hello - I’m new so please forgive my simplistic approach and no-scientific language!
I am using Invest to generate UHI scenarios where small forests are created in urban areas. I have used arcgis to merge the true LULC layer with my imagined forests with the appropriate LU code, and I’m happy with that. However the urban cooling model also takes evaoptranspiration as an input, so I have done a very simiilar thing here too. I reclassified my forests to a single notional ET value that seemed to fit with real forests in the area and then merged that with the true ET raster.
Cells in the original ET layer and my imagined forest layer are different sizes but I can see in the intermediate files that the ET is resampled during processing to match the LU. I hoping my distinct forest areas, (with higher ET values) will be absorbed into that resampling and affecting localised ET values, nudging the them higher.
I hope I explained that ok? What I’d like to know is if this is a valid approach?
I also find myself wondering to what extent LU and ET are co-correlated so why does invest need both inputs. I’m thinking ET must be affected by wind, aspect, surface material etc. and these subtlies must be important for the processing? For LU I am looking at sub 5m resolution whereas my ET layer is 30m, so that worries me.
Any thoughts from gurus very welcome,.
Thanks
