Impacts of Different Rainfall Estimates on Hydrological Simulation and Satellite Rainfall Retrieval Error Propagation

Accurate estimation of rainfall magnitudes and their spatial distribution plays an important role in hydrological applications such as flood risk analysis and river flow forecasts. In addition knowing the forcing data error structure is crucial part of hydrologic data assimilation system. Recent advances in radar and satellite based measurements have provided alternative methods for precipitation estimation. In this study, gauge, radar and satellite precipitation estimates are used to explore the impacts of different precipitation data sources on the accuracy of hydrological simulation over the Leaf River basin in Mississippi. The NEXRAD stage IV, 3B42 gridded TRMM, PERSIANN-CCS and rain-gauge data over the watershed are aggregated to similar resolution in time and forced to a conceptual hydrological model, SAC-SMA. The model performance with each of the rainfall estimates input has been assessed by statistical comparison of the simulated and measured streamflow at the watershed outlet. In this study, the standard error of satellite-based PERSIANN-CCS rainfall estimates conditioning on the assumed true field (i.e. radar rainfall) is obtained according to a multivariate function considering the spatial and temporal scales. Accepting the multiplicative nature of precipitation error, the Monte Carlo simulation based on log-normal distribution of error is conducted to generate the ensemble of precipitation and propagate them into a conceptual hydrologic model to investigate the impact of input error on streamflow simulation.