Future changes in meteorological drought characteristics over Bangladesh projected by the CMIP5 multi-model ensemble

Drought is an inconspicuous natural disaster. In a warmer world, the severity and coverage of drought are expected to change, and it is essential to study these changes at smaller scale. This study detected changes in drought frequency, severity, and intensity in Bangladesh from a bias-corrected CMIP-5 multi-model projection of 11 members under a business-as-usual RCP8.5 scenario. We have used two well-known meteorological drought indices, Standardized Precipitation Index (SPI) and Standardized Precipitation and Evaporation Index (SPEI). SPI is solely based on precipitation, while SPEI considers climatic water balance and incorporates the effect of temperature. Two different methods of estimation of potential evapotranspiration (PET), namely Thornthwaite and Hargreaves methods, are explored. SPEI-based drought identification is found to have high sensitivity among these PET estimation methods. In Bangladesh, SPI-based analysis suggests virtually no change in the long-term drought (12-monthly) condition and a minor change in short-term (6-monthly or less) droughts. SPEI evaluated with Hargreaves method projects a similar scenario for long-term droughts but an increase in both drought frequency and severity in short timescales. At seasonal scale, winter and pre-monsoon are projected to be potentially more affected by water stress in the future. A spatially coherent shift in wet-dry regime is also found over the northern part of Bangladesh under the warming world.