2003 Seattle Annual Meeting (November 2–5, 2003)

Paper No. 12
Presentation Time: 11:45 AM

HAZARD MAPS FOR RAPIDLY MOVING LANDSLIDES USING GIS AND LIDAR


GRISWOLD, Julie, USGS, Vancouver, WA and Dept. of Geology, Portland State Univ, P.O. Box 751, Portland, OR 97207-0751 and IVERSON, Richard, U.S. Geol Survey, Building 10, Suite 100, 1300 SE Cardinal Court, Vancouver, WA 98683-9589, griswold@pdx.edu

Hazard maps for lahars, non-volcanic debris flows, or rock avalanches can be prepared using a GIS methodology that relies on statistically calibrated predictive equations (e.g. Iverson et al., 1998, GSA Bull. v. 110, p. 972-984), LAHARZ software (Schilling, 1998, USGS OFR 98-638), and digital elevation models (DEMs) to compute and depict patterns of inundation downstream from potential source areas. Whereas coarse, low-resolution DEMs derived from USGS topographic maps are adequate for assessing hazards from large volcanic lahars, high-accuracy, high-resolution DEMs are necessary to compute and depict hazards from non-volcanic debris flows that are typically smaller than 100,000 cubic meters. Examples are shown of computed hazard zones for prospective lahars and rock avalanches at Mount Rainier, Washington and for non-volcanic debris flows along the Umpqua River in southern Oregon. The Mount Rainier DEMs have a minimum resolution of 10 meters. The Umpqua River DEM, generated from LIDAR (light detection and ranging technology) data obtained from the Oregon Department of Forestry, has a resolution of 1 meter.

Uncertainty and error in the hazard maps are addressed in two ways. Standard statistical techniques are used to assess the error in the semi-empirical equations that predict the maximum valley cross-sectional areas and total downstream planimetric areas likely to be inundated by landslides of various volumes. Nested hazard zones showing inundation limits for prospective landslides with a range of volumes depict the uncertainty in predicting the volumes of future landslides descending any particular drainage. Hazard zones generated in this way appear comparable to those generated by traditional field mapping techniques but are objective and reproducible.