Impact of Bias-correction Uncertainties in Quantifying Potential Future Hydrologic Response in Bago River Basin

Flooding due to heavy rainfall events as a result of the seasonal monsoon is an annually recurring problem in Bago River basin. The river basin is a critical water resource and livelihood area that has significant role in the future development in the region. Due to the increasing development in the basin and its critical role on water resource, assessment of current flood hazard and the potential future changes is a priority. High-resolution projections for present and future time slices are produced by using Weather Research and Forecast (WRF) model. The projections are downscaled by employing Pseudo-Global-Warming method using CMIP5. It is a common practice to bias-correct the high-resolution projections from RCMs. However, it has been shown that bias-correction can significantly affect the projections and can increase the uncertainty. This study focuses on the evaluation of multiple bias-correction techniques and available gridded precipitation products as reference for simulating hydrologic impacts on a sparsely gauged basin. It also investigates the influence of spatial interpolation and scale of the gridded precipitation product to the hydrological projections. The hydrologic impact is simulated by a water-energy budget distributed hydrological model (WEB-DHM) using the bias-corrected projections. The aim of the study is to show the impact of different uncertainty sources from bias-correction components (bias-correction method, reference data, and spatial interpolation used) to the hydrologic impact modeling in the basin. The study explores the impact of bias-correction component uncertainties in simulating the distinct high flow and low flow response of the river basin that is critical to water resource and hazard assessment. This has a huge implication in quantifying the flood risk and drought risk in the basin for future development planning purposes.