1. Introduction
Shallow groundwater has been shown to influence 22%–32% of the global land area (Fan et al. 2013). For the majority of these regions, precipitation and evapotranspiration (ET) profoundly affect the unsaturated storage and hence the groundwater table depths. Many regional numerical modeling studies with inclusion of groundwater table dynamics have shown that it generally increases the available soil moisture for ET, and hence increases the latent heat flux, leading to lower temperatures and increased humidity levels in the atmospheric boundary layer (Leung et al. 2011; Shrestha et al. 2014; Keune et al. 2016; Sulis et al. 2018). A recent global modeling study by Wang et al. (2018) showed that prescribed constant shallow water table depth increases ET over water-limited regimes, but decreases ET over energy-limited regimes due to the corresponding increase in cloud cover, which reduces downwelling radiation. In fact, clouds generally modulate the amount of solar radiation reaching the ground, and hence indirectly attenuate the ET. This has a direct effect on the dynamics of the shallow groundwater table for unconfined aquifers. Here, shallow groundwater is defined as groundwater levels that fluctuate within the root zone (<3-m depth measured from surface). For temperate climate, ET has been shown to play an important factor on the diurnal fluctuation of shallow groundwater levels (Gribovszki et al. 2010). At the diurnal scale, both precipitating and nonprecipitating daytime clouds can have indirect controls on the shallow groundwater dynamics. Nonprecipitating daytime clouds indirectly reduce the ET and increase the groundwater table (GWT) depth. In addition to this, precipitating daytime clouds decrease the GWT depths via an increase in infiltration. Depending on the climatology of these regions with shallow GWT, clouds could potentially influence the GWT dynamics at seasonal and interannual scales. On the other hand, vegetation with different physiology exhibits a strong local control on ET and infiltration after interception. For example, transpiration can account for almost 35%–90% of total ET depending on vegetation cover (Schlesinger and Jasechko 2014; Good et al. 2015; Maxwell and Condon 2016; Fatichi and Pappas 2017; Shrestha et al. 2018b). Hence, daytime clouds and vegetation combined can exert a significant control on the dynamics of shallow GWT depths along with other nonlocal effects.
Where observations are sparse, creating spatial and temporal gaps, studies using physically based groundwater models with integrated surface and groundwater lateral flows have mostly focused on the sensitivity of GWT depth on surface energy flux partitioning (e.g., Kollet and Maxwell 2008; Maxwell and Kollet 2008; Shrestha et al. 2015; Maxwell and Condon 2016; Shrestha et al. 2018a). Recently, a new study by Condon et al. (2020) has also explored the impact of shallow groundwater on climate warming scenarios, suggesting that it may buffer plant water stress to certain degree. However, little attention has been given to the impact of clouds and vegetation together on shallow GWT dynamics. In this study, a kilometer-scale hydrological model with surface–groundwater interactions over a temperate climate zone (northwestern part of Europe) is used to quantify the impact of daytime clouds and evapotranspiration on the modeled shallow GWT dynamics. The model is comprehensively evaluated with observations available over the region. Then, the model output is further analyzed to better understand the spatiotemporal variation of GWT depth over the region and directly evaluate its linkages with incoming solar radiation, precipitation and vegetation covers.
2. Methods
a. Model
The hydrological component of Terrestrial Systems Modeling Platform (TerrSysMP or TSMP; Shrestha et al. 2014; Gasper et al. 2014) consists of the NCAR Community Land Model CLM3.5 (Oleson et al. 2008) and the 3D variably saturated groundwater and surface water flow code ParFlow (Ashby and Falgout 1996; Jones and Woodward 2001; Kollet and Maxwell 2006; Maxwell 2013). The two models are coupled using the OASIS3-MCT coupler (Craig et al. 2017). The coupled model simulates both local and nonlocal controls on soil moisture and surface energy fluxes via explicit biogeophysical processes for 16 different plant functional types (PFTs) and interaction between surface and groundwater flow. A detailed discussion about the coupling is available in Shrestha et al. (2015).
The model, however, does not account for anthropogenic impacts (e.g., drainage, pumping, irrigation, etc.). Various research efforts are ongoing to integrate anthropogenic influence in models, which is a new frontier (e.g., Wada et al. 2012, 2016; Siebert et al. 2010; Siebert and Döll 2010). In a global sensitivity study with and without the human-induced change, Wada et al. (2016) shows that human impacts are limited for simulated total water storage for the Rhine basin compared to other major basins of the world. Further, since CLM3.5 does not have an urban land-use type, the urban area is simulated with PFT type 16 with fixed leaf area index (LAI), increased roughness height, and changes in soil water potential at full stomatal opening and closure. This simplified parameterization is designed to improve the surface energy flux partitioning over urban areas in the absence of a state of art urban canopy model and does not compensate for urban–rural surface temperature differences associated with population and infrastructure dynamics (Manoli et al. 2019).
b. Domain
The experiment is set up over a temperate region in the northwestern part of Germany bordering the Netherlands, Luxemburg, Belgium, and France. The region is characterized with multiple hills of the Rhine Massif with heights ranging from 600 to 800 m, and land cover including forest, agricultural land, and urban/rural area. Due to the availability of the twin polarimetric X-band research radars in Bonn (BoXPol) and Jülich (JuxPol) and overlapping measurements from four polarimetric C-band radars operated by the German Weather Service (Deutscher Wetterdienst, DWD), the region probably represents the best radar-monitored area in Germany (hereafter referred as Bonn radar domain).
