Doubts about urban cooling model

Hello, Community and everyone!
I am thinking about the input dataset for the urban cooling model.
The reference evapotranspiration dataset is required according to the user guide. The calculation of the cooling capacity index and urban heat mitigation index are based on evapotranspiration.
But in my case, my research area only focuses on the urban area (impervious surfaces), not covers the rural area or non-urban area. For the evapotranspiration products (e.g., MODIS), I think the values in the urban area always be filled(or zero). I hope to know, in this case, the urban cooling model is available or not?is there a significant influence on estimating urban cooling services if only assessing urban areas (almost all evapotranspiration values are zero)? If not available, any solution or alternative methods?

Thank you very much.
Best regards

Hi @Icey,
Urban areas commonly contain green spaces, which do have non-zero evapotranspiration. The Urban Cooling model relies on the evapotranspiration data to determine the cooling effect provided by green areas. You should make sure that your evapotranspiration raster captures even small green spaces. Global data products like MODIS may be too low-resolution for this.

If your area of interest really contains no green space at all, the Urban Cooling model may not be very helpful for you.


Hi @Icey ,

ET rasters are coarse as these values do not change much over short distances, so your experience is the same as is that of other users. Reference ET values will not be any different in a city versus outside of it. Remember, this is reference ET, not actual ET.

Reference ET refers to the evaporative power of the atmosphere, assuming water is not scarce. Similar to the baseline rural reference temperature model input, users should consider rural reference ET values and rasters applicable to adjacent cities (the sun radiates with the same intensity downtown as it does in the suburbs and beyond). The Urban Cooling model adjusts for the city based on the other model inputs (namely the land cover map and associated biophysical table). So to use this model, your research would have to consider some “rural” (city-adjacent) environmental factors in order to establish a baseline to compare the city against (just like experiments need a control).



Thanks for correcting me on that @jesseG!

1 Like

Hi @esoth and @jesseG,

Thank you very much for your detailed and clear explanation. It’s clear to me now. I’m comparing the several urban clusters’ UHI intensities. My areas of interest are large, covering numerous megacities. So, I am using coarse spatial resolution data(MODIS, CCI) and only focusing on the urban area. I will consider enlarging the study area, including more rural areas. Thanks!


1 Like