Controls on River Flashiness Across the Continental United States

Examining simple hydrologic signatures of river discharge at a continental scale can highlight drivers of runoff generation and generate hypotheses for future investigation. A widely used hydrologic signature is the Richards-Baker Index (RBI), which represents the flashiness of the discharge record of a river. Previous analyses have been performed at watershed to regional scales, but analyses at broader scales are currently lacking. In this study we calculated RBI using daily streamflow values for 1,144 USGS streamflow gages across the continental United States from 1999-2010. The sites used in the study were chosen to span a broad range of land cover and to have minimal influence from upstream dams. To explore empirical relationships between this hydrologic signature and watershed form, we compared RBI for these gages to a variety of watershed characteristics using stepwise multivariate regressions. Across all sites, factors most predictive of the flashiness of a river were metrics related to land use (e.g., developed, water, forest, and wetland land cover) and climate (proportion of precipitation falling as snow, potential evapotranspiration). Surprisingly, while increasing drainage areas for heavily urbanized watersheds resulted in reduced flashiness, this pattern was not present for reference watersheds. Furthermore, we found that significant predictors of RBI changed when watersheds were binned by size. The most successful regressions for small watersheds had more significant factors, including land use, climate, and basin size, whereas larger watersheds had fewer, and mostly climate related factors. Our analysis offers insights into what factors control stream flashiness across the continental United States and how these factors are governed by watershed size. A better understanding of controls on flashiness can help iterate our understanding of watershed response and relationships to watershed form in the context of how quickly (or slowly) watersheds respond to precipitation. Such information is key for benchmarking the different filters watersheds provide on precipitation across diverse landscapes.