A QUANTITATIVE SOLUTION TO THE DRUMLIN DEBATE?

The quantitative analysis of morphological data can provide valuable information on which to base interpretation of geomorphic processes and may aid in the resolution of ongoing debate regarding the origin of glacial landforms such as drumlins. This paper presents the results of spatial and non-spatial statistical analysis of over 1000 drumlins within the Peterborough Drumlin Field (PDF) of Ontario. The PDF, located 125 km north-east of Toronto, covers an area of over 5000km2 and contains many thousands of individual drumlins formed during late Wisconsin glaciation. The drumlins within the PDF have been interpreted both as erosional ridges associated with catastrophic subglacial meltwater floods and as geomorphological expressions of a deforming subglacial bed.

Drumlins were identified within a section of the drumlin field (approximately 1800 km2) from 10m resolution DEM data made available from the Ontario Geospatial Data Exchange (OGDE), air photos and remotely sensed satellite data. An automated process, created with a Geographic Information System (GIS), was designed to calculate landform variables such as length, width, circularity and long axis compass direction from the digital maps. This allowed relatively rapid identification of drumlin shapes which could be then classified and grouped according to their morphological and directional attributes.

Quantitative analysis of the morphological variables included the application of non-spatial techniques such as Pearson Correlation Coefficient and analysis of variance (ANOVA) to determine if any of the morphological variables are related. Spatial analytical techniques such as exploratory cluster analysis, analysis of direction of long axis and the visualization of kernel densities were also applied to determine the spatial distribution of drumlin forms.

Differences in directionality of drumlin long axes across the study area identify regions with complex ice flow histories, particularly close to the Lake Ontario shoreline, and raise the possibility of multiple phases of drumlin formation. Spatial variations in drumlin form within the study area do not show significant relationships with bedrock or sediment type but may be related to regional variations in drift thickness.