Submitted to: International Symposium on Recent Advances in Quantitative Remote Sensing
Publication Type: Abstract Only
Publication Acceptance Date: 6/1/2014
Publication Date: 9/22/2014
Citation: Alfieri, J.G., Kustas, W.P., Prueger, J.H., Neale, C. 2014. Using remote sensing to characterize the influence of field-scale heterogeneity on in-site measurements of evapotranspiration. Fourth International Symposium on Recent Advances in Quantitative Remote Sensing. Paper No. P3.02.
Technical Abstract: Accurate estimates of field-scale evapotranspiration (ET) are critical to maximizing the efficient use of water for agricultural production which, particularly in arid and semi-arid environments, is the largest consumptive user of fresh water. Often, these estimates are derived using numerical or remote sensing-based models that are developed, calibrated, and validated against in-situ data collected using techniques such as eddy covariance (EC). Consequently, the accuracy of the modeled estimates of ET are strongly linked to both the accuracy and representativeness of the in-situ measurements. Using data collected over adjacent irrigated cotton fields in the Texas High Plains as a part of the 2008 Bushland Evapotranspiration and Agricultural Remote Sensing Experiment (BEAREX08), the impact of within-field variability in vegetation density on in-situ measurements of ET was analyzed. Each field was instrumented with a pair of EC systems that were cross-calibrated immediately prior to the field study and a large weighing lysimeter (LY) located in the center of the field. Vegetation density was measured using both in-situ techniques and remote sensing methods with airborne imagery. Even after accounting for factors such as local advective effects and the imperfect energy balance closure associated the EC measurements, differences in the cumulative ET over the course of the growing season were between 10 and 20%. At the same time, analyses of the high-resolution leaf area index (LAI) maps derived from the airborne imagery showed significant variability in the vegetation density; coefficients of variation ranged as high as 40%. By comparing the LAI within the source area, i.e. footprint, of the different measurement systems, this study demonstrated the influence of fine-scale spatial variability in vegetation cover on measurements of ET which can only be quantified through high resolution remote sensing imagery. This study also underscores the need to account for the effects of surface heterogeneity on in-situ measurements used for ET model calibration and validation.