I am using the RUSLE model to estimate erosion. I have all the factors except SL which when i added to the RUSLE equation returned an error. So i decided to re-run the SL (slope length) calculation again but i am having problems with multiplying Flow accumulation by cell size and other constants. Attached is the equation i am using from (Lahlaoi et al,2015). Any suggestions will be welcome
It has been some time since my experience with RUSLE. In reading about the L factor, I remembered an unusual assumption was made. The L length factor does not go from ridge to stream. The distance may have been related to how long the slope would be when there are signs of concentrated flow erosion. The USLE was developed to address surface soil loss, and not to address rilling, gullying, channel or bank erosion. In addition, if you use L from ridge to channels, there is a difference if you have stream density based on LiDAR DEM or based on the old school topographic map DEM. We typically an L distance based on past measured values for physiographic area, but if I run across the Wischmeyer paper, I think that might have been the one, or a paper by Dissmeyer. A typical R factor for SC Piedmont is 275 in English units in 100 (ft tons/acre)(in/hour) and 45 inches rainfall, the L factors applied if we dont have better data are typically 100-150 feet, based on values gathered by Dissmeyer and Stump, 1978. I wish I remembered more, but the L factor has a specific intent that should be considered specific to your area. If you want to consider all forms of erosion including channel and bank, you might consider several other additional approaches. We have used one developed by Dave Rosgen with info at www.wildlandhydrology.com, the BEHI method that qualitatively evaluated bank height factors, and shear stress indicators of channel.
I googled on RUSLE L factor and easily found some guidance from USDA NRCS in Michigan. L is distance from where flow first concentrates to a recognizable channel or stream. I would suggest that in some circumstances, this is difficult to measure, such as in a hardwood forest after leaf fall. They warned that contour intervals over 2 feet elevation are a concern, the distance should be field verified to get appropriate value, distances over 400 feet are seldom, and distances over 1000 feet are probably outside of model feasibility. It also suggests the answer is not that sensitive to L, but should be in ballpark, but when they said plus or minus 10%, that is basically a site measure. The slope portion is the gradient of the ground. Since the USLE was done in USA with English units, be sure to check equations if conversions are made to metric units. There is another question that discusses issues with flow accumulation in ArcGIS10.1 that you might want to look over. There are flow accumulation issues to address, sometimes by hand.
@William, thanks, the slope length is estimated using between 100-150 as you stated, in this case the formular used 120 which is around 22.1m i think. I have been trying other formulars but still receiving the same error feed back. I have tried to rework the inputs but same results, For instance slope is defined in degrees, although some suggest %. not sure now what i might be missing.
Sorry, I have never touched such a software product. So I cannot help you.
I have been able to solve the problem, i think through more careful structuring of the input values. Thanks to all for your suggestions
- Badges
- Science topic
- Similar topics
- Geoscience
- Asia
- India