The hydrological model domain covers approximately 333 × 333 km2 area with a horizontal grid resolution of 1.132 km (Fig. 1). The model domain has 30 vertical levels with 10 stretched layers in the root zone (2–100 cm) and 20 constant levels (135 cm) extending to 30 m below the surface. The TSMP Preprocessing and Postprocessing System (TPS; Shrestha 2019) is used to remap/interpolate raw data and generate the input data for the model. The land surface data including vegetation cover and phenology were processed from Moderate Resolution Imaging Spectroradiometer (MODIS) remote sensing products (Myneni et al. 2015; Friedl and Sulla-Menashe 2019). The root-zone soil texture data were obtained from the DWD model analysis files. The soil type in these data is processed from the Digital Soil Map of the World (DSMW) available from FAO (Food and Agriculture Organization of the United Nations). The spatial pattern of soil textures below the root zone was obtained from the International Hydrogeological Map of Europe (IHME1500, https://www.bgr.bund.de/EN/Themen/Wasser/Projekte/laufend/Beratung/Ihme1500/ihme1500_projektbeschr_en.html). The slopes (associated with D4 method of flow directions) for the groundwater model were obtained using topography from Shuttle Radar Topography Mission (SRTM) data (Farr et al. 2007).
The spatial pattern of topography, plant functional types (PFTs), root-zone soil texture, and below root-zone soil texture for the Bonn radar domain. The major river networks processed from the SRTM topography and the extent of the BoXPol radar are also shown. PFTs: bare soil, bs; needle leaf evergreen trees, nle; broadleaf evergreen trees, ble; broadleaf deciduous trees, bld; grasslands, c3g; agricultural land, c1n; urban areas, urb. Root-zone soil texture: s-loam, sandy loam; c-loam, clay loam. Below root-zone soil texture: highly porous aquifers, HPA-1; low/moderate productive porous aquifers, LMPA-1; highly porous fissured aquifers, HPA-2; low/moderate productive fissured aquifers, LMPA-2; locally aquiferous rocks, LAR-3; practically nonaquiferous rocks, PNAR-3.
Citation: Journal of Hydrometeorology 22, 4; 10.1175/JHM-D-20-0171.1
The topography is dominated by the Rhine Massif and low-lying flat lands over northwestern part of the domain. The Middle Rhine Valley divides the Rhine Massif into two: Ardennes, Eifel, and Hunsrück to the west of the Rhine and Süder Uplands, Westerwald, and Taunus to the east of the Rhine. Within the domain, land cover is comprised of forested areas (58%), agricultural land (23%, c1n), urban areas (12%, urb), and grasslands (7%, c3g). The forested area contains needleleaf (nle), broadleaf evergreen (ble), and broadleaf deciduous (bld) trees. The root-zone soil texture is mostly dominated by loamy soil with patches of sand, sandy loam, and clay loam. The pattern of soil texture below the root zone matches well with the topography with valleys dominated by highly productive porous aquifers (HPA-1) and practically nonaquiferous rocks (PNAR-3) over the hills. Low/moderate productive porous and fissured aquifers (LMPA-1, LMPA-2), highly productive fissured aquifers (HPA-2), and locally aquiferous rocks (LAR-3) are spread out over the domain as well. The soil hydraulic parameters listed in Table 1 are determined based on literature (Schaap and Leij 1998; Maxwell and Condon 2016).
Soil hydraulic parameters for the Bonn radar domain.
c. Simulations
The atmospheric forcing data for the model were processed from the German weather forecast model COSMO-DE (Consortium for Small-Scale Modeling–Germany domain; Baldauf et al. 2011) analysis data available from DWD (https://www.dwd.de/DE/leistungen/pamore/pamore.html). The analysis has a spatial resolution of 2.8 km and hourly temporal resolution. The model has built in algorithm, which maps the atmospheric forcing data from their native resolution (coarse) to model resolution (fine) by finding every fine grid cell within a coarse grid cell and computing the weights depending on area of cell overlaps.
