Regional regression models of percentile flows for the contiguous United States: Expert versus data-driven independent variable selection

How is Baseflow Index (BFI) impacted by water resource management practices?

“…Baseflow, defined as streamflow fed from the deep subsurface and shallow subsurface storage between precipitation and/or snowmelt events (Tallaksen, 1995;Price, 2011;Zhang et al, 2017;Singh et al, 2019;Gnann et al, 2019), is a hydrological phenomenon that represents a whole catchment response to meteorological and other environmental signals (Bloomfield et al, 2011). It is important as it sus-Published by Copernicus Publications on behalf of the European Geosciences Union.…”

“…tains surface flows particularly during relatively dry periods and droughts (Smakhtin, 2001;Miller et al, 2016) because it supports ecological flows and ecosystem functioning (Poff et al, 1997;Boulton, 2003) and is a factor in regulating streamflow quality and temperature (Jordan et al, 1997;Gomez-Velez et al, 2015;Hare et al, 2021). It integrates the outcomes of a wide range of natural and human-influenced surface and subsurface catchment processes (Price et al, 2011;Gnann et al, 2019) that include geomorphological controls related to surface topography (Santhi et al, 2008) and soil processes (Vivoni et al, 2007;Price et al, 2011;Singh et al, 2019;Yao et al, 2021) and (hydro)geological processes that control baseflow (Longobardi and Villani, 2008;Bloomfield et al, 2009;Kuentz et al, 2017;Carlier et al, 2018). Land use and land cover (LULC) change may also have profound effects on baseflow generation (Zhang and Schilling, 2006;Wang et al, 2014), including effects of changing forest cover and agriculture (Juckem et al, 2008;Ahiablame et al, 2017;Zhang et al, 2017) and urbanization (Simmons and Reynolds, 1982;Chang, 2007;Dow, 2007;McGrane, 2016).…”

“…Despite this extensive work on baseflow generation dynamics, Gnann et al (2019) observed that there is still no general theory to explain variations in baseflow between catchments despite the strong evidence that it is largely controlled by the interaction of climate and landscape processes. They explored the role of climate in baseflow generation using baseflow data from the United States of America (USA) and the United Kingdom (UK) and found that in humid settings baseflow can be highly variable due to variations in catchment storage and wetting potential, whereas in more arid settings baseflow has much lower variability and is primarily controlled by vaporization limits.…”

“…They found that impacts of WRM were experienced during periods of low flows when the contribution of groundwater through baseflow is the largest and that return flows changed the timing and duration of the low-flow periods, causing baseflow to be maintained for longer. In summary, as with natural controls on baseflow (Gnann et al, 2019), there is as yet no general theory to explain the effects of WRM practices on baseflow, and the effect of a given WRM practice on baseflow may be contingent on a range of factors including climate, (hydro)geological setting, location, and timing of the activity.…”

“…This is despite new opportunities being offered to investigate and quantify catchment processes through open access, large-sample hydrology datasets (Addor et al, 2020). Such datasets have been used to provide insights into catchment processes and functioning across multiple climate and catchment settings (Beck et al, 2013;Ochoa-Tocachi et al, 2016;Fouad et al, 2018;Gnann et al, 2019;Dudley et al, 2020). CAMELS-GB, a recently published large-sample hydrology dataset for Great Britain (GB) (Coxon et al, 2020a;, is unusual in that it contains quantitative information on WRM practices including surface water and groundwater abstractions, discharges, and reservoir numbers and capacities at the catchment scale.…”