LITHOLOGIC ANALYSIS FROM MULTISPECTRAL THERMAL INFRARED DATA OF THE ALKALIC ROCK COMPLEX AT IRON-HILL, COLORADO

Airborne thermal-infrared multispectral scanner (TIMS) data of the Iro n Hill carbonatite-alkalic igneous rock complex in south-central Color ado are analyzed using a new spectral emissivity ratio algorithm and c onfirmed by field examination using existing 1:24 000-scale geologic m aps an petrographic studies. Color composite images show that the alka lic rocks could be clearly identified and that differences existed amo ng alkalic rocks in several parts of the complex. An unsupervised clas sification algorithm defines four alkalic rock classes within the comp lex: biotitic pyroxenite, uncompahgrite, augitic pyroxenite, and fenit e + nepheline syenite. Felsic rock classes defined in the surrounding country rock are an extensive class consisting of tuff, granite, and f elsite, a less extensive class of granite and felsite, and quartzite. The general composition of the classes can be determined from comparis ons of the TIMS spectra with laboratory spectra. Carbonatite rocks are not classified, and we attribute that to the fact that dolomite, the predominant carbonate mineral in the complex, has a spectral feature t hat falls between TIMS channels 5 and 6. Mineralogical variability in the fenitized granite contributed to the nonuniform pattern of the fen ite-nepheline syenite class. The biotitic pyroxenite, which resulted f rom alteration of the pyroxenite, is spatially associated and appears to be related to narrow carbonatite dikes and sills. Results from a li near unmixing algorithm suggest that the detected spatial extent of th e two mixed felsic rock classes was sensitive to the amount of vegetat ion cover. These results illustrate that spectral thermal infrared dat a can be processed to yield compositional information that can be a co st-effective tool to target mineral exploration, particularly in igneo us terranes.