The application of control charts to determine the effect of grid cell size on landform morphometry

Geoscientists have become increasingly dependent on digital elevation models (DEMs) to delineate and measure landforms and landscapes. However, the DEM grid cell size available may not be the optimum resolution; this can mask subtle changes in measurements and lead to erroneous results. This paper presents a standardized statistical technique (i.e. statistical process control charts (SPCC)) for determining the optimum DEM resolution (i.e. the coarsest resolution in which detail is not sacrificed) for landforms (e.g. drumlins). For this study, forty-four DEM resolutions, ranging from 1 to 80 m, were used to assess the effect of resolution on drumlin size, shape, and centroid. The results indicate that the optimum resolution for the size variables (width and length) was coarser than the optimum resolution for shape indices (elongation and rose curve). Drumlin location tends to drift in a predictable direction and rate as grid cell size coarsens above particular thresholds. The results prove that resolution plays a critical role in correctly evaluating drumlin morphometry and that care must be taken when utilizing DEMs to summarize drumlin characteristics. The creation of a standardized technique to describe drumlins will allow for scrutiny of previous work and straightforward comparative analyses between studies, while utilizing the optimum resolution will help decipher landform patterns, reveal relationships, and provide more insight into landform evolution. (C) 2009 Elsevier Ltd. All rights reserved.