[A-CG37] Satellite Earth Environment Observation

[ACG37-P01] Development of a correction method for 3D radiation effects in satellite observations of clouds

*Takayuki Masuko¹, Hironobu Iwabuchi¹ (1.Tohoku University )

Keywords:Cloud, Radiative transfer, Remote sensing

Clouds are an important component of the Earth's radiation budget and have a significant impact on the future climate of the Earth. The cloud optical thickness (COT), which is particularly important from the viewpoint of radiation budget, has been observed globally by satellite observations. However, current satellite-based COT estimation methods consider only plane-parallel, one-dimensional vertical radiation transport and do not take into account real three-dimensional radiation transfer and heterogeneous clouds. As an example of typical errors caused by this assumption, it has been pointed out that when the solar zenith angle is large, the COT overestimates and underestimates the illuminating effect, in which reflectance is higher on the side of the cloud illuminated by the sun, and the shadowing effect, in which reflectance is lower on the opposite side because the cloud is in shadow.
In this study, a deep learning model was developed to correct the reflectance by using the 3D radiative transfer calculation that takes into account cloud heterogeneity and the reflectance that could be calculated by the conventional method of 1D radiative transfer as the teacher data. Based on the corrected reflectance, the COT errors were compared using a method that mimics the COT estimation of conventional satellite observations. The results showed that the COT error estimated by the corrected reflectance was 33.2%, which was significantly lower than the RMSE of 93.8% obtained by the conventional method. The case analysis also confirmed that the reflectivity was corrected. Since this study was conducted under the condition of satellite nadir view, the next goal is to develop the method so that it can be applied to various viewing angles.