Use of high-resolution thermal infrared remote sensing and "three-temperature model" for transpiration monitoring in arid inland river catchment

Abstract

Based on high-resolution thermal infrared remote sensing and the three-temperature model (31 model), we developed a new algorithm for mapping transpiration. The necessary input parameters were surface temperature, air temperature, and solar radiation only. Therefore, in comparison with conventional methods, it is a simple and potentially valuable way to employ the thermal infrared remote sensing application. By using the proposed method, transpiration of sixteen types of typical vegetation in the upper and middle reaches of the Heihe River Basin in Northwestern China were calculated pixel by pixel. We evaluated modeled evapotranspiration with an eddy covariance (EC) result from the established regression equation, and a scatter correlation plot for the measured and estimated transpiration indicated that the model estimate is within acceptable limits, with a correlation coefficient of R-2 = 0.796. Compared with the desert-oasis transitional zone, the maximum transpiration rate at the Gobi Desert presented a little earlier but was smaller. This great difference may imply that different types of plants have different water-use abilities and drought tolerances. Thus, the transpiration estimation with the 31 model, using high-resolution thermal infrared remote sensing data, can provide not only a bridge between large-scale and point observation with a measure of m(2) from the infrared thermal imager, but also provide decision support for operational water management issues. (C) 2014 Elsevier B.V. All rights reserved.