Spatiotemporal Drivers of Runoff Event Timescales: An Event‐Based Clustering Approach Applied to 388 Catchments in the Contiguous United States

Abstract

Runoff responds differently to rainfall events depending on antecedent wetness conditions and catchment properties. One major characteristic of the response is the runoff event timescale, herein defined as the ratio of event volume to peak runoff rate. Since the event timescale essentially reflects the event runoff shape (e.g., prolonged or flashy), analyzing this metric and its main drivers can potentially help identify fast and slow runoff generation mechanisms in a catchment. Yet, large‐sample studies assessing the spatiotemporal variability of the event timescale and its controls in different climatic regions remain scarce. Here, we quantify how hydro‐meteorological event conditions and physical catchment attributes influence runoff event timescales in catchments with contrasting climates. We conducted an event‐based analysis in 388 snow‐free catchments in the contiguous US, separately for wet and dry climate types based on aridity. We demonstrate that spatiotemporal drivers of the event timescale significantly differ between wet and dry groups. In wet catchments, the temporal variability of the timescale was primarily determined by rainfall event characteristics, particularly rainfall event duration. While rainfall characteristics were important in the dry group as well, antecedent soil moisture was more important in predicting the timescale in dry than in wet catchments. The catchment‐median timescale was influenced by topography, soil and climate attributes. Physical attributes such as catchment area were mostly important in wet catchments, whereas mean potential evapotranspiration had the strongest influence on the median timescale in dry catchments. Our results imply that the potential controls on event timescales vary with catchment climate type.