Modeling the complexity of different, recently deglaciated soil landscapes as a function of map scale

Abstract

The scale at which a soil landscape (soilscape) is viewed has a significant impact on soil pattern and interpretations made from those patterns. Recently deglaciated soilscapes are particularly spatially complex. In order to understand how scale impacts pattern on complex soilscapes, we used a GIS to examine soil maps for 13 counties in the northern United States, all affected by Late Wisconsinan glaciation. We used an Arc™ macro language script to change the map scale and, when the change was to a smaller scale, group/dissolve soil map units based on similarities to a prescribed list of neighboring map unit characteristics. Similarity criteria included drainage class, taxonomic great group, parent material and slope. Soilscape complexity was measured at nine different scales and is based on various pattern metrics: number of punctate soil units km −2, map unit polygons km −2, map unit boundary length km −2, and boundary length polygon −1 km −2. Soilscape complexity as a function of scale was then examined by regressing pattern metric data against the size of the minimum map unit for each of the nine scales. Extrapolation of the regression lines to 1:10,000 (a scale larger than is typically mapped) illustrated how much additional information might accrue if these counties were to be mapped at that larger scale. In most cases, 2–10 times more map units would have been recognized and delineated at the two times larger map scale, but map unit boundary lengths would have increased by only about 1.5 times. Whether this additional information is of such a magnitude that it could justify remapping some of these complex landscapes at larger scales is an economic decision; our study provides much needed data on the magnitude of information gained by mapping soilscapes at larger scales.