An official website of the United States government
Abstract Integrated hydrological modeling is an effective method for understanding interactions between parts of the hydrologic cycle, quantifying water resources, and furthering knowledge of hydrologic processes. However, these models are dependent on robust and accurate datasets that physically represent spatial characteristics as model inputs. This study evaluates multiple data‐driven approaches for estimating hydraulic conductivity and subsurface properties at the continental‐scale, constructed from existing subsurface dataset components. Each subsurface configuration represents upper (unconfined) hydrogeology, lower (confined) hydrogeology, and the presence of a vertical flow barrier. Configurations are tested in two large‐scale U.S. watersheds using an integrated model. Model results are compared to observed streamflow and steady state water table depth (WTD). We provide model results for a range of configurations and show that both WTD and surface water partitioning are important indicators of performance. We also show that geology data source, total subsurface depth, anisotropy, and inclusion of a vertical flow barrier are the most important considerations for subsurface configurations. While a range of configurations proved viable, we provide a recommended Selected National Configuration 1 km resolution subsurface dataset for use in distributed large‐and continental‐scale hydrologic modeling.
more »
« less
Water Table Depth Estimates over the Contiguous United States Using a Random Forest Model
Abstract Water table depth (WTD) has a substantial impact on the connection between groundwater dynamics and land surface processes. Due to the scarcity of WTD observations, physically‐based groundwater models are growing in their ability to map WTD at large scales; however, they are still challenged to represent simulated WTD compared to well observations. In this study, we develop a purely data‐driven approach to estimating WTD at continental scale. We apply a random forest (RF) model to estimate WTD over most of the contiguous United States (CONUS) based on available WTD observations. The estimated WTD are in good agreement with well observations, with a Pearson correlation coefficient (r) of 0.96 (0.81 during testing), a Nash‐Sutcliffe efficiency (NSE) of 0.93 (0.65 during testing), and a root mean square error (RMSE) of 6.87 m (15.31 m during testing). The location of each grid cell is rated as the most important feature in estimating WTD over most of the CONUS, which might be a surrogate for spatial information. In addition, the uncertainty of the RF model is quantified using quantile regression forests. High uncertainties are generally associated with locations having a shallow WTD. Our study demonstrates that the RF model can produce reasonable WTD estimates over most of the CONUS, providing an alternative to physics‐based modeling for modeling large‐scale freshwater resources. Since the CONUS covers many different hydrologic regimes, the RF model trained for the CONUS may be transferrable to other regions with a similar hydrologic regime and limited observations.
more »
« less
- PAR ID:
- 10473641
- Publisher / Repository:
- Wiley
- Date Published:
- Journal Name:
- Groundwater
- ISSN:
- 0017-467X
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Shallow groundwater in the Prairie Pothole Region (PPR) is recharged predominantly by snowmelt in the spring and may supply water for evapotranspiration through the summer/fall. This two-way exchange is underrepresented in land-surface models. Furthermore, the impacts of climate change on the groundwater recharge are uncertain. In this paper, we use a coupled land and groundwater model to investigate the hydrologic cycle of shallow groundwater in the PPR and study its response to climate change at the end of the 21st century. The results show that the model reasonably simulates the water table depth (WTD) and the timing of recharge processes, but underestimates the seasonal variation of WTD, due to mismatches of the soil types between observations and the model. The most significant change under future climate occurs in the winter, when warmer temperature changes the rain/snow partitioning, delay the time for snow accumulation/soil freezing while bring forward early melting/thawing. Such changes lead to an earlier start to a longer recharge season, but with lower recharge rates. Different signals are shown in the eastern and western PPR in the future summer, with reduced precipitation and drier soils in the east but little change in the west. The annual recharge increased by 25% and 50% in the eastern and western PPR, respectively. Additionally, we found the mean and seasonal variation of the simulated WTD are sensitive to soil properties and fine-scale soil information is needed to improve groundwater simulation on regional scale.more » « less
-
Abstract This study synthesizes two different methods for estimating hydraulic conductivity (K) at large scales. We derive analytical approaches that estimate K and apply them to the contiguous United States. We then compare these analytical approaches to three‐dimensional, national gridded K data products and three transmissivity (T) data products developed from publicly available sources. We evaluate these data products using multiple approaches: comparing their statistics qualitatively and quantitatively and with hydrologic model simulations. Some of these datasets were used as inputs for an integrated hydrologic model of the Upper Colorado River Basin and the comparison of the results with observations was used to further evaluate the K data products. Simulated average daily streamflow was compared to daily flow data from 10 USGS stream gages in the domain, and annually averaged simulated groundwater depths are compared to observations from nearly 2000 monitoring wells. We find streamflow predictions from analytically informed simulations to be similar in relative bias and Spearman's rho to the geologically informed simulations.R‐squared values for groundwater depth predictions are close between the best performing analytically and geologically informed simulations at 0.68 and 0.70 respectively, with RMSE values under 10 m. We also show that the analytical approach derived by this study produces estimates of K that are similar in spatial distribution, standard deviation, mean value, and modeling performance to geologically‐informed estimates. The results of this work are used to inform a follow‐on study that tests additional data‐driven approaches in multiple basins within the contiguous United States.more » « less
