Retrieval of GF-4 Satellite Image Data Surface Albedo Based on Angular Bin Algorithm
摘要
Albedo has important significance to the research of the surface energy balance and climate. GF-4 satellite is the earth synchronous orbit satellite, whose spatial resolution is 50 m, much better than other same orbit satellites. Because GF-4 only has one observation angle on a site, it's difficult to use bidirectional reflectance distribution function (BRDF) inversion method to retrieve the albedo. So, we use angular bin (AB1) algorithm to retrieve the albedo. AB1 is based on priori knowledge. Firstly, we use the data of Polarization and Directionality of the Earth's Reflectances (POLDER) to found the empirical explanation between albedo and the reflectivity of one angle, and then convert POLDER's empirical explanation to GF-4 by the spectral response functions of POLDER and GF-4.Next Plug in one GF-4's image and calculate the albedo using the empirical explanation. Last we use MODIS and Landsat 8 albedo products to verify the calculation results. The results show that the root mean square error (RMSE) of MODIS albedo and GF-4 albedo is 0.027 3, the RMSE of Landsat 8 albedo and GF-4 albedo is 0.017 6, verify that the accuracy of albedo inversion based on AB1 is reliable, proving that using the AB1to retrieve the GF-4's albedo has good precision.