Albedo is an important parameter in surface energy balance, and it is also used in the research of global change and weather prediction. In recent years, due to the abnormal weather changes and the increase of extreme weather, there emerged a great demand for high-quality observation data in environmental and climate research. To better understand albedo, researchers have been constantly updating albedo products. The purpose of this paper is to develop a 250 m albedo product based on the 1 km resolution Global Land Surface Satellite (GLASS) albedo product. Compared with GLASS 1km albedo product, our product has higher resolution and precision. The main idea is to use the texture information from the 250 m band data of Moderate-resolution Imaging Spectroradiometer(MODIS) to scale down the 1km resolution GLASS albedo.There are three steps for the algorithm: simple direct inversion, simple combination of primary products, and downscaling fusion algorithm. Followingly, the 250 m albedo product was validated by comparing to the ground data from 30 sites in North America. It is found that the 250 m albedo product data is closer to the ground measurements. The accuracy of GLASS 1km product is 0.0257, while that of the 250 m product is 0.0235. Finally, an application of the product was demonstrated through analyzing the inter-annual albedo change trend of Beijing based on the 250m albedo product from 2003 to 2013. The results show that the urbanization of Beijing decreased the average albedo rather than increased, and that albedo changes caused by natural processes or changes in planting patterns and varieties were more significant than those caused by urbanization. Besides, the result shows that the 250 m albedo product is not only with better visual presentation and clearer texture, but also capable to reveal the spatial details of albedo changes.