Analyzing the Reliability of Synthetic Rating Curves for Continental Scale Flood Mapping

Abstract:

Empirical approach to flood inundation mapping using Height Above Nearest Drainage method in conjunction with Synthetic Rating Curves (HAND-SRC) can provide fast and computationally efficient alternative to traditional flood models at continental scale due to their lower complexity and data requirements. However, SRC are based on one-dimensional 1D steady state and normal depth assumption which may introduce errors in estimates of flood stage and extent. This study investigates the reliability of SRC across conterminous Unites States (CONUS) by evaluating the accuracy of these model assumptions. Specifically, SRC are compared to observed rating curves at the United States Geologic Survey (USGS) gauges. Errors due to model assumptions are estimated by accounting for total error in SRC and errors in hydraulic simulations from incomplete bathymetric representation. These errors are then analyzed spatially and correlated with watershed and stream characteristics to map the reliability of HAND-SRC method across CONUS.