Algorithms for the extraction of chemical absorption signatures in lidar time series

Chemical absorption signatures in lidar data can be difficult to identify when the signal to noise ratio is small. For lidar interrogation of unknown chemical mixtures it is advantageous to sample with many different wavelengths, covering the largest possible absorption bandwidth. A more effective DIAL measurement can be made if one knows a priori which wavelengths will be absorbed by the unknown chemical(s). An algorithm has been developed which quickly identities the absorbed laser lines by examining the temporal cross-correlation between wavelengths. Once this determination has been made the remote chemical mixture can be re-sampled with fewer wavelengths resulting in higher data rates at the sensitive wavelengths. This algorithm teas shown to be successful with actual DIAL measurements of remote chemical mixtures. A second detection algorithm will also be presented that uses the temporal autocorrelation at a single wavelength to detect the presence of a time dependent chemical absorption, e.g. from a chemical plume, in a noisy time series. The overall DIAL sensitivity using these algorithms will be compared with standard methods.