Impact of climate change on droughts in Maduru Oya river basin in Sri Lanka over the 21st century

Abstract

Drought is an creeping hazard that is least understood and the most complex of all-natural hazards. The drought study requires large historical climatological and meteorological datasets and their complex inter-relationships. Its impacts are prominently observed on a local scale only when the severity becomes high, and the coherent onset and persistence of mild droughts may go unnoticed. The current study investigates the existing drought conditions and future drought risk in the Maduru Oya River Basin over the 21 st -century in terms of meteorological and hydrological drought indices (i.e., SPI, SPEI, RDI, EDI and SRI). The future hydrology over the basin is simulated for this research, using bias-corrected precipitation and potential evapotranspiration outputs under RCP 4.5 and 8.5 of the MPI-M-MPI-ESM-MR. The relevant drought-related indices were computed in monthly and seasonal timescales over the 19512099

period. The time series have been classified for drought characterization, including drought frequency, severity, trend, and probability computation. Further, to assess the impact of these droughts on the basin's response, a hydrological model (i.e., HEC-HMS) was developed to simulate the discharge at the Padiyathalawa outlet considering 2008-2012 as validation period. The results of the monthly timescale for SPI (approximately similar drought frequency and severity by RDI and EDI) depicted that the severe and extreme droughts (45) occurred in March (8), August (5), September (4), October (6) and November (9) in the historical period. Severe and extreme droughts (110 under RCP 4.5,104 under RCP 8.5) are projected in January (17), February (12), April (10), May (12) and December (13) under RCP 4.5 and January (12), February (10), April (13), June (13) and August (10) under RCP 8.5 over the 21 -century. The SPEI at monthly timescale identified highest number of severe and extreme drought (67) events in the historical period and projected highest severe and extreme drought (128 under RCP 4.5,122 under RCP 8.5) events over the 21 st -century in the study area. The hydrological drought index, SRI projected severe and extreme droughts under RCP 4.5 (65) and RCP 8.5 (62) over the 21 st -century that is about 50 % frequency of the meteorological drought indices. The Northeast Monsoon season had the least drought episodes (~20) in the historical period, and on a seasonal time scale, high drought frequency (~30 using meteorological drought indices and ~20 using SRI under RCP 4.5 and RCP 8.5) and severity(severe and extreme droughts) are projected in the Northeast Monsoon. It is also observed that there is a consistent mild drought throughout the mid (~70) and end (~65) century for a maximum duration compared to the historical (~50) period. The accuracy of results obtained from the continuous HEC-HMS model (NSE, RMSE Std. Dev, and R 2 st values of 0.59, 0.72, and 0.60 achieved in validation)highlights the efficient way to simulate a basin's hydrological parameters. The model can project the future variation of streamflow of the Maduru Oya River Basin under varied climatic conditions. The discharge is projected to have a decreasing trend (Sen’s slope=0.008) for future years, identified as droughts. It can be concluded that the impact of climate

change on meteorological drought will affect the discharge of the basin. Moreover, due to time

lag between meteorological and hydrological drought, about 50 % of meteorological droughts

may lead to a severe and extreme hydrological drought in the Maduru Oya River Basin over

the mid-century (14) under RCP 8.5 and end-century (13) under RCP 4.5 scenarios. This study

will begin with quantitative investigations

including streamflow variability and climatology over the basin incorporating the application of regional circulation models.