High-resolution modeling of the spatial heterogeneity of soil moisture: Applications in network design

Abstract

The spatial heterogeneity of soil moisture remains a persistent challenge in the design of in situ measurement networks, spatial downscaling of coarse estimates (e.g., satellite retrievals), and hydrologic modeling. To address this challenge, we analyze high-resolution (9 m) simulated soil moisture fields over the Little River Experimental Watershed (LREW) in Georgia, USA, to assess the role and interaction of the spatial heterogeneity controls of soil moisture. We calibrate and validate the TOPLATS distributed hydrologic model with high to moderate resolution land and meteorological data sets to provide daily soil moisture fields between 2004 and 2008. The results suggest that topography and soils are the main drivers of spatial heterogeneity over the LREW. We use this analysis to introduce a novel network design method that uses land data sets as proxies of the main drivers of local heterogeneity (topography, land cover, and soil properties) to define unique and representative hydrologic similar units (subsurface, surface, and vegetation) for probe placement. The calibration of the hydrologic model and network design method illustrates how the use of hydrologic similar units in hydrologic modeling could minimize computation and guide efforts toward improved macroscale land surface modeling.

Soil Moisture for Hydrological Applications: Open Questions and New Opportunities

“…This important point is evident if we compare the capability of hydrological models in reproducing soil moisture spatial variability with respect to temporal variability. Frequently, the magnitude of variability is severely underestimated [90,93,94], and at regional scale soil moisture spatial patterns from modelling and satellite observations are found to be not consistent [95]. In this context, some recent modelling studies are demonstrating the value of soil moisture data for a more in-depth validation of distributed hydrological models [96,97].…”

“…In time, soil moisture is mainly driven by precipitation and evapotranspiration, and its temporal variability is also a function of soil characteristics, vegetation, topography and groundwater [87,89]. Several modelling studies obtained very good performances in simulating point-(grid-) scale soil moisture temporal evolution [65,67,90], thus showing that a good knowledge has been gained about the factors influencing point scale soil moisture temporal variability. Figure 3 shows a comparison between observed and modelled data at two sites in central Italy using the Soil Water Balance Model [67].…”

Improving the representation of hydrologic processes in Earth System Models

“…For instance, while insights can be gained from detailed studies over specific river basins, a substantial amount of additional work is needed to extend such insights to all river basins (e.g., from tropical to polar river basins). To this end, effort is required to estimate important geophysical attributes on the global scale (e.g., soil information, bedrock depth, hydraulic geometry) [Tesfa et al, 2009;Chaney et al, 2014;Fan et al, 2015;Allen and Pavelsky, 2015], and use the available geophysical information to estimate spatially variable model parameters [Samaniego et al, 2010]. The broad applicability of modeling advances also requires a synthesis across catchment-scale observatories, to pull out processes and interactions that are common in a wide range of climatic-hydrologic-ecosystem regimes [McDonnell et al, 2007;Wagener et al, 2007].…”

Robust Initial Wetness Condition Framework of an Event-Based Rainfall–Runoff Model Using Remotely Sensed Soil Moisture

“…The amount of soil moisture is influenced by differences in the infiltration capacity according to the surface soil properties and land cover [43,44]. Chaney et al [45] noted that the topography, land cover, and soil properties are the main drivers of spatial heterogeneity of soil moisture, and soil moisture variability is attributed to the complex interactions between the drivers of heterogeneity. The results of this study support the findings that both the remotely sensed soil moisture (SSM and SWI) have the potential to mimic soil moisture parameters and can be characterized as the IWCs before rainfall events, as noted in several previous studies [10,19,46].…”