close

Вход

Забыли?

вход по аккаунту

?

610

код для вставкиСкачать
INTERNATIONAL JOURNAL OF CLIMATOLOGY
Int. J. Climatol. 18: 1273–1284 (1998)
SPATIO-TEMPORAL VARIATIONS IN THUNDERSTORM RAINFALL
OVER NIGERIA
IBIDUN O. ADELEKAN*
Department of Geography, Uni6ersity of Ibadan, Ibadan, Nigeria
Recei6ed 21 February 1997
Re6ised 16 February 1998
Accepted 16 February 1998
ABSTRACT
Thunderstorm rainfall is a major form of convective rainfall over Nigeria. Its contribution to total wet season rainfall
increases from about 18% in the south to 36% in the north. The importance of thunderstorms lies in the fact that they
contribute significantly to flooding episodes and soil erosion processes due to their high intensity and torrential
characteristics.
This paper examines the spatial and temporal variations in thunderstorm rainfall over Nigeria using daily rainfall
data and associated weather information over a 30 year period (1960 – 1989) for 19 synoptic stations. Results show
that the dry season distribution pattern of rainfall due to thunderstorms shows a general decrease from south to north
similar to the total rainfall distribution pattern in the country. At the peak of the wet season (July – September)
thunderstorm rainfall increases northwards up to around latitude 11°N when it begins to decrease. The effect of
orography in enhancing thunderstorm rainfall across the country is also shown. Elementary linkage analysis was used
to group the 19 synoptic stations into thunderstorm rainfall regions. Six regions were identified which showed internal
coherence in terms of temporal fluctuations of thunderstorm rainfall. © 1998 Royal Meteorological Society.
KEY WORDS: Nigeria;
thunderstorm rainfall; synoptic climatology; elementary linkage analyses
1. INTRODUCTION
Thunderstorm rainfall due to local convection is one of the three major rainfall types experienced in
Nigeria and West Africa at large. The other two types are squall line rainfall and ordinary monsoon
rainfall due to large scale convergence (Kamara, 1986; Desbois et al., 1988). This classification is based
on the synoptic origin of rainfall. Thunderstorms are an elementary unit of larger tropical disturbances
and produce a large proportion of the rainfall in most tropical areas. In Nigeria, the percentage
contribution of thunderstorms to total wet season rainfall increases significantly from the coastal areas
inland, with the coastal, middle belt and northern sections of the country recording 17.8%, 30.4% and
35.8%, respectively.
Thunderstorms are highly localized and largely stationary weather systems affecting a limited area of
about 20–50 km2, depending on the size of the cumulus tower. They are associated with shower clouds
in which electrical discharges can be seen as lightning and heard as thunder on the ground, and they
represent an advanced stage in the development of convection in moist air. The importance of rainfall
generated by thunderstorms lies in the fact that it is largely torrential and of high intensity, and as a result
much is lost as runoff which causes flooding.
Most of the existing studies on Nigerian rainfall have concentrated on the spatial and seasonal
distribution of total rainfall without reference to its synoptic origin (Ayoade, 1970, 1974, 1977; Oyebande
and Oguntoyinbo, 1970; Adedokun, 1978; Olaniran, 1983; Adefolalu, 1986; Anyadike, 1993). Similar
work in temperate regions has however frequently adopted the synoptic approach (Goldie, 1936; Sawyer,
1952; Shaw, 1962; Smithson, 1969) in the study of rainfall climatology. Previous climatological studies of
* Correspondence to: Department of Geography, University of Ibadan, Ibadan, Nigeria.
CCC 0899–8418/98/111273 – 12$17.50
© 1998 Royal Meteorological Society
1274
I.O. ADELEKAN
Nigerian thunderstorms have been carried out but these are limited and restricted to a few stations.
Mulero (1973) looked at the seasonal distribution of thunderstorm days in Nigeria for the period of
1962–1971 without reference to its association with rainfall. Balogun (1981) went on further to examine
total rainfall distribution in Nigeria in relation to thunderstorm activity for the period 1950–1976.
Oladipo and Mornu (1985) also looked at the climatological characteristics of thunderstorms at a
northern station, Zaria, for the period 1969 – 1982 and finally, Salau (1986) carried out a temporal and
comparative analysis of thunderstorms and related phenomena (hail, squall and lightning) for three
northern stations, namely Kaduna, Jos and Zaria.
