Detection of Unusual Atmospheric and Surface Features by Employing Principal Component Image Transformation of GOES Imagery

The detection of unusual weather events and quasi-weather phenomena is often quite challenging. Many such events are often short-lived and the usual spectral bands available on weather satellites are not designed to look at these phenomena. Instead, many of these events are only detected by differencing the usual spectral bands to obtain the features of concern. The question then arises how some of these features can best be detected. Either the methodology has been previously proven or the satellite analyst experiments to determine which spectral bands and which combinations are optimal. Knowledge of the characteristics of various spectral bandshelps in the process, but with increasing numbers of spectral bands, the choices are sometimes beyond experimental reason and may be more a matter of serendipity.

In this paper the technique of Principal Component Image (PCI) transformation will be used to analyze multi-spectral satellite imagery. Focus will be on how the PCI transform is able to help with the analysis of various atmospheric and land-surface features. Many of these features are not normally what the satellite was designed to detect. The features, however, are of concern and can be hazardous events. Events include: volcanic activity, forest and range fires, dust, snow and ice discrimination, separation of different cloud layers, land/water boundaries, cloud phase, and water vapor features in the atmosphere.

Another useful application of PCI analysis is to simulate multi-spectral products from future satellites whose spectral bands will be modified or expanded. For example, GOES-M, to be launched in 2002, will have a new Imager infrared channel at 13.3 µm, replacing a 12.0 µm band. Potential applications of the new Imager channel can be previewed by PCI analysis using a similar wavelength available on the GOES Sounder.

The PCI examples given will focus on applications to GOES data, but with implications for other multi-spectral satellite imagery as well.