Nonlinear wavelet compression of ion mobility spectra from ion mobility spectrometers mounted in an unmanned aerial vehicle.

Linear and nonlinear wavelet compression of ion mobility spectrometry (IMS) data are compared and evaluated. IMS provides low detection limits and rapid response for many compounds. Nonlinear wavelet compression of ion mobility spectra reduced the data to 4-5% of its original size, while eliminating artifacts in the reconstructed spectra that occur with linear compression, and the root-mean-square reconstruction error was 0.17-0.20% of the maximum intensity of the uncompressed spectra. Furthermore, nonlinear wavelet compression precisely preserves the peak location (i.e., drift time). Small variations in peak location may occur in the reconstructed spectra that were linearly compressed. A method was developed and evaluated for optimizing the compression. The compression method was evaluated with in-flight data recorded from ion mobility spectrometers mounted in an unmanned aerial vehicle (UAV). Plumes of dimethyl methylphosphonate were disseminated for interrogation by the UAV-mounted IMS system. The daublet 8 wavelet filter exhibited the best performance for these evaluations.