The only attempt made to study Nigerian total rainfall in relation to the synoptic origin was the work
of Omotosho (1985). His study was however restricted to only five stations (four in northern Nigeria and
one in the south) and a relatively short 5 year period (1972–1976). It is not possible to make adequate
generalizations about the rainfall of a large country like Nigeria, with a land area of 924000 km2, from
the analysis of only five weather stations. Furthermore, the period 1972–1976 consisted mainly of dry
years in the north and therefore did not present a typical set of conditions.
The present study examines the spatial and temporal variation in the occurrence of thunderstorm
rainfall in Nigeria for a period of 30 years, 1960–1989. Unlike previous studies, it is based on data for
19 stations spread throughout the country. It also seeks to identify regions which are similar with respect
to the amount of rainfall produced by thunderstorm activity during the wet season period (April–
October).
2. DATA
The data used for this study were extracted from daily weather registers kept by the Nigerian
Meteorological Services, Lagos. Daily rainfall data were processed for nineteen synoptic stations (Figure
1) chosen in such a way as to have uniform spatial distribution and also to reflect the varied relief of the
country. The number of years for which a synoptic station has been in existence and the continuity of the
record were also taken into consideration. A sampling frame dividing the country into squares of 2° by
2° (latitude by longtitude) was used in selecting stations.
Rainfall was classified as being generated by thunderstorm on the basis of the following:
(i) Associated windgusts less than 23 knots. This value was chosen to coincide with the lower limit of
23 knots for squall lines given by the WMO (1965). In most cases, winds associated with isolated
thunderstorms do not usually go beyond 20 knots.
(ii) A record in the daily weather registers of thunder no later than 15 min before or after the start of
rain.
(iii) Main storm duration of usually not more than 1 h.
It is found that the duration of moderate or heavy rain from a single cell may vary from a few minutes
to almost an hour in a large active one (Byers, 1951; Landsberg, 1964). Longer storms lasting up to 2 h
with criteria (i) and (ii) were however also included as Kamara (1986) noted that in West Africa, the
lifetime of local thunderstorms may extend to 2 h.
3. NIGERIAN RAINFALL
Nigeria enjoys a tropical climate which is influenced mainly by two air masses. These are the cooler rain
bearing southwest monsoon which blows from the Atlantic Ocean and the hot, dry continental airmass
from the Sahara Desert. The fluctuating boundary zone between the two airmasses is called the
‘intertropical discontinuity’ (ITD). The amount, seasonal distribution and type of rainfall as well as the
length of the wet season at any place in Nigeria depend largely on its location with respect to the
fluctuating ITD and the associated weather zones.
© 1998 Royal Meteorological Society
Int. J. Climatol. 18: 1273 – 1284 (1998)
1275
THUNDERSTORM RAINFALL OVER NIGERIA
From March onwards, the southwest monsoon winds begin to advance inland in response to the
increasing insolation and decreasing pressure in the continental interior and the surface location of the
ITD shifts northwards correspondingly. As the depth of this maritime air increases, it brings cloud and
rain which begins at the coast and advances inland until, between June and September, when the whole
country experiences abundant rainfall.
3.1. Seasonal distribution
Thunderstorm rainfall in Nigeria occurs in two distinct patterns; one during the dry season months of
November–March and the other during the main wet season of April–October (Figure 2). The dry season
thunderstorm rainfall pattern reflects the general pattern of total rainfall in Nigeria, decreasing from
south to north despite the intense surface heating experienced during the dry season in the northern part
of the country. This suggests that during the dry season, thunderstorm rainfall events are mainly
dependent on the moisture content of the prevailing airmass and not just on local surface heating. In the
southern part of the country, the maritime winds occur all the year around, unlike the interior where they
are a seasonal phenomenon, with the result that it is their moisture content that tends to be crucial
(Anyadike, 1981). Much of the dry season rainfall over the country is convective in origin, resulting in
thunderstorms even though the rainfall amounts are low.
