| Abstract: | Debris flow frequency and magnitude were determined for 33 basins in southwest British Columbia. Basins were first classified as either weathering-limited or transport-limited using a discriminant function based on debris-contributing area, an area-weighted terrain stability number, and drainage density. Multiple regression was used to predict magnitude, peak discharge, frequency and activity (frequency times magnitude) within each group of basins. Model performance was improved by stratifying the total sample of debris flow basins into weathering-and transport-limited groups. Explained variance increased by an average of 15 per cent in the transport-limited sample, indicating that sediment supply conditions in the more active basins are fundamental in predicting debris flow activity. An independent test of the regression models with 11 basins yielded generally good results for debris flow magnitude and peak discharge. Prediction of debris flow frequency proved problematical in weathering-limited basins. The methods developed here provide estimates of debris flow attributes in basins for which few data on past events are available. Copyright © 1999 John Wiley & Sons, Ltd. |
