DIFFERENCES IN LANDSLIDES AND DEBRIS FLOWS ASSOCIATED WITH WINTER STORMS OF 2005 AND 1997 ON THE KLAMATH NATIONAL FOREST, CENTRAL KLAMATH MOUNTAINS CA

The December 30, 2005 storm event triggered landslides in much of NW California, resulting in widespread damage to roads & infrastructure. A similar storm in January, 1997 caused significantly more damage. Hourly precipitation measurements at Orleans, CA revealed that in the three days prior to peak flows, more rain fell in 2005 than in 1997. However, there was a lull in precipitation during that time in 2005 which was not mirrored in 1997. Although the 2005 event had greater total rainfall, the 1997 rainfall event had a greater 1–hour max rainfall intensity of 1.0 inches, compared to 0.56 inches in 2005. Peak discharge at Indian Creek (Happy Camp, CA) was 8 percent higher in 1997 than in 2005, and the lull in precipitation in 2005 was reflected by a trough in the hydrograph.

During the 1997 storm, landslides & debris flows were widespread throughout the Klamath National Forest, though there were concentrations in some areas. Many of the landslides were on roads, and on de-vegetated slopes, but there were also some on undisturbed forested slopes. Many wildfire-related debris flows occurred in the Elk Creek watershed (Wooley Creek Batholith), which had burned 10 years earlier. In contrast, landslides during the 2005 storm were mostly associated with roads, and few occurred on undisturbed forested slopes. A few wildfire-related debris flows occurred in a granitic portion the Swillup Creek watershed (Pony Peak Batholith) which had burned 4 years earlier. Though more localized than those in 1997, one of the debris flows in Swillup Creek temporarily blocked CA Highway 96.

The large differences in landslide response between these two very similar storms could be due to a variety of factors, such as local variations in storm intensities, land management practices, antecedent moisture conditions, or snow melt. The existing rain gages in the Klamath National Forest do not adequately describe the spatial and orographic variability of rainfall intensity during intense winter storms. An array of tipping bucket rain gages has been installed which should provide this information during the next landslide episode. This information will allow for better isolation of the effects of management practices from those of rainfall variations, and also serve as the foundation of a landslide warning system.