Figure 1. Relief map of Nigeria showing the 19 synoptic stations
© 1998 Royal Meteorological Society
Int. J. Climatol. 18: 1273 – 1284 (1998)
1276
© 1998 Royal Meteorological Society
I.O. ADELEKAN
Int. J. Climatol. 18: 1273 – 1284 (1998)
Figure 2. Distribution pattern of mean monthly thunderstorm rainfall (mm) in Nigeria (1960–1989)
© 1998 Royal Meteorological Society
THUNDERSTORM RAINFALL OVER NIGERIA
1277
Int. J. Climatol. 18: 1273 – 1284 (1998)
Figure 2 (Continued)
1278
© 1998 Royal Meteorological Society
I.O. ADELEKAN
Int. J. Climatol. 18: 1273 – 1284 (1998)
Figure 2 (Continued)
1279
THUNDERSTORM RAINFALL OVER NIGERIA
Figure 3. Trend of national thunderstorm rainfall averages in Nigeria (1960 – 1989)
By May, the coastal section of the country records as much as 60–80 mm of thunderstorm rainfall
while the northern section receives only about 10 mm. The area around Jos Plateau (1286 m) however
stands out as a region of high thunderstorm rainfall from its surroundings with 60 mm of thunderstorm
rainfall. This pattern confirms the fact that orography and moisture content of the atmosphere play
important roles in the occurrence of thunderstorms (Mulero, 1973). In the month of June, the pattern of
thunderstorm rainfall distribution changes from the dry season south–north pattern to an almost
east–west pattern. Thunderstorm rainfall declines east and west of longitude 9°E but north of latitude
11°N, thunderstorm rainfall decreases northwards. This distribution pattern reveals that thunderstorm
rainfall is not ITD based as long as there is enough moisture in the atmosphere.
At the peak of the rainy season in July the maritime airmass is over Nigeria, so moisture is not a
limiting factor. The coastal areas experience reduced thunderstorm activity as well as minimum rainfall
due to the upwelling process which begins to occur at this time and is well marked in the coastal waters
of West Africa by August. This upwelling process leads to colder waters at the ocean surface which
Table I. Coefficient of variation, C.V. (%) for wet season and annual thunderstorm
rainfall in Nigeria (1960 – 1989)
Station
Calabar
Gusau
Ibi
Ikeja
Ilorin
Jos
Kaduna
Kano
Lokoja
Maiduguri
Makurdi
Minna
Nguru
Oshogbo
Portharcourt
Potiskum
Sokoto
Yelwa
Yola
© 1998 Royal Meteorological Society
Coefficient of variation, C.V. (%)
Wet season
Annual
33.0
39.4
57.7
49.6
47.1
31.4
46.7
54.5
28.8
51.6
48.3
60.6
48.7
38.7
31.8
41.1
33.3
39.9
36.2
30.6
39.4
54.7
44.8
44.0
31.2
46.2
54.5
28.1
51.6
47.4
58.6
48.7
34.4
27.9
41.0
32.7
39.4
35.4
Int. J. Climatol. 18: 1273 – 1284 (1998)
1280
I.O. ADELEKAN
stabilizes the lower atmosphere over the region. The area around Jos Plateau however records the highest
thunderstorm rainfall with totals up to 130 mm. The Lokoja area (63 m) records the next highest total
with 70 mm of thunderstorm rain. In September, the area of high thunderstorm rainfall shifts westward
with Kaduna (645 m) recording 80 mm and the Oshogbo (302 m)–Lokoja (63 m) axis recording 80 mm
thunderstorm rainfall. Again, these areas are highlands relative to their surroundings and the effect of
orography on thunderstorm rainfall is clear. By October, the distribution of thunderstorm rainfall changes
and a NW–SE pattern is produced which decreases from south to north.
3.2. Interannual 6ariations
The interannual variation in rainfall due to thunderstorms is shown in Figure 3. The mean annual
thunderstorm rainfall averaged over the 19 synoptic stations for the study period is 322.2 mm. The
maximum national average of 419.8 mm was recorded in 1969 while the minimum national average of
207.3 mm was recorded in 1989. This gives a range of 212.5 mm or 66.0% of the long period mean. The
coefficient of variation for the west season and annual thunderstorm rainfall for the nineteen stations is
shown in Table I.
3.3. Spatial 6ariations
The spatial variability of rainfall due to thunderstorm rainfall was studied using correlation analysis, a
method which has been used in several studies (Longley, 1974; Sharon, 1974, 1979, 1981; Jackson, 1978).
