Potential Water Deficits in Uganda

Potential water deficits have been computed, using a generalized water balance based on meteorological data, from sites in Uganda, with sufficient data for an estimate of evaporation using the Penman formula (Eo). Available water capacity of Ioo mm was adopted for a general water balance, since, for most of the country, monthly excess of rainfall over evapotranspiration was insufficient to give higher soil-water storage. The averages of potential water deficit values for each calendar month from 1931-69 give a more realistic estimate than water balances using average monthly meteorological data which tend to underestimate the potential water deficit in the wet months and overestimate it during dry months. When the go per cent lower confidence limits of monthly potential water deficit exceeded zero, the month was considered wet, and when the 90o per cent upper confidence limit was below 50 mm for the month, it was assumed dry. Using these criteria, the duration of the wet and dry seasons were mapped for Uganda. This assessment is of value in rangeland management and agricultural planning, particularly in areas of expansion where an estimate of the incidence and length of periods of drought and water availability is needed. THE unreliable nature of the seasonal rainfall regime has been a longstanding problem facing agriculture in East Africa. The annual rainfall and its reliability have been used to indicate broad patterns, though it has been recognized that what is needed for agriculture is an assessment of the distribution within the year (Kenworthy, 1964). There are great problems in attempting to quantify seasonal patterns throughout a region in which the timing of the wet and dry seasons is not uniform. The bimodal rainfall pattern has been analysed by separating the first and second half of the year, but such a device would bisect the main rainy season in northern Uganda. Shorter intervals, months, Io-day periods or pentads, would seem to provide the best basis for analysing the seasonal pattern. If statistical analysis is to be of most value in agriculture, forestry or rangeland management in East Africa, it must be concerned with reliability. Manning's (1956) confidence limits based on 3-week moving totals provide a useful method of assessing the seasonal pattern of rainfall expectancy, but must be applied individually for the requirement of each crop. This has been done successfully in planning optimum planting dates for cotton in Uganda (Manning, 1960). Huxley (1965) has indicated the dangers of using Manning's confidence limits without reference to other climatic considerations. The level of rainfall expected must be seen in relation to the atmospheric demand and the unique way in which the crop responds to this. Since the publication of Manning's work much more information on potential evapotranspiration and crop water requirements has accumulated. Only an approach using the total water balance can adequately assess the seasonality of moisture and is what Russell (1960) had in mind in suggesting the need for a soil moisture statistic. This study attempts by a computer analysis of the water balance over a period of almost 40 years to assess the expectancy of soil-water deficits. The water balance was based on the simple model P = E, + AS + R, where P = precipitation, E, = evapotranspiration, AS = infiltration or change in storage and R = run-off and percolation. Evapotranspiration has been estimated from meteorological data using Penman's formula for

Cartography · Dry season · Economics · Geography · Socioeconomics · Water resource management · Climate variability and models · Environmental Science · Hydrology and Drought Analysis · Plant Water Relations and Carbon Dynamics

• An Approach toward a Rational Classification of Climate

  Open Access•C W Thornthwaite•Geographical Review•1948

  Add to list