Soil Water Content Spatial Correlation Estimation Using GPR

Spatial correlation estimates of water content in the shallow subsurface are needed as input for stochastic generation of representative vadose zone models. Improved vadose zone characterization is also important for applications such as precision agriculture, environmental monitoring, and optimizing data collection strategies. However, water content in the vadose zone is often highly variable. As such, using only traditional vadose zone sampling techniques, it is difficult to collect accurate point estimates of water content, and it is even more challenging to obtain accurate estimates of spatial correlation. We are investigating the use of Ground Penetrating Radar (GPR) to provide accurate, very high resolution estimates of water content and its spatial correlation in the very shallow subsurface. Using surface GPR groundwave techniques, estimates of shallow water content (~10 to 20 cm below ground surface) can be obtained quickly; these high resolution measurements can then be used to estimate the spatial structure and correlation length of the water content. This concept is being investigated in a five acre field located within the Robert Mondavi vineyards in Napa County, California. At several times during the year, we collected a grid of GPR measurements along selected rows using 10 cm sampling intervals. Using petrophysical relationships, these measurements were converted to water content and then analyzed for spatial and temporal correlation. To calibrate and verify the GPR measurements, we are comparing the GPR-obtained information with data collected at the same site using multispectral remote sensing, time domain reflectometry, electrical resistivity, neutron probe, and soil texture analysis techniques. This presentation will focus on water content spatial correlation estimation and its variation with season using the different measurement techniques. Based on this analysis, we will discuss the spatial support scales associated with the different types of vadose zone measurement techniques, the benefits of incorporating GPR into the spatial correlation estimation procedure, and the practical implementation of these methods for spatial correlation estimation.