I have a basic knowledge of GIS and arcgis sofware and I need to generate C factor (cover management factor of RUSLE) by using NDVI values. I know there is an exp formula but I dont know the steps to create the map. It would be perfect if I know that.
I believe that the Vander Knijff, et al., (1999) equation is
C = exp(-alpha(ndvi / (beta - ndvi))) where; alpha and beta define the shape of the curve and are commonly defined as 2 and 1 respectively.
I have seen the application of an OLS regression, based on field or ancillary data, to calibrate the curve to local conditions and correctly bound the C-factor values. Literature that fits a linear regression directly to NDVI produces somewhat dubious results. If you look at the actually fit, the authors are functionally fitting the regression between two points. There is actually quite weak correlation between NDVI and measured C-factor values. Thus, I believe that the transformation is an imperative.
If you are using OLI (Landsat 8) you may want to consider using the Atmospherically Resistant Vegetation Index, as an alternative to NDVI. The ARVI incorporates the blue-edge band and addresses any potential aerosol effects as well as being robust to index fluctuations due to parallax.