Soil Moisture Spatial-Temporal Patterns and Their Controls In Two Contrasting Landscapes.

Modeling and prediction of soil hydrologic processes require the identification of dominant controls on soil moisture spatial-temporal patterns and specific hydrological responses under different conditions. We investigated the space-time organization (and/or heterogeneity) of soil moisture and the associated physical controls in two contrasting landscapes in central Pennsylvania: steep-sloped forested Shale Hills Catchment vs. gently-rolling cropped Kepler Farm. Based on both spatially extensive manual and temporally extensive automated soil moisture datasets, sets of analysis of geostatistics, time stability, and Empirical Orthogonal Function (EOF) as well as frequency of preferential flow were applied. For the forested Shale Hills Catchment, the analysis resulted in significant spatial structure (primary EOF), which explained 76% to 89% of the spatial variability associated with soil properties and topography. In contrast, in the Kepler Farm the explained variance by the primary EOF is weak, only 31-66% of the total variability. For the both investigated landscapes, the statistical weight of the spatial EOFs varied by season, soil depth, and the degree of soil-topography variability at a given spatial scale. During the growing season, vegetation and soils exerted significant influences on soil moisture, especially in the surface and near surface; whereas during the non-growing season, terrain attributes showed more significant impacts on soil moisture variation, particularly in deeper soils. In areas of steeper slope, topography dominated over soil properties in controlling soil moisture variation, while in relatively flat areas, soil properties dominated over topography. Subsurface preferential flow was significant in both landscapes, but was particularly prominent in the forested steeper slope landscape with better drained soils. Initial soil moisture played different roles in regulating the preferential flow occurrence at different slope positions, with higher moisture in valley and lower moisture in hilltop favoring to preferential flow occurrence. Overall, the controls on soil moisture spatial organization at the near-surface (<0.3-m) fluctuate seasonally between evapotranspiration and topography; that at intermediate depths (0.3- to 0.7-m) the soil moisture organization is controlled significantly by lateral subsurface flow; and that the organization at deeper depths (> 0.7-m) becomes more temporally persistent and is primarily a function of both topography and soil depth. Based on EOF analysis, an integrate way was suggested to combine the manual datasets with the automatic datasets and the results showed that the fitness of the two data sets was both site-specific and season-specific. This study demonstrated the intertwined spatial complexity and temporal dynamics in soil moisture as controlled by climate, terrain, soil, and vegetation across the contrasting landscapes, the degree of which is a function of spatial scale, soil depth, and season.