Closed davemfish closed 8 months ago
Chris has the building footprint data on hand, so will share with us. And he will add building_type
to the lucode parameter table.
Footprint data is available here: https://github.com/microsoft/GlobalMLBuildingFootprints
So I can work on integrating that. And Chris will work on the parameters.
I have the building footprint data now and a script that extracted it.
I'm closing this in favor of #89, which can include the work to incorporate the building data.
Without building footprint data, the current valuation method has a lot of uncertainty around energy costs/savings associated with building heating & cooling because it relates the presence of buildings on a pixel to the LULC type of that pixel. This might be okay for types like "Dense Urban" or "Natural Park", but for types like "Commercial, Industrial, or Transportation" and many others, there could be a wide variety of building densities occurring on land of those types.
Also, right now we don't yet have energy cost parameters that vary by LULC code. We have one-size-fits-all energy parameters. If we don't want to incorporate building footprint data, should we at least parameterize the energy costs table for different LULC types? Or associate LULC to a "building type" label?