A FOSSIL LOCALITY PREDICTIVE MODEL FOR THE EARLY CRETACEOUS CEDAR MOUNTAIN FORMATION, UTAH, USA

Hard work and chance are nearly always among the deciding factors in finding new, important, and productive paleontological localities. Fossil locality predictive models have the potential to reduce unproductive field time and maximize hard work thus increasing the chances researchers have to find important localities. This study uses remotely sensed data to design and test a fossil locality predictive model for the Early Cretaceous Cedar Mountain Formation as mapped on the 2002 USGS geologic map of the Moab, Utah 30'x60' quadrangle. Landsat 8 OLI/TIRS spectral reflectance data covering areas of known fossil localities (provided by the Brigham Young University Museum of Paleontology) were summarized, reclassified and used in a weighted suitability analysis to categorize fossil locality potential of the study area.

Ground-truthing through field testing was conducted to test model functionality. Physical and environmental factors beyond spectral reflectance play a role in determining the chance of finding fossils. Very steep slopes are dangerous to traverse and very shallow slopes are often recent depositional surfaces. Northern aspects in the study area are more heavily vegetated and offer fewer opportunities to expose fossils due to regional dip. Landsat 8 spectral reflectance data alone offers a less accurate prescription for fossil locality potential. Field observations concerning surface slope and aspect were used to refine the weighted suitability analysis. Slope and aspect data from the BYU localities were summarized, analyzed, and used to refine the model. The model was tested through additional field work.

The usefulness of fossil locality predictive models is dependent upon the quality of input data and methods used to determine fossil locality potential. There is no one predictive model solution for all types of field areas and those creating them must have a strong understanding of local environmental conditions. Using predictive models is no substitute for field work, but is an additional tool to further paleontological research and to aid in resource allocation.