Data Fusion and Optimization Based on Multiple Satellites XCO2 Data
With the continuous development of satellite observation technology,more and more atmospheric composition data,including CO2,are available.Multiple satellites data fusion is an effective means to improve the accuracy and spa-tiotemporal resolution of data.In this paper,compressed sensing and deep learning techniques were used to optimize and fuse the remote sensing observed data of three passive detection satellites,namely GOSAT,OCO-2,and OCO-3.Com-pressed sensing is used to generate initial fused data,and the deep learning model further enhanced the seasonal cycle feature of CO2 data to optimize the initial fused data.Experimental results showed that the fused data generated by com-pressed sensing were superior to the original satellite observations,and the deep learning model could effectively denoise and optimize the fused data.
万方数据
