关键词: Landsat8, 垂直干旱指数, 土壤含水量, 植被含水量, 黄土高原

Abstract:

Mixed pixel is not only a common phenomenon in remote sensing but also an important factor influencing accuracy of surface parameters retrieval. To understand the influence of mixed pixel on monitoring ability of remote sensing drought index, using the Perpendicular Drought Index (PDI) as an example, the ground-satellite experiment data were used to verify the applicability of PDI on surface moisture monitoring under different coverage conditions in Longdong loess plateau, and a new surface water content (named SMsur) considering mixed pixel was proposed, and on this basis, the remote sensing monitoring model for surface water content  was established. The results are as follows: (1) Without considering mixed pixel, the maximum correlation coefficients between PDI and soil water content were 0.5 and 0.05 for bare soil and vegetation conditions, respectively. (2) Considering mixed pixel, the maximum correlation coefficient between PDI and SMsur was 0.6, which indicated that it could describe the applicability and monitoring effect more accurately about remote sensing indexes considering mixed pixel in monitoring ability evaluation in study area. (3)The surface water content estimated by PDI model considering mixed pixel could reflect the status of surface moisture under different vegetation coverage underlying surface, and their values were close to the observed data, the correlation coefficient of them was 0.85, the mean absolute percent error was less than 8.0%, and the minimum root mean square error was 2.17%. The regional distribution of the estimated SMsur was consistent with the investigation result in the study area, and the moisture of the farmland was largest with SMsur more than 30%.

Key words:  Landsat8, Perpendicular Drought Index, soil water content, vegetation water content, loess plateau