DEVELOPMENT OF SOIL SLIP SUSCEPTIBILITY MAPS FOR SOIL SLIPS GENERATED DURING WINTER RAINS, SOUTHWESTERN CALIFORNIA

Empirically derived soil slip susceptibility maps were made for 128 7.5’ quadrangles in southern California (http://geopubs.wr.usgs.gov/open-file/of03-17). Prompted by a prediction of a wet 1997-98 winter, 15 aerial photography transects were selected for study. These 1:24,000-scale transects, 5 miles wide and 8 to 20 miles in length include a wide variety of geology and physiography. During February 1998 soil slips were generated in 12 of the transects and over 20,000 soil slips-debris flows were mapped and systematically digitized. The susceptibility maps were based on data obtained from analyses of soil slips in these transects that extended from Lake Cachuma area south to Dana Point and inland to the Beaumont area. An additional transect was added after heavy rainfall in March, 2001in the Las Cruces area, western Santa Barbara County. Detailed analyses were made in three areas, Santa Paula (Hauser, 2000), Sunland (Koukladas, 1999), and in the San Timoteo Badlands, a semiarid area.

Geology, slope, and aspect, were the three most important variables in determining the location of soil slips. Using digital geologic maps of the Southern California Areal Mapping Project, SCAMP, more than 700 geologic units were grouped into 8 susceptibility categories based on geology. The 8 catagories were given values from 0 to 25 which were then assigned to over 800 million 5-m cells. 5 slope and 5 aspect values were assigned to more than 200 million 10-m cells derived from 10-m DEMs. A susceptibility algorithm was developed from the product of these three variables. Calculated susceptibility values, assigned to 10-m cells, ranged from zero to 5,000. For map depiction, these values were converted to four map categories: 0-5, no soil slips (uncolored); 6-999 low susceptibility (green); 1,000-3,250 moderate (yellow); and 3,251-5,000 high (red). Slope calculation is a major problem; conventional procedure is to calculate the slope for a cell from the eight surrounding cells. This results in an incorrect low slope value for cells near a summit. To compensate, a high susceptibility value cell was added to the topographically-highest upslope high susceptibility value cell; this resulted in an increase of capturing soil slips in the moderate and high value cells from 69% to 94% for the Santa Paula area, 55%-81% for the Sunland area, and 53% to 86.5 % for the San Timoteo Badlands.