Applications of remotely sensed data to geological problems

Despite the widespread applications of remote sensing in other fields, systematic geology has made only limited use of these techniques. This is largely because instruments capable of performing mineralogic discrimination are of relatively recent vintage. Moreover, the instruments capable of the degree of resolution necessary to differentiate spectral lithofacies must be carefully calibrated if results are to be comparable between data sets.;This study performs systematic calibration and coregistration of four multispectral instruments (TM-4, TM-5, TMS, and TIMS) and applies them, along with a coregistered digital topographic data set, to a study area of the Wind River Basin, Wyoming. The spectral units identified using TMS and TIMS data are applied to TM data to develop an overview of a large section of the basin. Apparent anomalies in the elevations of recurrent beds in the scene appear to be resolved in harmony with paleontological results by applying a topographic trending technique to the data.;Higher resolution instruments are currently under development for more detailed airborne mineralogy. These new instruments are much more sensitive to both the spectral characteristics of targets and to distortion of spectral signatures by atmospheric absorption. A number of past approaches have been made to calibration of such data sets, but these techniques are not always practical or possible, and do not lend themselves to automated processing techniques.;This study develops a technique, based on LOWTRAN line parameters, for removing atmospheric effects from the data based on the information in the data set itself. Because atmospheric effects are consistently relatable to the density of the absorbing species along the radiation path, it is shown to be possible to in turn estimate the concentration of the various absorbing species, and then remove their contribution from the remotely sensed data. This technique is demonstrated using convolved laboratory spectra and LOWTRAN parameters to show its applicability to spectra of geological relevance. The results indicate that this technique could be of considerable value when applied to hyperspectral instruments with sufficient primary systems calibration.