Studies of High Spectral Resolution Atmospheric Sounding Data Compression and Noise Reduction Based on Principal Component Analysis Method

The principal component analysis uses the variance as a standard for the measure of information content. By linear transformation using the principal component analysis method, the components with more information can be reserved while the ones with less information can be discarded, thus the feature extraction or data compression can be realized. In this paper, using the principal component analysis, studies and experiments of the data compression and noise reduction for high spectral resolution atmospheric sounding data have been carried out. The results show that the algorithm is efficient data compression and noise reduction, and the precision of the retrieved atmospheric parameters is also improved.