Retrieval of cloud liquid water from satellite passive microwave sensor AMSR-E

A new method to retrieve cloud liquid water (CLW) over land using AMSR-E brightness temperatures (T B) was presented in this paper. A surface emission model, Advanced Integral Equation Model (AIEM) and an one-dimensional atmosphere transfer model (1DRTM) were combined to generate a database. Through analysis of the simulated dataset, it is found that the ratio of the polarization difference obtained from 36.5 and 89 GHz (ΔT B(36.5)/ΔT B(89), called PDR_CLW later) is sensitive to CLW. Firstly, the 36.5 GHz vertically polarized measurements were used to derive the surface temperature (T s) with a precision high enough for the algorithm. Secondly, the ratio of the polarization difference between 18.7 and 23.8 GHz (ΔT B(18.7)/ΔT B(23.8), called PDR_WV later) was used to assess the effect of water vapor. Lastly, an index called HI (Height Index), which assumes the form of (T s-T v(89 GHz))/(T s+T v(89 GHz)), is developed to minimize the influence of cloud height. The relationship between PDR_CLW and CLW were regressed and parameterized in terms of PDR_CLW, CLW, PDR_WV, HI and. The algorithm was validated using AMSR-E observations over the Southern Great Plains. The CLW data retrieved by five ground based microwave radiometers (MWR) were used to validate the algorithm, with RMS error of 0.11 mm. Finally, a comparison was made with MODIS-retrieved CLW values, and a good correlation was found with an RMS error of 0.14 mm.