Abstract: Satellite data is of signi ficance on the accuracy of numerical forecast. The satellite data were used for global atmospheric reanalysis collected from more than 80 satellites, about 60 kinds of space-borne sensors since 1979. There are a number of dif ficulties in the satellite observational data format conversion, long time series data application and homogenization. So far, 83.5 TB in 53 kinds (115 sub kinds) of meteorological satellite data are collected and integrated for researches on the CMA's global atmospheric Re-Analysis (CRA-40, 1979-2018) in the first stage. This paper reviewed and summarized the satellite data collection and integration for global atmospheric reanalysis, the data pre-processing such as data format conversion and the pre-evaluating satellite data quality. We focus on the CRA-Interim (2007-2016) reanalysis products. The reprocessed satellite products released for recent years will replace the old operational products. These historical and real-time satellite data are achieved and shared gradually via the China Integrated Meteorological Information Service System (CIMISS).