Detection of Sulfur Dioxide in AIRS Data With the Wavelet Packet Subspace

Abstract

A generalized method for trace-gas detection in hyperspectral data using the wavelet packet transform is being developed. This new method decomposes the input signal using a wavelet packet transform. A best basis that is optimized for the target signature is selected for pattern matching. The wavelet packet transform, an extension of the wavelet transform, fully decomposes a signal into a library of wavelet packet bases. The application of the wavelet packet transform to hyperspectral data for the detection of trace gases takes advantage of the ability of the wavelet transform to locate spectral features in both scale and location. By analyzing the wavelet packet tree of a specific target gas, the nodes of the tree that represent an orthogonal best basis are selected. The best basis represents the significant spectral features of that gas. This is then used to identify pixels in the scene using existing matching algorithms such as spectral angle. Using data from NASA's Advanced Infrared Sounder, this method is used to detect sulfur dioxide. Initial results demonstrate a promising wavelet-packet-subspace technique for trace-gas-detection applications.