I have an idea to create the values of the biophysical table of the SYW model. Mainly the Kc values using the NDVI as input and using equations 13 and 14 from the following paper:
Kc = 1,25 * NDVI + 0,2 (Ec. 13)
Kc = 1,5625 * NDVI - 0,05 (Ec. 14)
For this I would use the Landsat 8 sensor and the corresponding bands to obtain the NDVI.
Since in my area of study it is very wide and has different LULCs
I would like to know if it is okay
I think it’s a good idea. I recently also think that The KC values in the SWR model are just the weighted way of LUCC and could not describe the differences between the same landscape types. as we all know, in the same landscape types, there is a difference because of vegetation structure. for example, big trees with high NDVI values and small trees with low NDVI values have different Water yield services.
Hi @Nicolas -
At the moment, the SWY model only takes in Kc parameters through the biophysical table, which is mapped to each LULC type. We have been hearing increasing interest in being able to provide these parameters as rasters instead, but the models don’t currently support that.
You could use the NDVI process you described, then assign the average Kc value within each LULC class in the LULC raster. This may work ok if your LULC classes are specific enough to capture differences in different, far-away parts of the basin. If, however, you only have one, say “Forest” class in the whole LULC map, then taking an average won’t preserve any differentiation. Then, you could consider ways to divide more coarse classes, using an ecosystem map or climate zone map or something similar, to create different Forest classes that would then be assigned different Kc values.
In the LULC raster, I have 14 different types of coverages, renoval (juvenile forest), adult and young plantation (pine monoculture), stunted forest, scrubland, agricultural, bare agricultural, bare permanent, water, wetland, urban, native forest, among others.
I believe that with that amount of coverage I could have good results.