A 10-yr period from 2008 to 2017 was chosen for the study, which also coincides with the TR32 project with a test bed over the region (Simmer et al. 2015). The model simulation is used to advance our understanding of the physical processes and provide consistent initial conditions for hindcast simulations using the fully coupled model over the region. As such, the model has not been calibrated and uses default parameters based on the land surface data. Besides, the use of modeled forcing data has uncertainty and biases (e.g., Böhme et al. 2011; Lindau and Simmer 2013). Based on 2 years of data (2007–08), Böhme et al. (2011) found that the COSMO-DE precipitation exhibited a wet bias of 20% during winter, mainly over orographic regions. Similarly, Lindau and Simmer (2013) also found that the regional climate version of COSMO model had overall wet bias of 26% over Germany (1960–2000). Also, previous studies by Shrestha et al. (2015, 2018a) have shown that the nonlocal controls of soil moisture are highly grid resolution dependent, and coarsening the grid resolution from 100 to 1000 m shifts the GWT depth distribution toward shallower depths. This is a limitation in this study, as hyper-resolution simulations require large computational resources that were not available here.
The model spinup was conducted for 4 years, using the 2008 atmospheric forcing data recursively. The spinup soil–vegetation state refers to equilibrium model state, which is defined using thresholds for energy and water balance. For energy balance, a threshold of 0.1 W m−2 was used for annual average difference in surface energy fluxes. Similarly, for water balance, a threshold of 0.1% was used for 1-yr difference in total water storage.
The model was then initialized with spinup soil–vegetation states, and a 10-yr transient run from 2008 to 2017 was conducted using the offline atmospheric forcing data. The vegetation phenology (consisting of monthly data) was updated yearly based on the MODIS remote sensing product. The model was integrated at hourly frequency, and the outputs were generated at 5-day intervals.
d. Observations
The near-surface soil moisture from the Soil Moisture and Ocean Salinity (SMOS; Kerr et al. 2001) mission and terrestrial water storage anomaly from the Gravity Recovery and Climate Experiment (GRACE) provide valuable long-term data for bulk comparison with the model. The monthly aggregated Level 3 near-surface soil moisture data (CATDS 2016; Jacquette et al. 2010; Kerr et al. 2013, 2016; Al Bitar et al. 2017) at 25-km spatial resolution from 2010 to 2017 were downloaded from https://www.catds.fr/sipad/. The Level 3 SMOS data are only available from 2010 onward. Similarly, the Level 3 monthly GRACE land data (Swenson and Wahr 2006; Swenson 2012; Landerer and Swenson 2012) from 2008 to 2017 at spatial resolution of 1.0° (~111 km at the equator) were downloaded from http://grace.jpl.nasa.gov. The model and satellite data are processed to obtain monthly area averages over the Bonn radar domain, and the anomalies are then obtained relative to a consistent base period from January 2010 to December 2015 (when data from both satellites are available) for comparisons.
Additionally, available in situ observations of surface energy fluxes, streamflow, and groundwater table depth (see online supplemental material) over the modeled domain are also used for model evaluation.
3. Results
a. Precipitation and incoming solar radiation
Clouds and precipitation are main source of uncertainties in weather prediction and climate simulations. While precipitation is an important forcing for integrated hydrological models, the modulation of incoming solar radiation by clouds also affects the available energy for evapotranspiration and root-water uptake. Figure 2 shows the time series of domain averaged annual accumulated precipitation
Time series of spatially averaged annual precipitation
Citation: Journal of Hydrometeorology 22, 4; 10.1175/JHM-D-20-0171.1
The anomaly in the 10-yr average spatial patterns of precipitation
Spatial anomalies of 10-yr average precipitation and incoming solar radiation over the Bonn radar domain. The solid and dashed contour lines indicate the positive and negative anomalies, respectively. The contours are overlaid on the topography of the region.
Citation: Journal of Hydrometeorology 22, 4; 10.1175/JHM-D-20-0171.1
The Süder Uplands and northern part of the domain exhibit negative anomalies for incoming solar radiation (Fig. 3b), while the southeastern part of the domain exhibit positive anomalies. In general, the forested hill ranges over the Süder Uplands, which nominally receive the highest precipitation, have relatively lower incoming solar radiation. But, the low-lying flat lands in the northwestern region exhibit negative anomalies for both precipitation and incoming solar radiation, suggesting lower precipitation efficiency.
b. Model evaluation
Figure 4a shows the comparison between simulated and observed terrestrial water storage anomaly from GRACE. The monthly GRACE land data for each grid represents the anomaly for that month relative to the baseline average over January 2004–December 2009. For comparison with model data, the monthly anomalies were recomputed over the region (4.5°–9.5°E, 49.5°–52.5°N) relative to the new base period (January 2010–December 2015). Additionally, for Level-3 GRACE land data, retrievals are available from NASA’s Jet Propulsion Laboratory (JPL), The University of Texas Center for Space Research (CSR), and Deutsches GeoForschungsZentrum (GFZ). The three retrievals vary due to the choices of solution strategy. A study by Sakumura et al. (2014) has suggested using the ensemble mean to reduce the noise in the solution, but here the model data are separately compared with all the three solutions, to show the uncertainty in observations. The anomaly for the region exhibits a strong seasonal cycle in both simulated and observed data. The total water storage starts to increase from autumn through winter, reaches a maximum over early spring and then decreases, reaching minimum at the end of the summer. In comparison to GRACE observations, the model underestimates the decrease in total water storage for 2008 and 2009 while it underestimates the increase for 2013. But in general, both the model and GRACE data exhibit similar seasonal anomalies from 2010 to 2017.
