Estimating dominant runoff modes across the conterminous United States

Abstract

Effective natural resource planning depends on understanding the prevalence of runoff generating processes. Within a specific area of interest, this demands reproducible, straightforward information that can complement available local data and can orient and guide stakeholders with diverse training and backgrounds. To address this demand within the contiguous United States (CONUS), we characterized and mapped the predominance of two primary runoff generating processes: infiltration‐excess and saturation‐excess runoff (IE vs. SE, respectively). Specifically, we constructed a gap‐filled grid of surficial saturated hydraulic conductivity using the Soil Survey Geographic and State Soil Geographic soils databases. We then compared surficial saturated hydraulic conductivity values with 1‐hr rainfall‐frequency estimates across a range of return intervals derived from CONUS‐scale random forest models. This assessment of the prevalence of IE versus SE runoff also incorporated a simple uncertainty analysis, as well as a case study of how the approach could be used to evaluate future alterations in runoff processes resulting from climate change. We found a low likelihood of IE runoff on undisturbed soils over much of CONUS for 1‐hr storms with return intervals <5 years. Conversely, IE runoff is most likely in the Central United States (i.e., Texas, Louisiana, Kansas, Missouri, Iowa, Nebraska, and Western South Dakota), and the relative predominance of runoff types is highly sensitive to the accuracy of the estimated soil properties. Leveraging publicly available data sets and reproducible workflows, our approach offers greater understanding of predominant runoff generating processes over a continental extent and expands the technical resources available to environmental planners, regulators, and modellers.

Random Forest Ability in Regionalizing Hourly Hydrological Model Parameters

“…Although the RF algorithm has demonstrated its outperformance in comparison with other ML algorithms, and despite its application to many problems in several environmental sciences [34][35][36], its use in water sciences is still limited [30]. Examples of RF applications in hydrology include precipitation downscaling [37,38], flood prediction and risk assessment [39][40][41], estimating runoff modes or hydrological signatures on a continental scale and predicting flow regimes [42][43][44][45][46] as well as predicting flow characteristics at ungauged locations [47][48][49]. RFs are constructed by growing a number of regression and classification trees.…”

Possible Increases in Flood Frequency Due to the Loss of Eastern Hemlock in the Northeastern United States: Observational Insights and Predicted Impacts

“…Rogger et al () suggest that the influence of forest cover on flooding frequency may be better estimated with methodologies that isolate the physical mechanisms by which land cover partitions infiltration and surface runoff. Hydrological land surface model development for prediction of hydrologic extremes typically maintains a strong focus on infiltration mechanisms (e.g., Buchanan et al, ) with less emphasis on capturing the complexity of plant dynamics (e.g., Giuntoli et al, ), often neglecting to properly represent the functional traits that govern plant hydraulic regulation (Matheny et al, ). Perhaps, as a result, relatively few model‐based studies have been conducted examining the role of forest cover on runoff.…”

Going Beyond the Ensemble Mean: Assessment of Future Floods From Global Multi‐Models

“…If on the one hand GCMs are responsible for regional runoff biases due to uncertainties in the representation of precipitation and sub‐grid soil infiltration and flow; on the other hand the GIMs' total runoff include contributions from surface runoff—function of saturation (SE) and infiltration excess (IE)—and subsurface runoff—function of impermeable area and water table depth (Kooperman et al., 2018). For instance, throughout the domain of study portions of Texas, Louisiana, Kansas, Missouri, and Iowa are more likely dominated by IE runoff; on the other hand SE runoff is more likely in the southeast (e.g., Florida, south Georgia) and coastal areas of the Great Lakes region (Buchanan et al., 2018). The prevalence of IE or SE excess runoff depends on the type of soil and its capacity to become saturated/infiltrate.…”