An Introduction to an Algorithm for Extracting Precipitable Water Vapor Over Land From AMSR2 Images

Accurate estimation of atmospheric precipitable water vapor (PWV) with high spatial and temporal resolutions is challenging. The main objective of this study is to develop an algorithm of retrieving PWV from advanced microwave scanning radiometer 2 (AMSR2) images over the land using local data. In this regard, the western part of Iran was selected as the study area and one-year period including 80 days of four seasons since 2015 to 2016 was utilized as training and testing days. Brightness temperatures from 18.7-, 23.8-, 36.5-, and 89-GHz channels of AMSR2 in horizontal and vertical polarizations were used for implementations. Moreover, reflectance of bands 1 and 2 of moderate resolution imaging spectroradiometer was used to investigate vegetation cover conditions. Six radiosonde stations of Tehran, Tabriz, Kermanshah, Ahvaz, Shiraz, and Isfahan in the study area were used as ground truth data. Then, seven experiments were conducted, which two first implementations were done using proposed algorithms for AMSR-E and AMSR-2 sensors over the land. In the five other experiments, the influences of cloud liquid water, land surface temperature, and ground elevation were considered in the measured brightness temperatures. In the fourth to seventh experiments, the normalized-difference vegetation index was imported as a linear, quadratic, power, and exponential parameter in the equation of PWV extraction, respectively. Based on the validation results, the best correlation was observed in the sixth experiment so that the coefficient of determination (R
²
), RMSE, and the ratio RMSE/STD were 0.648, 3.891 mm, and 0.6, respectively.