A Framework for Evaluating the Propagation of Errors in Rainfall to Hydrographs

The ability to accurately capture the spatial and temporal variability of precipitation is important to estimate hydrographs from hydrologic models. This study explores the key aspects of radar based rainfall that contribute to the differences in simulated hydrographs. We employed two rainfall products; MRMS and Iowa Flood Center derived rainfall (IFC) to systematically evaluate the errors in their magnitudes and consequent implications to hydrologic responses. The model domain for this study is the state of Iowa using the IFC flood forecasting hydrologic model, also referred to as the hillslope-link model (HLM). This study builds on key radar rainfall aspects such as spatial and temporal resolution, river network based rainfall aggregation, interplay between radar range and river network width function, moments and quantiles associated with spatial and temporal distribution of rainfall, and rainfall-weighted flow distance (RWD) measure. Results from spatial and temporal aggregation of rainfall products show that the impact of spatial and temporal variability decrease with the increase in basin sizes. We find the error due to temporal variability more sensitive to hydrologic response than the spatial variability. Based on simulation experiments across resolutions, we observe that basin sizes ³ 1000 [km²] are less prone to error in streamflow response. The effect of resolutions is more prominent in the IFC rainfall than the MRMS rainfall. We evaluate the difference of event based network aggregated rainfall across basin scales and its translation into the difference in peak flows and their timings. The radar visibility range expressed in relation to the river network width function is also analyzed in relation with errors in stream flow response. Moreover, preliminary results based on the RWD approach show some promising results in terms of errors in movement and position of storms from two rainfall products and their translation into stream flow predictions. Overall, this study provides a comprehensive framework in identifying the errors in rainfall and their propagation to hydrologic response of the system.

Key Terms: IFC flood forecasting hydrologic model, HLM Model, spatial and temporal resolution, rainfall-weighted flow distance, radar range, width function