I think the most interesting scale on that weather page is the red-blue one where the extremes of both then veer into other colors - often the same colors, but that's OK because they will only appear within 'moats' of red or blue covering the typical temperature distributions. So you have increasingly deep shades of blue or red to indicate very cold or hot weather, and then purple to white within those areas for record-setting-breaking extremes.
I guess you could just do normal vs deviant scales, but given our sensitivity to the dichotomies of weather people have an inherent need to know what sort of extreme it is.
Yes I noticed that too, and I agree with your guess about why it’s ok. Fundamentally the functions here are smoothly varying, which avoids ambiguities and some some contrast effects that affect chloropleth maps. Long path length in colorspace == more opportunities for nearby discrimination, even if with overlapping hues
I think the most interesting scale on that weather page is the red-blue one where the extremes of both then veer into other colors - often the same colors, but that's OK because they will only appear within 'moats' of red or blue covering the typical temperature distributions. So you have increasingly deep shades of blue or red to indicate very cold or hot weather, and then purple to white within those areas for record-setting-breaking extremes.
I guess you could just do normal vs deviant scales, but given our sensitivity to the dichotomies of weather people have an inherent need to know what sort of extreme it is.