The coefficient of correlation gives a measure of the tendency of two parameters to vary together. In this
analysis, the wet season (April – October) thunderstorm rainfall totals between station pairs are used.
These are months with the greatest absolute variability from year-to-year.
A classification of the stations into regions based on their homogeneity of rainfall due to thunderstorms
was carried out based on the results of the correlation matrix produced between stations (Table II).
Attempts have been made to classify total rainfall of some African countries using principal component
analysis (Van Regenmortel, 1995; Basalirwa, 1995). Jackson and Weinand (1994) also employed principal
components and cluster analyses in the classification of tropical stations. The elementary linkage analysis
developed by McQuitty (1957) is however employed in this classification. This method has been
successfully employed by Gregory (1965) to delimit regions of related rainfall fluctuations in Sierra-Leone,
West Africa.
The elementary linkage analysis gives a grouping result that approximates to that of orthogonally
rotated factor analysis and factor analytic techniques and allows grouping to be carried out by the
inspection of the matrix of correlations among the objects to be classified (Lawley and Maxwell, 1963;
Johnston, 1976). Every station is assigned to a group in terms of its highest index of association with
another station and the lower limit of association used in building the clusters is exclusively determined
by the data. Each station in any group is therefore more highly correlated with another station in the
same group than it is with any other station in any other group. In the majority of cases, the second
highest correlation for any station also falls within the same group whilst at times all members of the same
group are highly interlinked with little or no significant correlations outside such a group.
Figure 4(a) and (b) show station groupings of high correlation of thunderstorm rainfall within the
country. Six discrete regions were obtained. During the wet season, stations within a region have an
internal coherence in terms of the correlation of temporal fluctuations. All regions with the exception of
region II also display the characteristic of having either a single maximum or double maxima in
thunderstorm rainfall.
Region I:
Region II:
Calabar, Ikeja and Port-Harcourt (These are coastal stations with double maxima in
May/June and October).
Kaduna, Yelwa, Ibi, Lokoja, Oshogbo, Ilorin and Potiskum.
© 1998 Royal Meteorological Society
Int. J. Climatol. 18: 1273 – 1284 (1998)
© 1998 Royal Meteorological Society
Table II. Correlation matrix of wet season (April – October) thunderstorm rainfall
1 1.00
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
−0.01
1.00
0.44
0.46
1.00
0.93
0.08
0.61
1.00
0.49
0.57
0.54
0.58
1.00
0.73
0.31
0.71
0.68
0.64
1.00
0.62
0.49
0.69
0.45
0.66
0.71
1.00
0.23
0.62
0.54
0.37
0.43
0.48
0.34
1.00
0.68
0.46
0.67
0.60
0.66
0.77
0.77
0.20
1.00
0.27
0.53
0.62
0.35
0.46
0.35
0.60
0.69
0.29
1.00
0.08
0.52
0.60
0.76
0.52
0.80
0.47
0.65
0.69
0.39
1.00
0.49
0.47
0.58
0.46
0.35
0.51
0.37
0.79
0.28
0.66
0.75
1.00
0.20
0.68
0.36
0.35
0.35
0.22
0.51
0.61
0.30
0.76
0.49
0.63
1.00
0.47
0.51
0.67
0.43
0.69
0.57
0.86
0.25
0.77
0.55
0.35
0.14
0.43
1.00
0.95
0.15
0.45
0.85
0.56
0.68
0.74
0.27
0.76
0.38
0.52
0.48
0.32
0.60
1.00
0.50
0.55
0.49
0.64
0.78
0.63
0.66
0.43
0.78
0.48
0.50
0.38
0.38
0.76
0.59
1.00
0.18
0.65
0.64
0.25
0.47
0.43
0.65
0.58
0.58
0.71
0.45
0.47
0.56
0.54
0.28
0.46
10.0
0.46
0.54
0.72
0.51
0.68
0.55
0.88
0.40
0.78
0.73
0.47
0.44
0.62
0.87
0.58
0.64
0.76
1.00
0.09
0.86
0.54
0.20
0.56
0.50
0.53
0.70
0.54
0.67
0.66
0.72
0.75
0.42
0.17
0.50
0.77
0.69
1.00
1281
Int. J. Climatol. 18: 1273 – 1284 (1998)
Calabar
Gusau
Ibi
Ikeja
Ilorin
Jos
Kaduna
Kano
Lokoja
Maiduguri
Makurdi
Minna
Nguru
Osogbo
Port-Harcourt
Potiskun
Sokoto
Yelwa
Yola
1
THUNDERSTORM RAINFALL OVER NIGERIA
Station
1282
I.O. ADELEKAN
Figure 4. (a) Thunderstorm rainfall regions of Nigeria and (b) linkage analysis diagram
The spatial extent of Region II is mostly in the middle belt of the country since it represents a transition
zone between the coastal area to the south and the extreme northern section of the country. This results
in a combination of stations with a single maximum (Kaduna, Yelwa and Potiskum) and double maxima
(Ilorin, Ibi, Lokoja and Oshogbo) occurring at different times during the wet season.
