.R. Gat, 1990:
Patterns o f the isotopic
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
154
composition o f precipitation in tim e and space: data from the Israeli storm w ater
collection program. Tellus, 42b, 263-271.
Rodi, A.R., A.S. Heymsfield and N. Prasad, 1985: Precipitation Form ation in Dry Ice Seeding
Plumes. 4* WMO Conf. on Weather M odif, Honolulu, WMO, 71 - 76.
Schwerdtfeger, W. (1979): M etreological aspects of the drift of ice from W eddell sea toward the
mid-latitude westerlies. /. Geophy. Res. 84, 6321 • 6327.
Scorer, R.S. and H . Klieforth (1959): Theory o f mountain waves o f large amplitude. Quart. J.
R. MeteoroL Soc. 85,131 - 143.
Scorer, R.S. (1949): Theory o f waves in the lee of mountains: II The flow over a ridge. Quart.
J. R. MeteoroL Soc. 79, 70 - 83.
Scorer, R.S. (1954): Theory of airflow over mountains: III - Airstream characterstics. Quart.
J. R. MeteoroL Soc. 80, 417 - 428.
Smith, R.B., 1982: A Differential Advection M odel of Orographic Rain. Mon. Weather Review,
110, 306 - 309.
Smith, R.B., 1979: The Influence of Mountains on the Atmosphere. Advances in Geophysics,
Vol. 21, Academic Press, 87 - 127.
Smith, R.B., and Y-L. Lin, 1982: The Addition of H eat to a Stratified Airstream With
Application to the Dynamics of Orographic Rain. Quart. J. Roy. Meteor. Soc., 108,----- .
Smith, G.L., L. Friedman, H . Klieforth and K. Hardcastle, 1979: A real distribution of Deuterium
in eastern California precipition, 1968-1969. J. AppL Meteor., 18,172-188.
Snider, J.B.,F.O. Guiraud and D.C. Hogg, 1990: Comparison o f cloud liquid content measured
by two independent ground-based systems. J. Geophys. Res.,71,2875-2881.
Stewart, R.E., and J.D . Marwitz, 1982a: The Downwind Spread of An Initially Vertical Column
of Particles in a Sheared Environment. J. AppL Meteor., 21,1191 -1193.
Stewart, R.E., and J.D . Marwitz, 1982b: Microphysical Effects o f Seeding W intertime Stratiform
Clouds N ear The Sierra Nevada M ountains. /. Appl Meteor., 21,874 - 880.
Sugimoto, A., K. Higuchi and M. Kusakabe, 1988: Relationship between D and lsO values o f
falling snow particles from a separate cloud. Tellus, 40B, 205-213.
Super, A.B. and E.W. Holroyd, 1989: Temporal variations of cloud liquid w ater during winter
storms over the Mogollon rim of Arizona. J. Wea. Modif. ,21,35-40.
U ttal, T., R.M. Rauber and L.O. Grant, 1988: Distribution o f Liquid, Vapor and Ice in an
Orographic Cloud From Field Observations. /. Atmos. ScL, 7,1110 - 1122.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
155
Varidman, L., 1978: The G eneration of Secondary Ice Particles in Clouds by Crystal-Crystal
Collision. /. Atmos. ScL, 35, 2168 -2180.
W arburton, J.A., and T.P. DeFelice, 1986: Oxygen isotopic composition o f Central Sierra
Nevada precipitation, I. Identification of ice-phase water capture regions in winter storms.
Atmos. Res., 20,11-22.
, B.B. Demoz and R.H. Stone 1992: Oxygen isotopic variations of snowfall from winter
storms in the central Sierra Nevada. Relation to ice growth microphysics and mesoscale
structure, submited to Atmos. Res.
, R.H. Stone and B.B. Demoz 1988: A conceptual model for determining seeded/notseeded ratios in weather modification experiments based on stable oxygen isotopic ratios
and snow chemistry. 5th WMO Scientific conference on W eather Modification and
Applied Cloud Physics, Beijing, China, 15-20 May
Weight, K.T., III, 1984: A Numerical Study of The Sierra Nevada Barrier Jet. Atmospheric
Science Paper No. 148, University of Wyoming, Laramie, WY.
Westwater, E.R., and M.T. Decker, 1977: Inversion methods in atmospheric rem ote sounding,
A D eepak ed., Acad. Press, Inc.,pp395-425.