(a) Comparison of area averaged monthly total water storage anomaly between TSMP and Level 3 GRACE land data from CSR, JPL, and GFZ. (b) Comparison of area averaged monthly soil moisture anomaly between TSMP and Level 3 SMOS data for both ascending and descending orbits. The monthly area averaged anomalies are with reference to base period from January 2010 to December 2015.
Citation: Journal of Hydrometeorology 22, 4; 10.1175/JHM-D-20-0171.1
Figure 4b shows the model simulated near-surface soil moisture (0–5 cm) anomaly comparison with the SMOS data. As the surface soil moisture from satellites and models can exhibit different statistical characteristics (Reichle and Koster 2004), the comparison of more consistent information like soil moisture anomaly is presented here. The anomaly for SMOS data was obtained over the region (4.80°–9.21°E, 49.15°–52.22°N) relative to the same base period (2004–09) as for total water storage anomaly. The soil moisture anomaly for both model and satellite data also exhibits a strong seasonal cycle, with a decrease in summer and an increase in winter. For the time period of available SMOS data (2010 onward), the model simulates similar soil moisture anomaly as observed by SMOS, within its range of variability for ascending and descending orbits. Since SMOS data are available at higher resolution compared to GRACE, the correlations coefficient was also computed for all grid points in SMOS data over the region. For SMOS-A, 76% and 17% of data exhibited statistically significant (p < 0.05) weak (r = 0.2–0.4) and moderate (r = 0.5–0.6) correlation, while remaining data were either missing or had statistically insignificant correlation. Similarly, for SMOS-D, 61% and 39% of data exhibited statistically significant (p < 0.05) weak and moderate correlation, respectively. The variability in ascending and descending orbits for SMOS is mainly introduced by the difference in the impact of radio frequency interferences and sun corrections dependent on the geographical location (Al Bitar et al. 2017).
The above bulk comparison suggests that the model is generally able to capture the monthly anomalies of near-surface soil moisture and terrestrial water storage for most of the years. Additionally, the model comparison with in situ observations of surface energy flux, streamflow and groundwater table depth (Figs. S1–S5 in the online supplemental material) also show that the model is able to capture the observed patterns, which adds confidence in the model results.
c. Groundwater table depth
The GWT depth exhibits a strong seasonal cycle corresponding to the incoming solar radiation and plant phenology (Fig. 5). The domain averaged monthly GWT depth anomaly
Time series of modeled and observed domain average monthly GWT depth anomaly for regions with shallow GWT depth. The monthly anomaly for model is color coded for different seasons of the year.
Citation: Journal of Hydrometeorology 22, 4; 10.1175/JHM-D-20-0171.1
The model exhibits high magnitude of interannual variability in spring, summer, and autumn relative to winter season (based on standard deviation of seasonal data, not reported here). Example,
Figure 6 shows the spatial pattern of 10-yr average fluctuation in GWT depth anomaly
The 10-yr average fluctuations in GWT depth anomaly over the Bonn radar domain and their frequency distribution. The spatial patterns are categorized into four zones based on the spatial anomalies of precipitation and incoming solar radiation.
Citation: Journal of Hydrometeorology 22, 4; 10.1175/JHM-D-20-0171.1
Spatially averaged fluctuation in GWT depth anomaly in four zones for different PFTs (bare soil, bs; needleleaf evergreen trees, nle; broadleaf evergreen trees, ble; broadleaf deciduous trees, bld; grasslands, c3g; agricultural land, c1n; urban areas, urb). The differences in the fluctuation of GWT depth anomaly in the four zones are significant (p < 0.05). The bold font highlights the average for each zone.
Within each zone,
For all zones, PFTs or vegetation distribution plays an important role in the amount of root water uptake via evapotranspiration and the corresponding
Seasonal cycles of GWT depth anomaly
Citation: Journal of Hydrometeorology 22, 4; 10.1175/JHM-D-20-0171.1
While vegetation cover modulates ET and root water uptake, thereby directly modulating the shallow GWT depth, clouds directly modulate the available energy in the ground for surface energy flux partitioning and thereby indirectly controlling the fluctuation in shallow GWT depth. Figure 8 shows the scatterplot between monthly anomalies of domain averaged GWT depth
Scatterplot between anomalies of monthly averaged GWT depth and monthly accumulated incoming solar radiation (non-rain and rain affected). Also shown is the linear regression fit including the 95% confidence and prediction intervals. The monthly data are colored to represent different seasons of the year. The incoming solar radiation (non-rain/rain affected) was accumulated over a monthly time scale and spatially averaged over the Bonn radar domain for periods when precipitation was absent/present. The 10-yr averaged seasonal cycle was removed from the monthly time series to generate the anomalies.
