Technical reportAbstract: The granite tors on Dartmoor are characterized and described by a combination of 21 geomorphic, petrographic and structural variables using two statistical approaches: analysis of significant correlations and multivariate analyses. Whereas an analysis of correlations enables relations between pairs of individual variables to be determined, the multivariate procedures addresses all variables at the same time. Spearman's rank correlation coefficient was used to correlate the variables so that nominal and ordinal variables could be included. The characteristics of summit tors and spur tors were defined by analyzing the significant correlations as well as by analyzing indirect relations between related variables. The multivariate procedures separated the sample of 58 tors into five clusters. Principal coordinates analysis and a non-hierarchical classification were used to determine the characteristics of each cluster. The joint spacing frequency distributions for the clusters were compared to determine statistically significant differences between them; each group is significantly different from at least two others. Comparing the results of the two procedures indicates that although all variable types are important to landform development, the geologic factors are most important. The most important variable types are rock texture, grain size, composition and structure.

U.S. Army Topographic Engineering CenterEngineer Research and Development Center (U.S.)

Map background raster; Color separates; Resolution; Color normalization; Declutter

TEC ; 0027.

Approved for public release; distribution is unlimited.

DOCUMENT