Characteristics of meteorological drought in Bangladesh

Abstract

Meteorological drought events occur in Bangladesh are diagnosed using monthly rainfall and mean air temperature from the surface observations and Regional Climate Model (RegCM) by calculating Standardized Precipitation Index (SPI) and Palmer Drought Severity Index (PDSI) for the period 1961–1990. The historical records of drought event obtained from the Bangladesh Bureau of Statistics and International Disaster Database are used to verify the SPI and PDSI detected events. The SPI and monthly PDSI are obtained for 27 station data across Bangladesh as well as for two subregions over the country. Result based on the observed data shows that regional information is better in drought diagnosis compared to the point information. The regional analysis is able to detect about 80 % of the drought events occurred during the study period. Frequency of moderate drought is higher for all over the country. The SPI calculated from RegCM rainfall shows that the detection of moderate drought events is 10, 7, and 21 % overestimated for 1-, 3-, and 6-month length, respectively, compared to using of observed data. For extreme drought cases, detection is overestimated (underestimated) by 25 % (79 %) for 1-month (6-month) length. The PDSI results for model and observed data are nearly same to SPI calculations. Model monthly PDSI result is overestimated (underestimated) by 29 % (50 %) for moderate (severe) drought events with reference to the observed PDSI. Hence, RegCM output may be useful to detect 3–6-month (monthly to seasonal) length moderate drought events over a heavy rainfall region likely Bangladesh.