Comparison of process-based and lumped parameter models for projecting future changes in fluvial sediment supply to the coast

Abstract

Fluvial sediment supply (FSS) is one of the primary sources of sediment received by coasts. Any significant change in sediment supply to the coast will disturb its equilibrium state. Therefore, a robust assessment of future changes in FSS is required to understand the coastal system’s status under plausible climatic variations and human activities. Here, we investigate two modelling approaches to estimate the FSS at two spatially heterogeneous river basins: the Irrawaddy River Basin (IRB), Myanmar and the Kalu River Basin (KRB), Sri Lanka. We compare the FSS obtained from a process-based model (i.e., Soil Water Assessment Tool: SWAT) and an empirical model (i.e., the BQART model) for mid- (2046–2065) and end-century (2081–2100) periods under climate change and human activities (viz, planned reservoirs considered here). Our results show that SWAT simulations project a higher sediment load than BQART in the IRB and vice versa in KRB (for both future periods considered). SWAT projects higher percentage changes for both future periods (relative to baseline) compared to BQART projections in both basins with climate change alone (i.e., no reservoirs) and vice versa when planned reservoirs are considered. The difference between the two model projections (from SWAT and BQART) is higher in KRB, and it may imply that empirical BQART model projections are more in line with semi-distributed SWAT projections at the larger Irrawaddy River Basin than in the smaller Kalu River Basin.