Citation: Journal of Hydrometeorology 22, 4; 10.1175/JHM-D-20-0171.1
The
4. Discussion
The modeled shallow GWT depth over the Bonn Radar domain exhibits a strong seasonal cycle and compares well with observed shallow GWT depth anomalies. This is associated with increase in diurnal fluctuation of shallow GWT depth, which is strongly influenced by increase in ET with replenishment of the depleted groundwater storage possible through a local hydraulic gradient in the saturated zone (Kollet and Maxwell 2008; Gribovszki et al. 2010; Rahman et al. 2014). Kollet and Maxwell (2008) proposed a critical GWT depth (1–5 m) where the GWT dynamics exerted a strong control on surface energy fluxes, based on integrated hydrological modeling for Little Washita watershed in Oklahoma, U.S. The shallow GWT depth (<3 m) dynamics, explored in this study, also lies within this critical depth. The ET starts to increase during the growing season, with corresponding increase in solar radiation. During this period, the magnitude of the ET is generally very large relative to the lateral flow of soil water over the regions with shallow GWT (Emanuel et al. 2014), and hence produces more influence on the seasonal GWT dynamics. However, the different PFTs have different photosynthetic, optical and aerodynamic properties including LAI, which results in variable partitioning of solar radiation absorbed by vegetation and ground and root-water uptake (Sulis et al. 2015; Shrestha et al. 2018a,b). This is also observed as variability in the GWT fluctuations for different PFTs, besides the influence of lateral flow associated with topography.
The seasonal dynamics in GWT depth is also well captured in terms of the near-surface soil moisture anomaly and the terrestrial water storage anomaly in SMOS and GRACE data, respectively, for the entire region. An earlier study by Rahman et al. (2014) also showed the presence of such seasonal dynamics in GWT depth over the Rur catchment in Germany. They also pointed that the ET variability is driven by the radiative forcing and the variability of ET influences the subsurface hydrodynamics and creates the diurnal GWT depth fluctuation. In fact, the variability in the radiative forcing at surface is mainly driven by the clouds and vegetation—hence their distribution together changes the partitioning of GWT dynamics influence by ET and lateral flow—thereby modulating the spatiotemporal pattern of fluctuations in GWT depth.
The Rhine Massif as such appears to play an important role in the zonal segregation of clouds and the corresponding precipitation and incoming solar radiation, including the vegetation distribution in the region. The Süder Uplands and Westerwald east of the Rhine have relatively high cloud cover and precipitation (zone 3) with shallow distribution of GWT depth fluctuations, while parts of Eifel and Hunsrück to the west of the Rhine have relatively low cloud cover and precipitation (zone 1) with relatively high GWT depth fluctuations (distribution shifts rightward). The lowland on the northwestern part of the domain including the Rhine Valley (zone 4) have relatively high cloud cover and low precipitation with GWT depth fluctuations shifting slightly rightward compared to zone 3. This rightward shift in the distribution is also clearly visible for the regions with different PFTs in zones 4 to 1 (see Table 2), indicating a larger influence of ET variability on GWT dynamics over the region.
5. Conclusions
The 10-yr annual average precipitation, incoming solar radiation (accumulated) and shallow GWT depth fluctuation for this temperate region are around 840 ± 94 mm, 1050 ± 54 kWh m−2, and 0.55 ± 0.21 m, respectively. The spatiotemporal distribution of clouds, partly influenced by the Rhine Massif, modulate the seasonal variability of incoming solar radiation and precipitation. The distribution of vegetation also exhibits a strong local control on the seasonal cycle of shallow GWT fluctuations, with higher magnitudes for broadleaf trees and crops compared to needle leaf trees, grasslands and urban areas. Thus, for regions with complex terrains and temperate climate, the intersection between clouds, topography and vegetation together determines the partitioning between the influence of ET and lateral flow on shallow GWT dynamics.
Clouds and precipitation processes constitute two of the major uncertainties in the state of the art weather and climate models. This study shows the importance of better estimates of incoming solar radiation and precipitation to accurately estimate and predict shallow GWT depth variability over the region. In future work, the current modeling study will be extended for additional years, which will provide valuable data on the spatiotemporal evolution of shallow GWT depth fluctuations for this region, which could be used by the scientific community as well as policy makers.
Acknowledgments
The research was carried out in the framework of the priority programme SPP-2115 “Polarimetric Radar Observations meet Atmospheric Modelling (PROM)” in the project ILACPR funded by the German Research Foundation (DFG, Grant SH 1326/1-1). I gratefully acknowledge the computing time (project HBN33) granted by the John von Neumann Institute for Computing (NIC) and provided on the supercomputer JUWELS at Jülich Supercomputing Centre (JSC). Finally, I would also like to thank Veronica Koess for processing the atmospheric forcing data. The data analysis including the pre-processing and post-processing of input data was done using the NCAR Command language (Version 6.4.0).
