Sibylles on the forums (https://community.naturalcapitalproject.org/t/pollination-python-without-pip-install-gdal-3-3-0/5479/4) identified an issue with the InVEST Pollination model where `farm_pollinators.tif` has `inf` values as nodata, and functionally infinity values in places where there should be reasonable numerical values. I am observing this on `natcap/invest`, version `3.14.2.post60+gaf8ce8be3`, having just used `make fetch` to get the latest sample data. To reproduce: * Run the InVEST pollination model using the sample data * Open `farm_pollinators.tif` using QGIS to observe values.