December, like all of 2024, had lower search interest than the last few years. This was especially true for the Upper Peninsula. December was warmer than forecasted, but Northern Michigan was snow covered for most of the month. This winter is forecasted to have relatively normal temperatures and normal or higher precipitation. If true, this will be considerably different from 2024 which was a very warm and snowless winter. The Upper Peninsula, especially, is sensitive to winter sports such as snowmobiling, ice fishing, and skiing.

I am currently two years into this project and one year into the current forecast methodology. The next three months will be a period of truth for the current methodology. I am planning some changes this year; however, how large these changes are will depend on these results. The first methodology was based on difference from previous months. This led to very inconsistent forecasts from month to month. The current methodology is based on difference from the previous year. The benefit of this method is it creates much more consistent forecasts from month to month and reflects my previous observation that variation in search interest in the previous season doesn’t explain search interest in the next season. However, this is not the case in a year like 2024 where overall search interest appears to have trended downward for the entire year.

The deviation from past trends has been larger for the Upper Peninsula. At the same time, the potential positive impact from a normal winter is highest for the Upper Peninsula. How I change the future forecast methodology will likely depend on to what extent the positive impact of a normal winter ends up being true as opposed to the observed downward trend in search interest.