When rain falls on a watershed, a diverse array of processes are set in motion, moving water from the canopy, into and across the ground surface, through pores in soil and rock, and eventually into streams as runoff. This event-driven discharge to streams is also termed quickflow. The total streamflow can be divided into quickflow, which reaches streams rapidly through runoff generation processes, and baseflow, which varies gradually and is contributed by slower pathways through the subsurface. Understanding runoff generation has been a keystone in hydrological research, as these canopy-to-stream processes aggregate over watersheds to determine the quantity and timing of flow in rivers, transport of sediment and nutrients, and water available for plant uptake. In addition, measuring runoff generation in the field is important for improving models that forecast floods, water supply, and agricultural conditions. However, runoff generation is difficult to measure directly, as it is highly dependent on processes happening below the ground surface. Some studies have used dense arrays of wells and soil moisture sensors to interpret runoff generation processes, though cost has limited these to very few sites. In an effort to identify the strongest predictors of quickflow volume and provide insight into runoff generation processes, Scaife et al. (2020) examined the quickflow response and runoff generation processes at three small forested watersheds using streamflow and precipitation records along with single timeseries measurements of soil moisture and groundwater level at each site.
Figure 1. Upland terrain at the Coweeta Hydrologic Laboratory. Photo by Greg Zaussen, 2013.
Two of the sites examined (WS2 and WS14) are located within the Coweeta Hydrologic Laboratory in North Carolina, USA, while the third site (SHW) is located in the Shale Hills Critical Zone Observatory in Pennsylvania, USA. The authors identified storms from timeseries of precipitation, and separated quickflow from baseflow using a constant slope method (Figure 2).
Figure 2. Streamflow response to precipitation, showing the constant slope method of separating quickflow from baseflow. (Adapted from Hewlett and Hibbert (1967), Figure 2).
Based on the analysis of Scaife et al. (2020), the strongest predictor of quickflow volume at the three sites was the sum of total precipitation and antecedent soil moisture (ASI; Figure 3). Neither total precipitation alone nor the maximum precipitation intensity were not as strong predictors of quickflow generated in a storm as the combined total precipitation and ASI. The authors suggest that the highly permeable forest soils at these sites are likely responsible for this, as even very high intensity rainfall can easily infiltrate into the soil.
Beyond predicting quickflow response, the authors examined the soil moisture and groundwater level timeseries to gain understanding about the processes generating quickflow. While all sites generated the most quickflow when soil moisture was high, suggesting the importance of shallow saturated and unsaturated flow, the dependence of quickflow on water table elevation varied between sites (Figure 4). The difference in sensitivity to groundwater elevation suggests differences in the role of deeper subsurface pathways in generating runoff.
Figure 3. Total event stormflow (the total quickflow during event period) show threshold relationships with precipitation and an antecedent soil moisture index (ASI) at the three sites. Scaife et al (2020), Figure 4.
Figure 4. Hourly soil moisture and groundwater level during storm events at three sites studied by Scaife et al. (2020). Open and closed circles indicate whether quickflow was or was not present that hour. Colors indicate hours with above the 75th percentile of quickflow, where blue colors have the greatest values. (a) WS2 generates the most quickflow when both water table elevation and soil moisture are high. (b) WS14 generates large volumes of quickflow at high soil moisture regardless of water table elevation. (c) SHW generates the greatest quickflow when the water table is close to the surface. Scaife et al. (2020), Figure 7.
The authors discuss how differences in the coupling between groundwater and soil moisture during storm events indicate the runoff generation mechanisms unique to each site. For example, that quickflow is not strongly dependent on groundwater level at WS14 suggests that flow through soil pipes or along the bedrock-soil boundary may be more important sources of runoff at this site than at SHW or WS2. These results suggest that it may be possible to determine some differences in runoff generation processes between sites with only single timeseries measurements of soil moisture and groundwater level, provided that the locations of these measurements are selected to be representative of what occurs in the whole watershed. Consequently, measurements of soil moisture and groundwater level could be important tools for characterizing and predicting the quickflow response to precipitation at sites with less intensive monitoring.
Charles Scaife is a PhD candidate in the Department of Environmental Sciences at the University of Virginia. He received a Master of Arts degree in geography from the University of North Carolina Chapel Hill in 2015. In his research, he studies climate change impacts on water resources in mountainous environments, with a focus on how vegetation and biodiversity affect the terrestrial water cycle. He is currently a Knauss Marine Policy Fellow with the U.S. Department of Energy.
Reference: Scaife, C. I., Singh, N. K., Emanuel, R. E., Miniat, C. F., & Band, L. E. (2020). Non-linear quickflow response as indicators of runoff generation mechanisms. Hydrological Processes, 34 (13). https://doi.org/10.1002/hyp.13780
Additional in-text citation: Hewlett, J. D., & Hibbert, A. R. (1967). Factors affecting the response of small watersheds to precipitation in humid areas. In Int. Symp. Forest Hydrology. https://doi.org/10.1177/0309133309338118
Comentarios