Westwater, E.R., and F.O. Guiraud, 1980: Ground-based microwave radiometric retrieval of
precipitable water vapor in the presence of clouds with high liquid content. Radio
ScL,15,947-957
W estwater, E.R., 1978: The accuracy o f w ater vapor and cloud liquid determination by dual­
frequency ground-based microwave radiometty. Radio ScL,13,677-685.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
156
142.98
128.68
114.38
100.09
85.79
Vo
71.49
§
\
j Lake
5L\ Tahoe
57.19
42.89
28.60
Placervilie
■ < su
14.30
0.00
0.00
14.05
28.10
42.15
56.21
70.26
84.31
98.36
112.41
126.46
Figure A l. Contour map of experiment area showing precipitation sampling stations. Contour lines are
drawn every 200 meters.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
at*
I
O^ndl* of
* !a * * n ta rt naadlaa
i
&
Solid trite* p lo t*
<
rpd
O an d rltlt c ry a ta l
• th a a c te r llk a and*
1
•p*
era
C onotnatlon
of n iid ln
Combination
o f eulw no
0
•it
cu
tjrm Id
He
CIO
Me
S olid o u iia t
cid
HoiIcm O ull«t
tti
Cl*
0
><
£
as
of
’. 0
1
*
0
0
3$e
Pit
P lat* Mtth
t a r t o r ll b a a ita n a lrn *
PP«
P lat* a l l h
d a n d r lt u «it*n*lona
f ia
« )—
Tvr.orancnad
n ry a ta l
e-
A
Mb
Thr**.Br*nr
er*»**i
"*1
1
P it
F«e
C ryacal Mlth
proad branch**
r o g r- b ra r r-* l
cr»»t*l
f-.a
rid
tre a d tran r* 'r e n a l
Mltn | / e r a r c 'a t
p i*
C rtflnarr a * n a rltte
e rra ta ;
rir
» * rn lis* < r)* t« l
PJ a
St**.lar e ra a ta i
• I f 91 at* r at *na*
&
%
s
■ft
=£
PP*
P la ta Mlth
o tapl* * it*nalena
ri*
Haaatonal p lata
C rcatal * ltn
l a c to r iia * branch**
Combination qf
lo in s o lid CO'.-JMO
Cup
n
S croll
Combination
o f b u lla ta
i
0
cu
Ci#
at*
lone i s l l d (flo w i
<£>
°T °
TMeo p U to
o r aaolton te i*
*>b
v$M
O a n d rltlt f M t i a l
«lti> p la ta a a t tnda
■ tia in tv r itn ith
*i«abSn»tlan
o f inaatna
P lat* Mlth
a p a tlo l dandrita*
PPc
°4 * ?
Cln
B undla o f
• iM M tu y i b i it h i
X
X
$
3
Cl*
n«
•U
X
rrt
S ta lla r c r f t u i
Mlth a a e to rtla * ana*
0
aio
Pd»
C lf
Hollo* eo|«an
A
l
%
¥
<^>
CPM
P la ta a t t h
a c ra lla a t *nda
Pa*
SI
S ia t la r cry a ta l
a t t h a p a tta l p la ta a
S lia planaa
PM
sp
S ta tic # c r y a ta l
* t t h a p a tla l dandrita*
S ealallba
atda planaa
s«
n*
( M l a a t l t n of aid*
e lc n aa. b u lle ts
and coitana
N pdlatla# a**a*bl*|*
a t plot**
P7h
• a d la tia j i i i i a » l i |i
o f d a a d rllia
CPI*
C a lw « t u p ia u a
CPI*
< * !■ * Mlth dandrita*
CPI*
1
1
0
au
aiM d naadla
cryacal
Nib
iU a d c a lim a r
c ry a ta l
•tc
S
*
fit*
*»b
M *nirltlc :r » » ta l
• Itn i • branrnaa
It*
Irokan eranot
P
Mb
MlMd aroaan branch
•Pa
lu
Oansslv riaa d
M lacalianaaua
S|
■ Inuta c e l w t
cpya
Ub
S to lla r c rv a ta i
Mlth naadlaa
S ta l la r c r y a ta l a m
ria a d a p a tla l t m m has
CPJ*
d
i p
m
CPJ«
p
IP
•la** p art lei*
IA
S ta l la r c r y a ta l
Mlth c a l iM i
H
ll
(ca p a r ttc l*
Dana*la H ead
s t a l l a r a ry atal
C P»
at
•
N.«
Conaltka irtu o a l
to ile t
M lth d aa d rlta*
* tir:r* « a r r r a ta l
Poa
Lu m frOMpal
a
vV
&
i
•A*
Hatayenal graupel
Q
«lc
•to ad o t a l l t r
c r y s ta l
-ra u a a lltb * aoe» « u n
noerlaad *at*nalen*
o
blnad
P la ta e r aa eto r
•Id
ttd lttp lP
IllM
CPU
Pub
P lat* - I f
•3«>l*l slo t* *
O
S ta i’ia r cry a ta l
• i t n t d r a lla a t oms
•la
C ravpalllb* ana*
o f hacaranal typa
•1*
Orawaalltk* p a *
*r la a In *
*
0
*
*
cp
Cam o f aaalton rer»
St
■inut* naaadonal p lat*
ca.
