Climatic drought characteristics of north west region of Bangladesh

Abstract

This study deals with statistical characteristics of climatic drought which is defined as the difference between potential evapotranspiration and rainfall. Droughts have been a matter of serious concern to man since ancient times and even today it is an outstanding example of man I s helplessness before nature I s large-scale and formidable phenomena. To the meteorologist, rought is a rainless situation for an extended period of time uring which some precipitation should have been normally received depending upon the geographical location of the region and season of the year. To the agriculturist, climatic or agricultural drought is a shortage of moisture availability for crops. Analysis begins with calculation of daily drought values. Then 10-day yearly drought maxima is calculated for the period of _eeard obtained for five meteorological stations in the Teesta Barrage Project area. The reason for choosing lO-day interval is that irrigation requirement is usually calculated on lO-day basis. Extreme Value Type 1 (EVl) distribution is to fitted to IO-day drought maxima for each station and goodness-of-fit is judge by visual inspection of probability of plots on extreme value paper. Overall, EVI distribution seems to fit the drought data reasonably well. Frequency relationships have also been presented and using these relationships the magnitude of IO-day d}:ough can be calculated for a given return period. Method of calculating irrigation requirement using these drought values has so been illustrated. The analysis of drought was performed both i for growing season of T.Aman (July 15 to November 15) and for the cri.tical growing period of the same crop (October 15 to November 15) . Another important parameter in the description of climatic drought is the dry spells or sequences of dry or non-rainy day and wet or rainy days. So the next part of the analysis deals wi th frequency of wet- and dry-day sequences having rainfall greater than or equal to and less than specified threshold rainfall val ues, respectively. Threshold rainfall values for this study were chosen as 6.0 mm, 3.0 mm and 1.0 mm for each station. For each year the consecutive dry- and wet-day sequences having maximum length(days) was selected. This yielded one dry-day and one wet-day sequence in each year. Results obtained show that the yearly maximum length of consecutive wet-day and dry-day sequences can be approximated by normal distribution. - Finally seasonal distribution of dry-day sequences has been studied. First average length of consecutive dry-day sequences for each month over the entire period of record was calculated for each threshold and plotted against respective months as a line graph. The largest dry-day sequences occur in January, February and March and again in october, November and December. The frequency of dry-day sequences having length greater than or equal t.O 25 days was also calculated for each month for the peri,d record for each threshold. Largest frequency of such sequences was found to occur in January through March and again in October through December.