Watershed

Oral

393820 - The Role of Rainfall Space-Time Structure and Resolution in Flood

Author(s)

• ZZ

  Zhihua Zhu

  Sun-Yat Sen University; University of Wisconsin-Madison

Co-Authors: Guo Yu, 1261 Engineering Hall (University of Wisconsin-Madison), Madison, WI, United States – University of Wisconsin-Madison; Daniel Wright, 1269C Engineering Hall (University of Wisconsin-Madison), Madison, WI, United States – University of Wisconsin-Madis; Ricardo Mantilla, Department of Civil and Environmental Engineering The University of Iowa, Iowa City, IA, 52242 – The University of Iowa

Extreme rainfall can vary dramatically in space and time, often with important impacts on resulting hydrologic response. How rainfall space-time variability (i.e. structure) influences flood frequency is a major unknown, however, due to severe limitations of existing observational records and stochastic modeling alternatives. In hydrologic engineering practice, flood frequency analyses based on stream gages bypass rainfall entirely, while design storm methods use highly idealized asssumptions regarding rainfall space-time structure. This study seeks to shed light on how flood frequency is shaped by rainfall space-time structure.

We use the RainyDay open-source Stochastic Storm Transposition (SST) software to generate large numbers of extreme rainfall scenarios based on a 15-year record of Stage IV radar rainfall. We have previously shown that, when combined with the distributed Hillslope Link Model, RainyDay can reproduce flood frequency estimates across a wide range of recurrence intervals and watershed scales. We will examine rainfall scenarios at four resolutions (4 km hourly; 4 km daily; basin-average hourly; and basin-average daily) to see how rainfall resolution, and, by extension, space-time structure, influence flood frequency estimates. We focus on the Turkey river basin, northeastern Iowa, USA and its subwatersheds, ranging in size from 900 to 4000 km2, and analyze the peak discharges for recurrence intervals of 10 to 500 years. Based on these findings, we make scale-dependent recommendations on the importance of rainfall space-time variability in flood frequency and what rainfall resolutions (and thus what rainfall data sources) are suitable for flood frequency analysis.