Climate–soil–vegetation control on groundwater table dynamics and its feedbacks in a climate model

Abstract

Among the three dynamically linked branches of the water cycle, including atmospheric, surface, and subsurface water, groundwater is the largest reservoir and an active component of the hydrologic system. Because of the inherent slow response time, groundwater may be particularly relevant for long time-scale processes such as multi-years or decadal droughts. This study uses regional climate simulations with and without surface watergroundwater interactions for the conterminous US to assess the influence of climate, soil, and vegetation on groundwater table dynamics, and its potential feedbacks to regional climate. Analyses show that precipitation has a dominant influence on the spatial and temporal variations of groundwater table depth (GWT). The simulated GWT is found to decrease sharply with increasing precipitation. Our simulation also shows some distinct spatial variations that are related to soil porosity and hydraulic conductivity. Vegetation properties such as minimum stomatal resistance, and root depth and fraction are also found to play an important role in controlling the groundwater table.Comparing two simulations with and without groundwater table dynamics, we find that groundwater table dynamics mainly influences the partitioning of soil water between the surface (0-0.5 m) and subsurface (0.5-5 m) rather than total soil moisture. In most areas, groundwater table dynamics increases surface soil moisture at the expense of the subsurface, except in regions with very shallow groundwater table. The change in soil water partitioning between the surface and subsurface is found to strongly correlate with the partitioning of surface sensible and latent heat fluxes. The evaporative fraction (EF) is generally higher during summer when groundwater table dynamics is included. This is accompanied by increased cloudiness, reduced diurnal temperature range, cooler surface temperature, and increased cloud top height. Although both convective and non-convective precipitation are enhanced, the higher EF changes the partitioning to favor more non-convective precipitation, but this result could be sensitive to the convective parameterization used. Compared to simulations without groundwater table dynamics, the dry bias in the summer precipitation is slightly reduced over the central and eastern US Groundwater table dynamics can provide important feedbacks to atmospheric processes, and these feedbacks are stronger in regions with deeper groundwater table, because the interactions between surface and subsurface are weak when the groundwater table is deep. This increases the sensitivity of surface soil moisture to precipitation anomalies, and therefore enhances land surface feedbacks to the atmosphere through changes in soil moisture and evaporative fraction. By altering the groundwater table depth, land use change and groundwater withdrawal can alter land surface response and feedback to the climate system.

Implementing and Evaluating Variable Soil Thickness in the Community Land Model, Version 4.5 (CLM4.5)

“…Recent research has shown that a model atmosphere would respond to groundwater dynamics (Anyah et al 2008;Jiang et al 2009;Leung et al 2011). Using a different forcing, CRUNCEP in VARCRU, also has an effect on the variable soil thickness simulation as seen by the drier uppermost soil moisture in the southwestern United States and the northern Great Plains in Fig.…”

“…This may not be the case if coupled to an atmospheric model because of potential feedbacks between LH flux and precipitation. Several studies (Anyah et al 2008;Jiang et al 2009;Leung et al 2011) have shown changes to LH flux in response to changes in precipitation caused by the inclusion of groundwater models. At Karakoram and southeastern Africa, LH flux in DEEP is almost the same as in CLM4.5, whereas LH flux in VAR is as much as about 12 W m 22 lower than in DEEP and CLM4.5 (Figs.…”

Improving the representation of hydrologic processes in Earth System Models

“…Similar approaches are used in the Total Runoff Integrating Pathways (TRIP) model Decharme et al, 2010;Decharme et al, 2012]. However, VIC-GW [Liang et al, 2003;Leung et al, 2011] allows for two-way soil-water table interactions in the vertical direction. 3.…”

Coupling a three-dimensional subsurface flow and transport model with a land surface model to simulate stream–aquifer–land interactions (CP v1.0)

“…al., 2015), cloud formation (Rahman et al, 2015), and climate (Leung et al, 2011;Taylor et al, 2013). Lateral subsurface processes are fundamentally important on multiple spatial scales, including hill-slope scales (McNamara et al, 2005;Zhang et al, 2011), basin scales in semiarid and arid climates where regional aquifers sustain baseflows in rivers (Schaller and Fan, 2009) and wetlands (Fan and MiguezMacho, 2011).…”