Remote sensing based evaluation of uncertainties on modelling of streamflow affected by climate change

Abstract

Assessment of the impacts of land-use and climate change on streamflow is vital to develop climate adaptation strategies. However, uncertainties in the climate impact study framework could lead to changes on streamflow impact. The aim of this study is to assess the uncertainties on Digital Elevation Model (DEM), Satellite Precipitation Product (SPP) and climate projection on the modelling of streamflow affected by climate changes. These uncertainties are evaluated and reduced independently. The climate projection uncertainty is addressed through the modification of the Quantifying and Understanding the Earth System - Global Scale Impacts (QUEST-GSI) methodology. Twenty-six modified QUEST-GSI climate scenarios were used as climate inputs into the calibrated Soil and Water Assessment Tool (SWAT) model to evaluate the impacts and uncertainties of climate change on streamflow for three future periods (2015-2034, 2045-2064 and 2075-2094). The selected study areas are the Johor River Basin (JRB) and Kelantan River Basin (KRB), Malaysia. The Shuttle Radar Topography Mission (SRTM) version 4.1 (90m resolution) DEM and the Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks – Climate Data Record (PERSIANN-CDR) SPP which show a better performance were selected for the SWAT model modification, calibration and validation. The results indicated that the modified SWAT model could simulate the monthly streamflow well for both basins. Land-use and climate changes from 1985 to 2012 reduced annual streamflow of the JRB and KRB by 5% and 4.2%, respectively. In future, the annual precipitation and temperature of the JRB / KRB are projected to increase by -0.4-10.3% / 0.1-11.2% and 0.6-3.2oC / 0.8-3.3oC, respectively, and that this will lead to an increase of annual streamflow by 0.5-13.3% / 4.4-18.5%. This study showed that satellite data play an important role in providing input data to hydrological models.