I know. Yet it goes in the books. Plus it is, at least historically, dominated by radiating nightime lows and the wind has precluded that. My suspicion is that if we had daily climo temp curves we'd find a greater area under the longterm mean than above it for Jan (2F bias aside perhaps). The max+min / 2 method is just really lame and it's so long since time to graduate from arithmetic to stats 101.