ARS | Publication request: Reflectance-Based Estimation of Soil Heat Fluxes in the Texas High Plains

Submitted to: American Geophysical Union
Publication Type: Abstract Only
Publication Acceptance Date: October 15, 2010
Publication Date: December 16, 2010
Citation: Gowda, P., Colaizzi, P.D., Oshaughnessy, S.A., Ha, W., Howell, T.A. 2010. Reflectance-Based Estimation of Soil Heat Fluxes in the Texas High Plains[abstract]. American Geophysical Union Meeting, December 13-17, 2010, San Francisco, California. Paper No. H31B-0984.

Technical Abstract: Soil heat flux (G) is one of the terms required for estimating evapotranspiration (ET) rates using an energy balance. Numerous reflectance-based models are available in the literature for estimating G fluxes. However, these models have shown wide variation in their performance. Therefore, operational ET remote sensing programs may require locally developed/calibrated models for accurately estimating G. The objective of this study was to develop and evaluate reflectance-based empirical G models for the semi-arid Texas High Plains. Soil heat flux was measured at 0.15 hz interval and averaged every 15 minutes at five different locations within a 4.7 ha lysimeter field with Pullman clay loam soil during the 2010 summer growing season. The field was planted to soybean and managed under dryland conditions. In each location, G was measured at 8 cm depth with two Campbell Scientific HFT3 soil heat flux plates. Soil temperature was measured at 2 and 6 cm above the soil heat flux plates. Soil moisture was measured in the 2-8 cm layer using Acclima SDI-12 sensors. Hourly G fluxes at the surface were calculated by adding the measured G fluxes at 8 cm to the energy stored above the heat flux plates. A multispectral radiometer, (MSR, CROPSCAN, Inc.) and hand-held thermometer (EVEREST Interscience Inc.) measured surface reflectance in red and near infrared bandwidths and surface temperature (ST), respectively, daily at 11:30 AM CST to be consistent with the Landsat 5 overpass time. Fraction crop cover (FC) was measured by digital photographs taken twice a week. A set of G models was developed for estimating hourly fluxes based on measured reflectance, net radiation, Normalized Difference Vegetation Index (NVDI), ST, and FC. Resulting models were compared for performance with existing models available in the literature. In this presentation, we will discuss our G models for the Texas High Plains and the statistical results.