Improving Hydrologic Modeling for Poorly Gauged Basins

Abstract

Water management in poorly gauged basins is subject to several difficult challenges. This is a serious problem particularly in developing countries where the resources and infrastructure are often very limited or even unavailable. The SERVIR program of NASA and USAID focuses on improving water resources management in several developing regions of the world by utilizing cutting-edge observations from space. Funded by the SERVIR program, this dissertation focuses on improving streamflow monitoring and forecasting capabilities for selected poorly gauged pilot basins in Africa. The dissertation includes three manuscripts. The first manuscript presents the development of a new real-time streamflow monitoring and forecasting platform, Multimodel and Multiproduct Streamflow Forecasting-Real Time (MMSF-RT), built using multiple hydrologic models and satellite-based precipitation products to produce more accurate streamflow forecasts and better characterize forecast uncertainty. It involves advanced methods for bias correction, model calibration, and forecast merging. The second manuscript focuses on implementation of the MMSF-RT platform for operational forecasting in the pilot basins. In the third manuscript, the goal is to improve individual components of the platform. In particular, the focus was on improving the HYMOD model (a component of MMSF-RT platform) using newly available satellite-based estimates of actual evapotranspiration. From a more general perspective, this part of the dissertation shows how hydrologic modeling can benefit in different ways from newly available satellite-based data sources and more specifically, how efficiently (and differently) satellite-based evapotranspiration products could be utilized to improve hydrologic model performance.