Data availability statement
The yearly simulations required an average of 20 000 core-hours using 192 compute cores in JUWELS machine at Jülich Supercomputing Center (JSC). The data volume produced was approximately 5.4 GB per year. The COSMO analysis data are available from German Weather Service (DWD). The SMOS monthly aggregated Level 3 data are available at https://www.catds.fr/sipad/ produced by CATDS. The GRACE land data are available at http://grace.jpl.nasa.gov, supported by the NASA MEaSUREs Program.
The data generated for model runs, including model outputs and analysis for the current study is available from Collaborative Research Centre/Transregio 32 database (https://doi.org/10.5880/TR32DB.40). The hydrological model of TSMP is an open source model available on Github (https://github.com/HPSCTerrSys/TSMP).
REFERENCES
Al Bitar, A., and Coauthors, 2017: The global SMOS Level 3 daily soil moisture and brightness temperature maps. Earth Syst. Sci. Data, 9, 293–315, https://doi.org/10.5194/essd-9-293-2017.
Ashby, S. F., and R. D. Falgout, 1996: A parallel multigrid preconditioned conjugate gradient algorithm for groundwater flow simulations. Nucl. Sci. Eng., 124, 145–159, https://doi.org/10.13182/NSE96-A24230.
Baldauf, M., A. Seifert, J. Förstner, D. Majewski, M. Raschendorfer, and T. Reinhardt, 2011: Operational convective-scale numerical weather prediction with the COSMO model: Description and sensitivities. Mon. Wea. Rev., 139, 3887–3905, https://doi.org/10.1175/MWR-D-10-05013.1.
Böhme, T., and Coauthors, 2011: Long-term evaluation of COSMO forecasting using combined observational data of the GOP period. Meteor. Z., 20, 119–132, https://doi.org/10.1127/0941-2948/2011/0225.
CATDS, 2016: CATDS-PDC L3SM Aggregated - 3-day, 10-day and monthly global map of soil moisture values from SMOS satellite. CATDS (CNES, IFREMER, CESBIO), accessed 10 June 2020, https://doi.org/10.12770/b57e0d3d-e6e4-4615-b2ba-6feb7166e0e6.
Condon, L. E., A. L. Atchley, and R. M. Maxwell, 2020: Evapotranspiration depletes groundwater under warming over the contiguous United States. Nat. Commun., 11, 873, https://doi.org/10.1038/s41467-020-14688-0.
Craig, A., S. Valcke, and L. Coquart, 2017: Development and performance of a new version of the OASIS coupler, OASIS3-MCT_3. 0. Geosci. Model Dev., 10, 3297–3308, https://doi.org/10.5194/gmd-10-3297-2017.
Emanuel, R. E., A. G. Hazen, B. L. McGlynn, and K. G. Jencso, 2014: Vegetation and topographic influences on the connectivity of shallow groundwater between hillslopes and streams. Ecohydrology, 7, 887–895, https://doi.org/10.1002/eco.1409.
Fan, Y., H. Li, and G. Miguez-Macho, 2013: Global patterns of groundwater table depth. Science, 339, 940–943, https://doi.org/10.1126/science.1229881.
Farr, T. G., and Coauthors, 2007: The Shuttle Radar Topography Mission. Rev. Geophys., 45, RG2004, https://doi.org/10.1029/2005RG000183.
Fatichi, S., and C. Pappas, 2017: Constrained variability of modeled T: ET ratio across biomes. Geophys. Res. Lett., 44, 6795–6803, https://doi.org/10.1002/2017GL074041.
Friedl, M., and D. Sulla-Menashe, 2019: MCD12Q1 MODIS/Terra+Aqua land cover type yearly L3 Global 500m SIN Grid V006. NASA EOSDIS Land Processes DAAC, accessed 11 February 2020, https://doi.org/10.5067/MODIS/MCD12Q1.006.
Gasper, F., K. Goergen, P. Shrestha, M. Sulis, J. Rihani, M. Geimer, and S. J. Kollet, 2014: Implementation and scaling of the fully coupled Terrestrial Systems Modeling Platform (TerrSysMP v1.0) in a massively parallel supercomputing environment – A case study on JUQUEEN (IBM Blue Gene/Q). Geosci. Model Dev., 7, 2531–2543, https://doi.org/10.5194/gmd-7-2531-2014.
Good, S. P., D. Noone, and G. Bowen, 2015: Hydrologic connectivity constrains partitioning of global terrestrial water fluxes. Science, 349, 175–177, https://doi.org/10.1126/science.aaa5931.
Gribovszki, Z., J. Szilágyi, and P. Kalicz, 2010: Diurnal fluctuations in shallow groundwater levels and streamflow rates and their interpretation–A review. J. Hydrol., 385, 371–383, https://doi.org/10.1016/j.jhydrol.2010.02.001.
Jacquette, E., A. Al Bitar, A. Mialon, Y. H. Kerr, A. Quesney, F. Cabot, and P. Richaume, 2010: SMOS CATDS level 3 global products over land. Proc. SPIE, 7824, https://doi.org/10.1117/12.865093.
