Can anybody share the Kc values they used in their Biophysical table for urban land use categories?
I would love to see someone else’s for comparison and hear how you calculated them, because the Kc Calculator tool seems overly complicated for Urban systems, but perhaps I’m misunderstanding how to use it.
Hi @nic.keller,
Thanks for posting to the forums. I know Kc values is a common topic among our models, unfortunately I don’t have much background in the area. But here is another thread on the forums that might help:
Let us know if that post is useful!
Cheers,
Doug
After thinking more deeply I think we have a more detailed question:
So one of the main input layers is the “Reference Evapotranspiration” for the study period. Can someone explain exactly what this means? Does it expect actual ET or potential ET? In our case, we acquired actual ET from the ECOSTRESS sensor on the ISS for our study area.
Calculation of the Kc of each land use category in the biophysical table seems to rely on PET (potential), if we are understanding Allen et al. correctly. That also suggests that the ET raster layer was supposed to be potential ET. Since our raster is actual ET, we are considering using a Kc value of 1 for all land use categories in the biophysical table. Is that valid? Is that doing what we think it’s doing - i.e. basically telling the model to accept the actual ET value from our ECOSTRESS raster as Kc?
Hi @nic.keller
Does it expect actual ET or potential ET?
Yes, I think that Potential Evapotranspiration is the required input. From what I understand Reference Evapotranspiration (ET_0) is Potential Evapotranspiration calculated at a nearby climatic station on a reference surface. ( Full disclosure, pulling from Wikipedia ).
Since our raster is actual ET, we are considering using a Kc value of 1 for all land use categories in the biophysical table. Is that valid?
Hmmm. Good question. The model creates and uses an evapotranspiration index which represents a normalized value of potential evapotranspiration (From the Users Guide - UCM)
I’m not comfortable saying that equation 1 in the Users Guide would actually reflect passing in AET with Kc values of 1…
Hopefully this is somewhat helpful. I’m not the model designer, but can reach out if you’d like more clarity.
Cheers,
Doug
Hi Nic,
I had the same doubt! I also have the actual ET raster data for my study site. I was about to use it with Kc=1 and then run a correlation analysis to see if there is any difference with the results using the Potential ET. Unfortunately the actual ET layer I have has a coarse resolution 1km2…so I am not sure.
Please keep me posted!
I also found a very useful the information from Hamel et al 2000. (Co-développement du module rafraîchissement offert par la végétation de l’outil InVEST) it is in French but a very useful report with a detailed tables with all the parameters for many different LULC.
Best,
Laura
Hi Doug, yes if you could reach out to the developers about this question, that would be much appreciated. We’re going to move forward with using 1 for our Kc values in the meantime, but I would feel much better with some kind of confirmation that this is valid. Thank you!
Hi @nic.keller,
Sorry for the really slow follow up! I asked around a bit but didn’t get a ton of feedback from the model developers.
I think some takeaways are that technically you should probably be providing reference ET. Potential ET is similar, but different, and the model wasn’t designed with that in mind. However, initial thought is that it would be ok to use actual ET and set Kc to 1, since it is creating an index of evapotranspiration (not an absolute value that it’s more important to get the actual value correct) that includes vegetation type, and that’s akin to Actual ET.
Perhaps @Perrine, @RafaSchmitt, @chris or others could pitch in?
Doug
I think I’ll use the same “trick”, if anything has changed in the meantime, please tell me!
Thanks,
Antonio
Hi everyone! I’m thinking about using actual evapotranspiration as well and set Kc values in the Biophysical Table to 1. However, the mentioned datasets for actual evapotranspiration, i.e. the ECOSTRESS ISS Sensor gives values in W/m2 and if I understand correctly the model requires mm/timeunit - how did you convert or calculate W/m2 to the one needed for the InVEST Model? any kind of help on this is very much appreciated!
Well, I’m no expert on this, but this stackexchange post suggests a method for converting from latent heat flux to actual evapotranspiration. This is assuming that you’re using the ECOSTRESS ETdaily product, as referenced by this USGS page, which has units of W/m2.