Earth observation uses spaceborne sensors to obtain complete and systematic information about the earth. With the help of advances in spatial information technologies, the field of earth observation has entered the era of “big data”. By analyzing the entire process of earth observation data processing and associated algorithms, this paper concludes that the challenges facing earth observation “big data” processing are those of data-intensive computing. We suggest that to solve this problem, there should be research and innovation in the areas of system platforms, processing algorithms, and services models.