Jones, J. E., and C. S. Woodward, 2001: Newton-Krylov-multigrid solvers for large-scale, highly heterogeneous, variably saturated flow problems. Adv. Water Resour., 24, 763–774, https://doi.org/10.1016/S0309-1708(00)00075-0.
Kerr, Y. H., P. Waldteufel, J.-P. Wigneron, J. Martinuzzi, J. Font, and M. Berger, 2001: Soil moisture retrieval from space: The Soil Moisture and Ocean Salinity (SMOS) mission. IEEE Trans. Geosci. Remote Sens., 39, 1729–1735, https://doi.org/10.1109/36.942551.
Kerr, Y. H., E. Jacquette, A. Al Bitar, F. Cabot, A. Mialon, P. Richaume, A. Quesney, and L. Berthon, 2013: CATDS SMOS L3 soil moisture retrieval processor Algorithm Theoretical Baseline Doc. SO-TN-CBSA-GS-0029, 73 pp., http://www.cesbio.ups-tlse.fr/SMOS_blog/wp-content/uploads/2013/08/ATBD_L3_rev2_draft.pdf.
Kerr, Y. H., and Coauthors, 2016: Overview of SMOS performance in terms of global soil moisture monitoring after six years in operation. Remote Sens. Environ., 180, 40–63, https://doi.org/10.1016/j.rse.2016.02.042.
Keune, J., F. Gasper, K. Goergen, A. Hense, P. Shrestha, M. Sulis, and S. Kollet, 2016: Studying the influence of groundwater representations on land surface-atmosphere feedbacks during the European heat wave in 2003. J. Geophys. Res., 121, 13 301–13 325, https://doi.org/10.1002/2016JD025426.
Kollet, S., and R. M. Maxwell, 2008: Capturing the influence of groundwater dynamics on land surface processes using an integrated, distributed watershed model. Water Resour. Res., 44, W02402, https://doi.org/10.1029/2007WR006004.
Kollet, S. J., and R. M. Maxwell, 2006: Integrated surface–groundwater flow modeling: A free-surface overland flow boundary condition in a parallel groundwater flow model. Adv. Water Resour., 29, 945–958, https://doi.org/10.1016/j.advwatres.2005.08.006.
Landerer, F. W., and S. C. Swenson, 2012: Accuracy of scaled GRACE terrestrial water storage estimates. Water Resour. Res., 48, W04531, https://doi.org/10.1029/2011WR011453.
Leung, L. R., M. Huang, Y. Qian, and X. Liang, 2011: Climate–soil–vegetation control on groundwater table dynamics and its feedbacks in a climate model. Climate Dyn., 36, 57–81, https://doi.org/10.1007/s00382-010-0746-x.
Lindau, R., and C. Simmer, 2013: On correcting precipitation as simulated by the regional climate model COSMO-CLM with daily rain gauge observations. Meteor. Atmos. Phys., 119, 31–42, https://doi.org/10.1007/s00703-012-0215-7.
Manoli, G., and Coauthors, 2019: Magnitude of urban heat islands largely explained by climate and population. Nature, 573, 55–60, https://doi.org/10.1038/s41586-019-1512-9.
Maxwell, R. M., 2013: A terrain-following grid transform and preconditioner for parallel, large-scale, integrated hydrologic modeling. Adv. Water Resour., 53, 109–117, https://doi.org/10.1016/j.advwatres.2012.10.001.
Maxwell, R. M., and S. J. Kollet, 2008: Interdependence of groundwater dynamics and land-energy feedbacks under climate change. Nat. Geosci., 1, 665–669, https://doi.org/10.1038/ngeo315.
Maxwell, R. M., and L. E. Condon, 2016: Connections between groundwater flow and transpiration partitioning. Science, 353, 377–380, https://doi.org/10.1126/science.aaf7891.
Myneni, R., Y. Knyazikhin, and T. Park, 2015: MCD15A2H MODIS/Terra+Aqua Leaf Area Index/FPAR 8-day L4 Global 500m SIN Grid V006. NASA EOSDIS Land Processes DAAC, accessed 11 February 2020, https://doi.org/10.5067/MODIS/MCD15A2H.006.
Oleson, K. W., and Coauthors, 2008: Improvements to the Community Land Model and their impact on the hydrological cycle. J. Geophys. Res. Biogeosci., 113, G01021, https://doi.org/10.1029/2007JG000563.
Rahman, M., M. Sulis, and S. J. Kollet, 2014: The concept of dual-boundary forcing in land surface-subsurface interactions of the terrestrial hydrologic and energy cycles. Water Resour. Res., 50, 8531–8548, https://doi.org/10.1002/2014WR015738.
Reichle, R. H., and R. D. Koster, 2004: Bias reduction in short records of satellite soil moisture. Geophys. Res. Lett., 31, L19501, https://doi.org/10.1029/2004GL020938.
