Dear all,
I was doing some more general reflections on the modeling of this service, in particular with regard to the well known disitinction between supply, demand and budget of ecosystem services.
In the case of this service, do you think it is correct to define:
Heat Mitigation index as supply of the service;
Average temeperature anomaly as demand for the service.
If the above assumptions are true, what could be the budget of the service?
Thanks for your attention,
Antonio
I continue my reflections, with the hope that someone will join the debate 
I was currently wondering: if I wanted to know the service supply by land use class, I could calculate the statistics on LULC of both HMi and CC.
I think maybe the result of CC by LULC class is the most appropriate, since it doesn’t consider the effect of surrounding green areas, but only the characteristics of that pixel (proxi of LULC class).
Hi @antobaro sorry for the delay! We’ve reached out to our science teams and are waiting to hear back.
James
Hi Antonio!
Thanks for reaching out. I like the approach of considering supply and demand for our urban services, but with all these things it gets pretty complicated pretty fast.
The supply of the service is, as you state, the Heat Mitigation index.
The maximum demand (without taking into account where people exist on the landscape) could be the UHI magnitude, which is not spatial. You could argue that spatialized demand is the modeled air temperature (found in the intermediate results folder as T_air{suffix}.tif), with high temperatures indicating highest demand–this is the temperature anomaly, but calculated as an absolute temperature. However, this incorporates the HMi into its calculation so we may be double counting when comparing the two. And as I hinted at earlier, these metrics have no sense of where people are actually experiencing demand for cooling (e.g dense residential areas). I would try to calculate demand as the air temperature weighted by population.
I have not conceptualized this model in terms of a ‘budget’ before, and to be frank I have a hard time seeing an easy way to do so. Typically, my use of this model has been in calculating the difference between two land use scenarios, so any discussions of value are about the total change rather than the service of a single baseline scenario.
Your approach for supply by LULC class seems appropriate, as CC is much more the impact of the land cover itself and not its spatial context.
Hope this helps!
Thank you @chris for your response!
I agree with everything and I too had the doubt about the double counting:
I’m still thinking about it and if I have any enlightenment :), I’ll post here.
In the meantime, thank you!
.