New Method for Estimating Landslide Losses for Major Winter Storms in California.

Abstract:

We have developed a prototype system for estimating the economic costs of landslides due to winter storms in California. This system uses some of the basic concepts and estimates of the value of structures from the HAZUS program developed for FEMA. Using the only relatively complete landslide loss data set that we could obtain, data gathered by the City of Los Angeles in 1978, we have developed relations between landslide susceptibility and loss ratio for private property (represented as the value of wood frame structures from HAZUS). The landslide loss ratios estimated from the Los Angeles data are calibrated using more generalized data from the 1982 storms in the San Francisco Bay area to develop relationships that can be used to estimate loss for any value of 2-day or 30-day rainfall averaged over a county. The current estimates for major storms are long projections from very small data sets, subject to very large uncertainties, so provide a very rough estimate of the landslide damage to structures and infrastructure on hill slopes. More importantly, the system can be extended and improved with additional data and used to project landslide losses in future major winter storms. The key features of this system—the landslide susceptibility map, the relationship between susceptibility and loss ratio, and the calibration of estimates against losses in past storms—can all be improved with additional data. Most importantly, this study highlights the importance of comprehensive studies of landslide damage. Detailed surveys of landslide damage following future storms that include locations and amounts of damage for all landslides within an area are critical for building a well-calibrated system to project future landslide losses. Without an investment in post-storm landslide damage surveys, it will not be possible to improve estimates of the magnitude or distribution of landslide damage, which can range up to billions of dollars.