2004 Denver Annual Meeting (November 7–10, 2004)

Paper No. 7
Presentation Time: 3:00 PM


PLOTNICK, Roy E., Earth and Environmental Sciences, Univ of Ilinois at Chicago, 845 W Taylor St, Chicago, IL 60607-7056, plotnick@uic.edu

The data used as the basis for the reconstruction of earth history are either implicitly or explicitly spatial; i.e., their coordinates in 1-, 2-or 3-dimensional spatial frameworks are relevant to their analysis and interpretation. For example, the field description of the occurrence of a particular fossil taxon includes not only the associated lithology, but its geographic coordinates and its location in the overall section. Similarly, the interpretation of the occurrence of a paleoclimatic indicator, such as an evaporite deposit, requires detailed knowledge of both its spatial and temporal location. Paleoenvironmental and paleogeographic reconstructions require accurate description and interpretation of the spatial distribution of the underlying data. Similarly, the output of many earth science models, such as basin evolution and paleoclimate models, is also spatial. These models are generally assessed by the perceived agreement between spatial patterns predicted by the model and corresponding values from the data. Reconstructing regional or global events and patterns and testing geoscience models thus requires the ability to integrate a wide variety of spatial information. STRATISTICS is a toolkit for the spatial characterization of data relevant to Earth history and for the quantitative comparison of these datasets with the models used to simulate them. The software makes available a broad suite of quantitative and visualization techniques for the characterization of the spatial properties of geologic data and for the comparison of data and model output derived from multiple sources. These techniques are based on methods developed in the field of spatial statistics and landscape ecology, which can be directly applied to the study of spatial heterogeneity in earth science. The new toolkit complements other proposed CHRONOS toolkits, which will focus on areas such as quantitative biostratigraphy and time-series analysis.