Masters Thesis

A landslide study in the Redwood Creek basin, northwestern California : effects of the 1997 storm

Analysis of aerial photographs following the January 1997 storm in the Redwood Creek basin, north coastal California, documented 250 fresh landslides greater than 400 m2 in area. Data on failure type, hillslope position, hillslope form, hillslope gradient, bedrock type, road association, timber harvest unit association, and percent sediment delivery to stream were collected for each landslide. Storm rainfall was of moderate intensity and the return intervals for the 1-, 6-, and 24-hour rainfall intensities were less than 2 years, except the 24-hour rainfall intensity at the O'Kane gage (upper Redwood Creek) which had a return interval of 8 years. The peak flow recurrence interval was 12 years. The antecedent precipitation index (API) was 188 mm, the second highest API at the Orick gage in the Redwood Creek basin since 1950. This API was higher than that for the 1964 storm, which is the storm of record in the basin. Total landslide material volume was about 570,000 m3, and estimates of delivery of landslide material to Redwood Creek and its tributaries range between 345,000 m3 and 496,000 m3. The most significant risk factors in the initiation of landslides in the Redwood Creek basin were the inner gorge hillslope position, the Grogan Fault Zone bedrock type, and the 'very steep' hillslope gradient categories. These factors are correlated because the Grogan Fault Zone bedrock type lies primarily along Redwood Creek, as does the inner gorge, and, by definition, the inner gorge has hillslope gradients greater than 50%. Preliminary results showed that landslide volume per unit area derived from the Grogan Fault Zone bedrock type was 100 times greater than the basin average of 930 m3/km2 after separation of hillslope gradient and bedrock type in the inner gorge. Commercial timber harvesting in the upper basin of Redwood Creek has occurred in the inner gorge since the 1930's and has continued through the present. Sediment input from inner gorge areas containing harvest units was over three times greater on a per unit length basis than in inner gorge areas not containing harvest units. Twenty-nine kilometers of the inner gorge was harvested from 1977 to 1997. If the inner gorge had not been harvested, there could have been a reduction of 11% (or 47,000 m3) of the sediment that was delivered to the study area. Sixty-three percent of the total landslides in this study were associated with roads. Landslides associated with roads contributed 2.6 times more sediment to Redwood Creek and its tributaries on a per unit area basis than the basin average.

Items in ScholarWorks are protected by copyright, with all rights reserved, unless otherwise indicated.