Working with pixel size larger than 30m

[JestradaG]

Hi, I wonder if there is any option to work with pixels of much larger dimension?

I want to use the output of your simulation as an input for an optimization model. I need data of an area on about 61km by size (square area) that is going to be composed of 600 grid cells. I wish to have each cell (2.4km x 2.4km) to be directly obtained form your simulation model.

Is there anyway to have aggregated data for larger areas to be simulated without a requirement of such granularity?

[doyled-it]

Apologies for the late reply, but this would be a feature that we would require you to make into a PR and we can review and merge in.

The size of the data is currently directly dependent on the LANDFIRE API, and that only goes up to 90m resolution at its coarsest. If you wanted to make cells of much larger size, this would have to be interpolated after getting the data from LANDFIRE.

This is doable, but we don't currently have the development capacity to put towards this capability.

If you wanted to include this functionality, I would recommend putting a scaling factor in the config that is then propagated to an internal method (say named _upscale_area) of the LandFireLatLongBox class and put into the _make_data method.

Any further specifics could be discussed in a PR.

[doyled-it]

Update, it looks like landfire-python actually does support resampling, so this might be easier than anticipated. Still not on our radar for now, but definitely makes this a lighter load:

https://landfire-python.firesci.io/en/latest/usage.html#resampling-data