I started this project trying to replicate a graph I saw on Twitter last year, which showed how the expected first and late frost dates had both shifted by about 2 weeks over the last 80 years - making winter shorter.
The climate data comes from NOAA's NCEI/NCDC GHCN-Daily
I found Get_NOAA_GHCN data to be very helpful in converting the Daily format to tabular data.
I've always thought it was interesting that people have different ideas about which months are "winter" - if you're measuring the darkest months, it's Nov-Jan, with the winter solstice as "midwinter"; if you're measuring the coldest months, it's Dec-Feb. The lag between insolation (amount of sunlight) and temperature is called the seasonal lag, and I wanted to graph it out with local weather and sunrise/sunset data.
The temperature range boundaries are the min/max averages for that day of the year from 1939-2020.