Spatial Process of Soil Hydrological State Variables in Pinyon-Juniper Woodland.
Ole Wendroth1, Inmaculada Lebron2, M. Madsen3, David Robinson4, J. Belnap5, and David Chandler3. (1) University of Kentucky, "N-122M Ag Sci N., Dep Plant & Soil", Lexington, KY 40546-0312, United States of America, (2) Stanford University, 397 Panama Mall, 397 Panama Mall, Stanford, CA 94305, United States of America, (3) Utah State University, Department of Plants, Soils and Biometeorology, Logan, UT 84322-4820, (4) Stanford Univ Dept Geophysics, 397 Panama Mall, Stanford, CA 94305-2215, United States of America, (5) USGS, Canyonlands Field Station, Moab, UT 84532
Soil water and crop status across the landscape are closely associated in space and time. Hydrologic functional subunits vary within landscapes, and even within existing soil classification units, differing functional subunits can be identified. It is very important to understand the spatial process of soil water transport properties of surface soils within and across different hydrological functional subunits. The purpose of this study was to identify whether spatial water content series showed stable patterns when sampled during four different seasons. Moreover, the relation of matric-saturated hydraulic conductivity K to underlying soil moisture status should be identified. It was especially interesting if K was closely related to soil water content during the sampling time or rather related to the spatial soil water content distribution at another characteristic time. During winter, spring, summer, and fall of the year 2005 land surface soil water content was measured every 10 cm along a 15-m- transect, which reached from a trunk of a juniper tree through a patch of biological soil crust, a disturbed zone affected by wheel traffic, and another biological soil crust under ephreda bushes and a pinion pine. Along the same transect and at the same four sampling dates of soil water content, K at a soil water pressure head of -2 cm was measured every 30 cm in the horizontal direction using a tension infiltrometer. Spatial processes of both soil water content and K were analysed with respect to their spatial covariance structure. Partial autocorrelation functions indicated that a choice of first order autoregressive model would be appropriate to describe the spatial process. State-space models of K and soil water content were developed and autoregression coefficients estimated for different sampling scenarios. Pronounced fluctuations of K could only be conserved, if a sufficient number of observations was incorporated in the model.