© 1998 Royal Meteorological Society
Int. J. Climatol. 18: 1273 – 1284 (1998)
1283
THUNDERSTORM RAINFALL OVER NIGERIA
Region III:
Region IV:
Region V:
Region VI:
Gusau, Yola and Sokoto (single maximum in August).
Jos and Makurdi (single maximum in July).
Kano and Minna (single maximum with that of Kano occurring July/August and
Minna in September).
Nguru and Maiduguri (single maximum in July/August).
4. CONCLUSION
This study has shown that thunderstorm rainfall in Nigeria presents two distinct patterns–one during the
dry season months of November to March and the other during the main wet season months of
April–October. The dry season thunderstorm rainfall pattern reflects the general pattern of total rainfall
in Nigeria which decreases from south to north. This suggests that during the dry season, thunderstorm
rainfall events are dependent both on the moisture content of the atmosphere and also on local heating.
During the wet season the pattern changes, especially for the peak months of July–September, with
thunderstorm rainfall increasing from south to north up to latitude 11°N. The areas around Jos Plateau
and east of it receive the highest amount of thunderstorm rainfall during this period. The country’s
thunderstorm rainfall is marked by year-to-year variations in amount with an emerging downward trend
below the mean since 1975.
Station groupings using elementary linkage analysis of wet season thunderstorm rainfall yielded six
regions. These regions show internal coherence in terms of temporal fluctuations of thunderstorm rainfall
during the wet season.
ACKNOWLEDGEMENTS
The author is grateful to Dr A.O. Aweto and the anonymous refrees for their comments and suggestions
which have been very helpful in the preparation of the final draft.
REFERENCES
Adedokun, J.A. 1978 ‘West African precipitation and dominant atmospheric mechanisms’, Arch. Met. Geoph. Biokl. Ser. A., 27,
289 – 310.
Adefolalu, D.O. 1986. ‘Rainfall trends in Nigeria’, Theor. Appl. Climatol., 37, 205 – 219.
Anyadike, R.N.C. 1981. ‘On the relative contribution of atmospheric parameters to the rainfall of West Africa’, Arch. Met. Geoph.
Biokl. Ser. A., 30, 87–98.
Anyadike, R.N.C. 1993. ‘Seasonal and annual rainfall variations over Nigeria’, Int. J. Climatol., 13, 567 – 580.
Ayoade, J.O. 1970. ‘The seasonal incidence of rainfall’, Weather, 25, 414 – 418.
Ayoade, J.O. 1974. ‘A statistical analysis of rainfall over Nigeria’, J. Tropical Geogr., 39, 11 – 23.
Ayoade, J.O. 1977. ‘On the use of multivariate techniques in climatic classification and regionalization’, Arch. Met. Geoph. Biokl.
Ser. B. 24, 257 – 267.
Balogun, E.E. 1981. ‘Seasonal and spatial variations in thunderstorm activity over Nigeria’, Weather 36, 192 – 196.
Basalirwa, C.K. 1995. ‘Delineation of Uganda into climatological rainfall zones using the method of Principal Component Analysis’,
Int. J. Climatol., 15, 1161–1177.
Byers, H.R. 1951. Thunderstorms, Compendium of Meteorology, American Meteorological Society Boston, MA, pp. 681 – 686.
