Daniel Wilson of the Fed of San Fran has written a high quality panel paper available here. His unit of analysis is a United States county/industry/year/month and he is studying how employment in these units of observation are affected by the county/year/month's snowfall, rain and temperature as well as the weather in the previous 3 months. Using data from 1975 to 2015, he explores how weather affects economic activity. The novel feature of his analysis is that he can include many, many fixed effects. His “identifying variation” is based on the fact that the same industry in the same place is exposed to different weather over time and that the same industry in the same year/month is exposed to different weather at a point in time (because industries are scattered across the country).
On the topic of adaptation, Dr. Wilson finds that we have adapted to snowfall but not to spring hot temperatures. This is an interesting reduced form fact that merits more research.