Agricultural flood and drought risk reduction by a proposed multi-purpose dam: A case study of the Malwathoya River Basin, Sri Lanka

Abstract

In this study, we investigate the role of a proposed multi-purpose dam for flood and drought risk reduction in the Malwathoya River Basin, Sri Lanka. The Malwathoya River Basin is the second largest basin in Sri Lanka with an area of 3,246 km 2 located in the dry zone of northern Sri Lanka. The river basin has an average annual rainfall of 1200 mm with two distinct dry and wet seasons and has an elevation range between 350 above mean sea level (AMSL) in the mountainous areas and 0 AMSL near the sea shore. For the lower part of the river basin, the dry season rice production of 170 km 2 area cannot be fully utilized due to annual precipitation of less than 750 mm and lack of irrigation water, which is diverted from the Malwathu River at the Thekkam diversion weir to large water storage tanks via a cascaded irrigation scheme. In the wet season, the rice paddies are frequently flooded due to intense rainfall from the North East monsoon between October and January as well as the inter monsoon during April causing severe economic losses during 2012 and 2014 floods. As a result, a multi-purpose dam with the capacity of 210 million cubic meters (MCM) was proposed in the middle of the Malwathoya River Basin to reduce flood peak discharges during wet season and to supply irrigation water for the Thekkam diversion weir during dry season. In our methodology, a holistic approach of combined drought and flood assessment is adopted for the multi- purpose dam operation. From this point of view, opposite actions are required for the risk reductions: the full reservoir storage should be maintained before the start of the dry season and the low levels are necessary for the flood peak reduction during flood season. For the hazard severity assessment, the long-term precipitation record has been investigated with the use of Standardized Precipitation Index (SPI) as well of Standardized Precipitation Evapotranspiration Index (SPEI) to select the most severe dry and wet seasons. In addition, past agricultural damages caused by 2012-2014 floods and droughts were correlated using the Normalized Difference Vegetation Index (NDVI) and Land Surface Water Index (LSWI) with field data at several locations. For the dam operation simulation, the 30-arcsec (about 1-km) grid distributed hydrological model BTOP was constructed from the global data sets and included reservoirs and medium water storage tanks with total capacity of 540 MCM located upstream of the Thekkam diversion weir. The BTOP model simulation with the ground-based precipitation was calibrated with observed river discharge data at the diversion weir and had a satisfactory statistical performance. Then, the calibrated BTOP model was used to simulate multi-purpose dam operation with the short- and long-term precipitation. From the BTOP model results, the proposed dam allowed us to establish an optimum operation for the flood and drought risk reduction and indicated a decrease in water related hazards. In addition, we estimated a standardized inflow index (SII) and a standardized reservoir storage index (SRSI) from the BTOP simulated reservoir inflows and water storage to demonstrate long-term dam performance on the same scale with SPI and SPEI values. As a result, this study demonstrates a combined flood and drought risk assessment from the water resources infrastructure point of view and allows us to reduce farmers' vulnerability in the northern region of Sri Lanka.