干旱区生态环境知识资源中心

The intercomparison of MetOp-A and MetOp-B Advanced Very High Resolution Radiometer visible and near-infrared (NIR) channels over the Libyan Desert with consideration of the effect from the bidirectional reflectance distribution function (BRDF) is used for MetOp-B postlaunch calibration update. In order to remove this effect from the comparison, two methods are investigated: BRDF effect modeling in direct comparison and seasonal oscillation regression. The BRDF modeling method employs both a simple linear model and an empirical model. The seasonal oscillation method has improved by adding a curve regression, where the seasonal oscillation is derived from MetOp-A measurement data since its launch and used for the regression of MetOp-B measurements. Using these methods, the reflectance ratios of MetOp-B overMetOp-A in three reflective solar channels have been derived, which are 1.034 for channel 1, 0.912 for channel 2, and 0.805 for channel 3A. The uncertainties of the derived ratios are estimated in the range of 4.7%-6.4%. The model accuracy and uncertainty have been discussed. Initial calibration updates based on these results have been delivered for the MetOp-B operational L1B product, and a routine update is performed monthly. The methods used in this work are also applicable to the intercomparison of other visible and NIR instruments.