Coupled hydrodynamic modelling approach to assess land use change induced flood characteristics.

Abstract

MIKE 11 rainfall-runoff model and MIKE 21 overland flow model were successfully coupled in the MIKE FLOOD platform for flood simulation in Gin catchment (932 km2) of Sri Lanka to assess land use change induced changes in flood discharge. MIKE 11 Nedbør-Afstrømings-Model (NAM) rainfall-runoff simulation depicted a good agreement with the observed discharge at Thawalama and Baddegama gauging stations. MIKE FLOOD, validated against the two major flood events that occurred in May 2003 and in May 2017, showed a reasonable agreement with the observed water depths and peak discharge values displaying more than 70% goodness of fit between the observed and simulated inundated extents. Dominant land use change processes in the catchment between 1999 and 2016 were identified as the forest area and built-up land expansion at the expense of agricultural land and bare land which possessed contradictory impacts on flood generation. The impact of a single factor, 17-year land use change, on flood formation was differentiated. In the upstream sub-catchment having 490 km2, despite a 0.74% increase in the built-up land, 2.85% increase in the forest area had significantly contributed to mitigate the overall flood formation with 34% and 40% reduction in the peak discharge and the flood volume, respectively. Overall reduction of the flood discharge attributed to the forest expansion emphasised the importance of preserving forest cover and pervious area. The modelling framework presented in this typical tropical monsoon catchment study could be effectively used to quantify the land use change induced flow regime variations in similar catchments.