Generating consistent satellite land surface albedo products across scales using a data fusion method

Land surface albedo is one of the key parameters in land surface modeling and climate change studies. In the past several decades, global surface albedo datasets have been developed from multiple satellite sensors. However, existing albedo products suffer from several problems, such as cloud contamination, algorithm limitation, and sensor failure, which may bring gaps or reduced accuracy for climate modeling applications. A novel approach was proposed in this paper by fusing multiple satellite albedo products across different spatial scales to reduce gaps and improve consistency. To implement the prototype algorithm, three satellite albedo products from the Multi-angle Imaging Spectro-Radiometer (MISR), Moderate Resolution Imaging Spectroradiometer (MODIS), and Landsat were used to generate consistent albedo datasets at different spatial resolutions simultaneously.