Open AccessAn algorithm for estimating downward shortwave radiation from GMS 5 visible imagery and its evaluation over China
Author(s) -
Lu Ning,
Liu Ronggao,
Liu Jiyuan,
Liang Shunlin
Publication year - 2010
Publication title -
journal of geophysical research: atmospheres
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.67
H-Index - 298
eISSN - 2156-2202
pISSN - 0148-0227
DOI - 10.1029/2009jd013457
Subject(s) - shortwave , shortwave radiation , remote sensing , environmental science , geostationary orbit , satellite , sky , meteorology , mean squared error , atmosphere (unit) , atmospheric radiative transfer codes , radiative transfer , radiation , mathematics , geology , physics , optics , statistics , astronomy
This paper presents an operational scheme to estimate downward shortwave radiation (DSR) over China from the visible‐band top‐of‐atmosphere reflectance of the Geostationary Meteorological Satellite (GMS) 5 imagery. The proposed algorithm retrieves surface reflectance and atmospheric parameters directly from GMS 5 images by searching lookup tables, which are created by the radiative transfer model SBDART and consider the effects of water vapor absorption and surface altitude variations. Experiments show that the DSR retrieval is more sensitive to the selection of aerosol type and less to that of the cloud type. Uncertainty in the reflectance of a bright surface leads to a considerable DSR retrieval error (±(6–9%)). The instantaneous retrieved DSR is evaluated by field measurements on the Tibetan Plateau, and the daily retrieved DSR is compared with one year's ground‐based measurements at 96 stations in China. The results show that the estimated DSR is in good agreement with ground measurements with a correlation coefficient of ∼0.9 and a bias of 1.5%. Root‐mean square differences in the daily DSR are 17.7% for all‐sky and 13.1% for clear‐sky conditions. These results suggest that the proposed method applied to the GMS 5 satellite data can accurately estimate temporally and spatially continuous instantaneous and daily DSR. These DSR data sets will be useful for a wide range of applications.