A Comparison of GIS Approaches to Slope Instability Zonation in the Central Blue Ridge Mountains of Virginia

Files
TR Number
Date
2004-10-22
Journal Title
Journal ISSN
Volume Title
Publisher
Virginia Tech
Abstract

To aid in forest management, various approaches using Geographic Information Systems (GIS) have been used to identify the spatial distributions of relative slope instability. This study presents a systematic evaluation of three common slope instability modeling approaches applied in the Blue Ridge Mountains of Virginia. The modeling approaches include the Qualitative Map Combination, Bivariate Statistical Analysis, and the Shallow Landsliding Stability (SHALSTAB) model. Historically, the qualitative nature of the first model has led to the use of more quantitative statistical models and more deterministic physically-based models such as SHALSTAB. Although numerous studies have been performed utilizing each approach in various regions of the world, only a few comparisons of these approaches have been done in order to assess whether the quantitative and deterministic models result in better identification of instability.

The goal of this study is to provide an assessment of relative model behavior and error potential in order to ascertain which model may be the most effective at identifying slope instability in a forest management context. The models are developed using both 10-meter and 30-meter elevation data and outputs are standardized and classified into instability classes (e.g. low instability to high instability). The outputs are compared with cross-tabulation tables based on the area (m²) assigned to each instability class and validated using known locations of debris flows. In addition, an assessment of the effects of varying source data (i.e. 10-meter vs. 30-meter) is performed. Among all models and using either resolution data, the Qualitative Map Combination correctly identifies the most debris flows. In addition, the Qualitative Map Combination is the best model in terms of correctly identifying debris flows while minimizing the classification of high instability in areas not affected by debris flows. The statistical model only performs well when using 10-meter data while SHALSTAB only performs well using 30-meter data. Overall, 30-meter elevation data predicts the location of debris flows better than 10-meter data due to the inclusion of more area into higher instability classes. Of the models, the statistical approach is the least sensitive to variations in source elevation data.

Description
Keywords
slope instability, GIS, terrain analysis, debris flows
Citation
Collections