/AnMtgsAbsts2009.54186 Evaluating Spatial Relationships Between Soil Properties and Terrain Attributes in An Oak Woodland Catchment.

Monday, November 2, 2009
Convention Center, Exhibit Hall BC, Second Floor

Alexandre Swarowsky, Dylan Beaudette, Anthony O'Geen and Randy Dahlgren, One Shields Ave., Univ. of California, Davis, Davis, CA
Abstract:

Topographic attributes have been successfully used in many different settings to describe patterns in the landscape-scale variation of soil properties. While the complex soil landscapes of California’s Sierra Foothill Region have been mapped by the NRCS, the map units do not adequately describe soil variation at the watershed-scale. A lack of detailed soils information in this region impedes accurate modeling of hydrologic behavior and water-soil interaction; critical for long-term surface water supply planning. Therefore, our objective is to evaluate the use of terrain attributes to predict soil properties at the catchment scale. A 35-ha watershed in the California Foothill Region was intensively investigated in order to understand the relationships between soil properties and topographic attributes. Soil morphologic data, site descriptions and total soil depth were recorded at each sampling location (n profiles = 100). Samples collected by genetic horizon were characterized using standard methods for physical and chemical properties, as well as elemental analysis (n horizons = 450). A 1-meter resolution digital elevation model (DEM) was constructed from 1000 real-time kinematic (RTK) GPS observations . GRASS GIS was used to generate several terrain shape and intensity parameters. Potential variation in microclimate was modeled with the European Solar Radiation Atlas (ESRA) solar radiation model, and tree canopy extracted from classification of color aerial imagery. While correlation within soil properties and within terrain properties was moderate to high (r = 0.6-0.9), correlation between soil properties and terrain properties was low (r = 0.1-0.2). Variation in soil properties was approximately 10 to 30 times greater with depth than in space. Segmentation of the landscape into hybrid geomorphic-microclimate groups and modeling of profile-summed values are expected to provide better predictive tools in this region.