Hlnuta a t a l l a r c ry a ta l
is
■Inuta aa**abl*«#
of e la ta a
Ob
I r r a i u i a r asim
157
Figure A2. Mogono-Lee classification o f natural snow crystals.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1086 to
1711 ( g m l ) 18 d e c
ICAUIAI. H O T 3 . OKI'.' Hi. I'Jllii
RADIAL PLOT 2 . DEC. 16. 1906
RADIAL PLOT I . DEC. 16. 1966
16 5 0 U m l> 16 d e c
10B6
U »»«
*B
*9BB
U m l)
18 d e c
1966
1911
1U d e c
19HU t o
1921 ( g n . l j
Hi d e c
1966
N
i
Wt
Vi
i
0 . }0 IT}I
RADUL PLOT 4 . DEC. 18. 1988
1 9 4 6 ( g r o t) 18 d e c
1986 to
1059 ( g m t ) 18 d e c
RAOIAL PLOT 5 , DEC. 1 8 . 1966
1086
2001 ( g i n t ) IB d e c 1 0 8 6 l o 2 0 1 4 ( g n i t ) 18 d e c
RAlllAI. Pl.OT 6 , DEC. Hi. I9U 6
1086
2 8 4 2 ( g l u t ) 16 d e e
1 DUlJ l o 2 6 0 4 I g m lJ
18 d e c
I9 6 0
N
-»•
Vi
Q.[0 n^i
158
Figure A3. V ector plot o f the radiom eter measured integrated cloud liquid w ater value over TRK on 18
December 1986.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
RADIAL PLOT 8 . OEC. 16, 1 9 8 6
RADIAL PLOT 7 . DEC. 18. 1966
2 0 5 7 ( g r o t) 18 d e c
1 6 8 6 l o 2 1 0 9 ( g m l ) 18 d e c
1986
2 1 1 2 ( g m l ) 16 d e c
1 9 6 6 ( o 2 1 2 4 ( g m l ) 18 d e c
RADIAL PLOT 9 . D l l '.
1986
2 1 2 7 ( g m l ) 18 d e c
|U .
I9UU
I9UB l o 2 1 1 0 ( g m l )
111 d u e
1986
■ft 1---*
° 1° ilT
RADIAL PLOT i l . DEC. 18. I9UQ
RADIAL PLOT 10. DEC. 16. 1966
2 1 4 2 ( g m l ) 16 d e c
1 9 8 6 l o 2 1 5 5 ( g m l ) 18 d e c 1986
2 2 0 5 ( ( m l ) 18 d e c
1980 lo 2 2 1 ? ( ,m l )
IB d e c
RADIAL PLOT 11!. DEC. 18. I08U
1980
“ ~u
18 d e c
1 0 0 0 t o 8 2 8 8 (Ci u l )
IB d e c
1086
III
Figure A3, (continued)
159
d ec
|9 0 6 t o 2 3 4 8 ( g m l ) 18 d e c
I9H 6
2250 U m l)
IB d e c
N
1988 ( o 2 3 0 3 ( g m l ) 18 d e c
1988
2305 (g m l)
N
IB d e c
I'JUO l o 23111 ( g m l )
18 d c i
IU0U
N
Ri
Vi
i
—
0 .1 0
CUMULATIVE RADIAL PLOT; DEC. 18. 1 9 8 6 (1 5 SCANS)
1 6 5 9 ( g m l ) 18 d e c 1 9 8 6 t o 2 3 1 9 ( g m l ) 18 d e c
1986
N
Vi
Figure A3, (continued)
160
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 2 J 5 (J m i j | 0
RADIAL PLOT 15. DLL'. JU. 1980
RADIAL PLOT 14. DEC. 1980
RADIAL PLOT 13. DEC. JB. 1988
Figure A4.1. Surface weather map and 500 mb layer heights at 1200 UTC, 19 December 1986.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
162
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
163
3 0 0 » M IL L I M It H EIG H T C ONTOURS
AT 7 : 0 0 A .M ;, S .3 .T .
Figure A4J3. Same as Fig. A4.2, but for 03 January 1986.
Reproduced with perm ission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
9 .0
8. 0
7 .0
Mo Data
2.0
T e rra in
0. 0
a *s.
•**
•H
09
06
03
Time (GMT)
Figure AS. Radar echo contours and precipitation rate values at KGV, on 03 January, 1987. Radar data
are contoured in 5 dBz increments beginning with 10 dBz. (from Hemmer et al.,1986)
164
Документ
Категория
Без категории
Просмотров
0
Размер файла
7 093 Кб
Теги
sdewsdweddes
1/--страниц
Пожаловаться на содержимое документа