Sakumura, C., S. Bettadpur, and S. Bruinsma, 2014: Ensemble prediction and intercomparison analysis of GRACE time-variable gravity field models. Geophys. Res. Lett., 41, 1389–1397, https://doi.org/10.1002/2013GL058632.
Schaap, M. G., and F. J. Leij, 1998: Database-related accuracy and uncertainty of pedotransfer functions. Soil Sci., 163, 765–779, https://doi.org/10.1097/00010694-199810000-00001.
Schlesinger, W. H., and S. Jasechko, 2014: Transpiration in the global water cycle. Agric. For. Meteor., 189–190, 115–117, https://doi.org/10.1016/j.agrformet.2014.01.011.
Shrestha, P., 2019: TerrSysMP pre-processing and post-processing system. CRC/TR32 Database (TR32DB), accessed 10 April 2019 https://doi.org/10.5880/TR32DB.37.
Shrestha, P., M. Sulis, M. Masbou, S. Kollet, and C. Simmer, 2014: A scale-consistent terrestrial system modeling platform based on COSMO, CLM and ParFlow. Mon. Wea. Rev., 142, 3466–3483, https://doi.org/10.1175/MWR-D-14-00029.1.
Shrestha, P., M. Sulis, C. Simmer, and S. Kollet, 2015: Impacts of grid resolution on surface energy fluxes simulated with an integrated surface-groundwater flow model. Hydrol. Earth Syst. Sci. Discuss., 12, 6437–6466, https://doi.org/10.5194/hessd-12-6437-2015.
Shrestha, P., M. Sulis, C. Simmer, and S. Kollet, 2018a: Effects of horizontal grid resolution on evapotranspiration partitioning using TerrSysMP. J. Hydrol., 557, 910–915, https://doi.org/10.1016/j.jhydrol.2018.01.024.
Shrestha, P., W. Kurtz, G. Vogel, J. P. Schulz, M. Sulis, H. J. Hendricks Franssen, S. Kollet, and C. Simmer, 2018b: Connection between root zone soil moisture and surface energy flux partitioning using modeling, observations, and data assimilation for a temperate grassland site in Germany. J. Geophys. Res. Biogeosci., 123, 2839–2862, https://doi.org/10.1029/2016JG003753.
Siebert, S., and P. Döll, 2010: Quantifying blue and green virtual water contents in global crop production as well as potential production losses without irrigation. J. Hydrol., 384, 198–217, https://doi.org/10.1016/j.jhydrol.2009.07.031.
Siebert, S., J. Burke, J. M. Faures, K. Frenken, J. Hoogeveen, P. Döll, and F. T. Portmann, 2010: Groundwater use for irrigation—A global inventory. Hydrol. Earth Syst. Sci., 14, 1863–1880, https://doi.org/10.5194/hess-14-1863-2010.
Simmer, C., and Coauthors, 2015: Monitoring and modeling the terrestrial system from pores to catchments: The transregional collaborative research center on patterns in the soil–vegetation–atmosphere system. Bull. Amer. Meteor. Soc., 96, 1765–1787, https://doi.org/10.1175/BAMS-D-13-00134.1.
Sulis, M., M. Langensiepen, P. Shrestha, A. Schickling, C. Simmer, and S. J. Kollet, 2015: Evaluating the influence of plant-specific physiological parameterizations on the partitioning of land surface energy fluxes. J. Hydrometeor., 16, 517–533, https://doi.org/10.1175/JHM-D-14-0153.1.
Sulis, M., J. Keune, P. Shrestha, C. Simmer, and S. J. Kollet, 2018: Quantifying the impact of subsurface-land surface physical processes on the predictive skill of subseasonal mesoscale atmospheric simulations. J. Geophys. Res. Atmos., 123, 9131–9151, https://doi.org/10.1029/2017JD028187.
Swenson, S. C., 2012: GRACE monthly land water mass grids NETCDF RELEASE 5.0, version 5.0. PO.DAAC, accessed 10 May 2020, https://doi.org/10.5067/TELND-NC005.
Swenson, S. C., and J. Wahr, 2006: Post-processing removal of correlated errors in GRACE data. Geophys. Res. Lett., 33, L08402, https://doi.org/10.1029/2005GL025285.
Wada, Y., L. Beek, and M. F. P. Bierkens, 2012: Nonsustainable groundwater sustaining irrigation: A global assessment. Water Resour. Res., 48, W00L06, https://doi.org/10.1029/2011WR010562.
Wada, Y., I. E. M. de Graaf, and L. P. H. van Beek, 2016: High-resolution modeling of human and climate impacts on global water resources. J. Adv. Model. Earth Syst., 8, 735–763, https://doi.org/10.1002/2015MS000618.
Wang, F., A. Ducharne, F. Cheruy, M. H. Lo, and J. Y. Grandpeix, 2018: Impact of a shallow groundwater table on the global water cycle in the IPSL land–atmosphere coupled model. Climate Dyn., 50, 3505–3522, https://doi.org/10.1007/s00382-017-3820-9.