Desbois, M., Kayiranga, T., Gnamien, B., Guessous, S., and Icon, L. 1988. ‘Characterization of some elements of the Sahelian
climate and their interannual variations for July 1983, 1984 and 1985 from the Analysis of Meteosat ISCCP Data’, J. Climate,
1(9), 867 – 904.
Goldie, A.H.R. 1936. Rainfall at fronts of depression, Geophysical Memoir, Great Britain Meteorological Office, 18 pp.
Gregory, S. 1965. Rainfall o6er Sierra-Leone, Research Paper No. 2, Department of Geography, University of Liverpool. 58 pp.
Jackson, I.J. 1978. ‘Local differences in the patterns of variability of Tropical rainfall: Some characteristics and implications’, J.
Hydrol., 38, 273 – 287.
Jackson, I.J. and Weinand H. 1994. ‘Towards a classification of tropical rainfall stations’, Int. J. Climatol., 14, 263 – 286.
Johnston, R.J. 1976. Classification in Geography, Catmog 6.
Kamara, S.I. 1986. ‘The origins and types of rainfall in West Africa’, Weather, 41, 48 – 56.
Landsberg, H.E. 1964. Physical Climatology, Grall Printing Company Incorporated, Pennsylvania.
© 1998 Royal Meteorological Society
Int. J. Climatol. 18: 1273 – 1284 (1998)
1284
I.O. ADELEKAN
Lawley, D.N. and Maxwell, A.E. 1963. Factor Analysis as a Statistical Method, London.
Longley, R. 1974. ‘Spatial variations of precipitation over the Canadian prairies’, Mon. Wea. Re6., 102, 307 – 312.
McQuitty, L.L. 1957. ‘Elementary linkage analysis for isolating orthogonal and oblique types and typal relevances’, Educational
Psychol. Measur., 17, 207–229.
Mulero, M.A. 1973. ‘On seasonal distribution of thunderstorm days in Nigeria’, Quart. Meterol. Mag. 3, 73 – 78.
Oladipo, E.O. and M.E. Mornu, 1985. ‘Characteristics of thunderstorms in Zaria, Nigeria’, Weather, 40, 316 – 321.
Olaniran, O.J. 1983. ‘The monsoon factor and seasonality of rainfall distribution in Nigeria’, Malaysian J. Tropical Geogr., 7, 38 – 45.
Omotosho, J.B. 1985. ‘The separate contributions of linesqualls, thunderstorms and the monsoon to the total rainfall in Nigeria’,
Int. J. Climatol., 5, 543–552.
Oyebande, B.C. and Oguntoyinbo J.S., 1970. ‘An analysis of rainfall patterns in the South Western States of Nigeria’, Nigerian
Geographical J., 13, 141–162.
Salau, O. 1986. ‘Temporal and comparative analysis of thunderstorms and some related phenomena in Zaria, Jos and Kaduna
(Nigeria)’, Theor. Appl. Climatol., 37, 220–232.
Sawyer, J.S. 1952. ‘A study of rainfall of two synoptic situations’, Q. J. R. Met. Soc., 78, 231 – 246.
Shaw, E.M. 1962. ‘A decade of precipitation at keele, 1952 – 1961’, N. Staffordshire J. Field Stud., 2, 132.
Sharon, D. 1974. ‘The spatial pattern of convective rainfall in Sukuma-Land, Tanzania — A ststistical analysis’, Arch. Met. Geoph.
Biokl., B22, 201 – 208.
Sharon, D. 1979. ‘Correlation analysis of the Jordan Valley rainfall field’, Mon. Wea. Re6.,109, 1042.
Sharon, D. 1981. ‘The distribution in space of local rainfall in the Namib Desert’, J. Climatol., 1, 69.
Smithson, P.A. 1969. ‘Rainfall variations in the synoptic origin or rainfall across Scotland’, Scottish Geographical Mag., 85(3),
182 – 195.
Van Regenmortel, G. 1995. ‘Regionalization of Botswana rainfall during the 1980s using principal component analysis’, Int. J.
Climatol., 15, 313 –323.
WMO, 1965. Analysis and Prognosis in Meterology, Technical Note, 71.
© 1998 Royal Meteorological Society
Int. J. Climatol. 18: 1273 – 1284 (1998)
Документ
Категория
Без категории
Просмотров
4
Размер файла
411 Кб
Теги
610
1/--страниц
Пожаловаться на содержимое документа