-
Abstract Floodplains are essential ecosystems that provide a variety of economic, hydrologic, and ecologic services. Within floodplains, surface water‐groundwater exchange plays an important role in facilitating biogeochemical processes and can have a strong influence on stream hydrology through infiltration or discharge of water. These functions can be difficult to assess due to the heterogeneity of floodplains and monitoring constraints, so numerical models are useful tools to estimate fluxes, especially at large spatial extents. In this study, we use the SWAT+ (Soil and Water Assessment Tool) ecohydrological model to quantify magnitudes and spatiotemporal patterns of floodplain surface water‐groundwater exchange in a mountainous watershed using an updated version of thegwflowmodule that directly calculates floodplain‐aquifer exchange rates during periods of floodplain inundation. Thegwflowmodule is a spatially distributed groundwater modelling subroutine within the SWAT+ code that uses a gridded network and physically based equations to predict groundwater storage, groundwater head, and groundwater fluxes. We used SWAT+ to model the 7516 km2Colorado River headwaters watershed and streamflow data from USGS gages for calibration and testing. Models that included floodplain‐groundwater interactions outperformed those without such interactions and provided valuable information about floodplain exchange rates and volumes. Our analyses on the location of floodplain fluxes in the watershed also show that wider areas of floodplains, “beads” (e.g., like beads on a necklace), exchanged a higher net and per area volume of water, as well as higher rates of exchange, compared to narrower areas, “strings.” Study results show that floodplain channel‐groundwater exchange is a valuable process to include in hydrologic models, and model outputs could inform land conservation practises by indicating priority locations, such as beads, where substantial hydrologic exchange occurs.more » « less
-
Groundwater and surface water are interconnected in most climatic regions. Baseflow, the contribution of streamflow not directly associated with precipitation forcing, is a critical component of streamflow prediction and water resource allocation. Baseflow is often considered to be a low-frequency component of streamflow and many of the methods for estimating it are based on this premise. The climatic and physiographic attributes of a region will contribute to the low-flow behavior of its surface waterways. For example, baseflow in a snowmelt-driven basin may produce a distinct hydrologic signature compared to baseflow in a precipitation-driven basin.In this study, we developed a unique metric based on the variable drought threshold method (VDTM) for characterizing historical streamflow timeseries and performed cluster analysis on a large set of gages in the continental United States (CONUS). Our study goal was to observe correlations between low-flow characteristics and distinct hydrologic, physiographic, and climatic regions to provide insight into the underlying mechanisms influencing baseflow.The VDTM applies a non-exceedance percentile (NEP) computed based on the distribution of flow recorded at a stream gage over a given time frame (i.e., month, season) throughout the complete record of measurement. This study used daily streamflow records for 1,462 reference quality gages across the CONUS from the USGS GAGES-II data set; each gage contained at least 20 years of complete daily streamflow measurements. We computed the 10th NEP for each month at all 1,462 gages and normalized this value by the mean streamflow to develop the parameter r10. We performed K-means clustering on the monthly r10 values, forming seven clusters of low-flow behavior.We observed clusters with distinct low-flow behavior across different ecoregions related to possible mechanisms driving streamflow and baseflow in those regions. For example, a cluster located in the intermountain-west shows unique behavior largely seen nowhere else in the CONUS, possibly a result of the predominantly snowmelt-driven shallow subsurface flow that contributes to baseflow seen in that region. Conversely, clusters located in the Pacific Northwest and parts of the Appalachians show a different behavior, possibly a result of the predominantly rainfall-driven streamflow observed in those regions. Principal components analysis suggests that the critical months associated with clustered gages are during the summer (June, July) and winter (January, February).The spatial distribution of the clusters largely adheres to the defined physiographic and climatic regions of the CONUS despite the absence of any physiographic or climatic variables used for clustering, suggesting a possible linkage between these attributes and the low-flow behavior of surface waterways. Analysis of the trend and magnitude of r10 may provide insight into whether (and when) a stream is losing water to or gaining water from groundwater as well as the magnitude of the transfer. The results of this study suggest that using NEPs and the r10 metric may be an effective method for defining regionalization based on low-flow metrics.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Journal Article:
- https://doi.org/10.1111/gwat.13362
