The purpose of this installment is to determine whether the search interest data introduced during the first two installments and the weather data introduced during the third installment are correlated. The general idea is that people will search for outdoor recreation search terms differently in some weather conditions than others.

Note there are 10 search interest keywords: atving, boating, camping, fishing, hiking, kayaking, rving, hunting, skiing, and snowmobiling. There are four weather variables: maximum temperature, minimum temperature, precipitation, and snow depth

One of the greatest drivers of search interest for outdoor recreation is seasonality, largely due to weather. However, it’s generally to a smaller degree that unseasonal weather patterns impact unseasonal search interest outside of recurring seasonality. Finally, changes in search interest from the previous day are generally not explained by changes in weather from the previous day.

While the general ideas concerning the relationship between weather and search interest are described above, there are obviously ways to improve upon this framework. For instance, in many cases these impacts are seasonal in themselves. Changes in temperature during the middle of winter are unlikely to impact search interest for boating. Changes in temperature are likely to have the opposite effect for summer fisherman vs ice fisherman. Aggregating the data for the entire year dilutes the results we see.