USING GIS TO ASSESS MAPPING INCONSISTENCIES ALONG THE GLACIAL MARGIN IN WESTERN NEW YORK

Depending on whether soil characteristics or surficial geology is the major tool used, the mapped position of late-Wisconsin ice extent in Cattaraugus County, New York, varies by several kilometers. Erosional histories should have created landscapes with different morphometry; smooth, round hills and minimal elevation range in the glacial regions; steep, rectilinear slopes, in the fluvially dissected terrain south of the margin. A regional-scale morphometric analysis using 10x10 m DEMs in ArcGIS and ArcInfo may provide a way to visualize and analyze patterns not observable from field-based examination. The purpose of this study is to assess the value of digital terrain analysis using a GIS for improving confidence in ice margin delineation.

Digital data sets used in the analyses were bedrock and surficial geology and soils. SQL statements in ArcGIS on surficial geology enabled selection and classification of kames, kame moraines, till and till moraines as “glacial”, and colluvium as “nonglacial”. A hand-digitized map of ice extent, mapped according to soil characteristics (Bryant 1955), was classified as “glacial” and “nonglacial”. Overlay analysis identified the mismatch between the two reclassified maps and was used to create polygons designated as “glacial”, “nonglacial” or “difference”. These “zones” were used as the independent variable in further analyses. The DEMs were used to calculate morphometric parameters of elevation, slope, terrain ruggedness (standard deviation and range of elevation calculated in a 5x5 neighborhood), and slope planform curvature and profile curvature. Zonal statistics using “zones” as a mask were run in GRID, and ascii data files were extracted for running ANOVA in a conventional statistical package.

Two-way crosstabs indicates that entisols are the predominant soil order in the glacial areas and areas of difference. ANOVA using morphometry shows significant differences in all zones in elevation, slope, and the two ruggedness parameters. No significant difference exists between measures of slope form. For slope and ruggedness parameters, the area of difference is most similar to the nonglacial area. The results from soil and morphometry prove contradictory; however, they do provide insight into why precise delineation of the margin has been so difficult.