They did. As they said, they use (weather) station data + statistics to compensate for missing data. We have temperature records from weather stations going bac

They did. As they said, they use (weather) station data + statistics to compensate for missing data. We have temperature records from weather stations going back to the 18th century. Just quickly googling I found this paper that works to make historical records usable for today: https://rmets.onlinelibrary.wiley.com/doi/full/10.1002/gdj3.11 https://rmets.onlinelibrary.wiley.com/doi/full/10.1002/gdj3.... "Daily observations of weather and climate for the province of Québec, Canada, start in the 18th century and continue to the present day. Daily temperature observations from 12 observers ranging from 1742 to 1873 are described here." That's why the 1880 forward data is different from other reconstructions going further back. These are measurement instruments by scientists recording temperature.