Hi all, I’m working with some temperature data, and wanted to see if anyone has recommendations for how to better deal with my situation. I am comparing a rare historical data set with only single mean/max/min values for each month to full records taken continuously over the past 10 years. Assume the data are directly comparable . We basically want to see if the most recent 10 years differ from the one historical year, with the hypothesis that the recent years would be warmer in most months .

Obviously having a single year’s worth of data isn’t ideal, but it is valuable for the system I’m studying as there are no comparable records elsewhere.

So far I have calculated the same metrics for the recent data , modeled these and put a 95% confidence interval around it, and visually looked at which historical data points are outside that interval. From this it appears some months are warmer in the recent years than in the historical year, but I can’t really say much beyond qualitative statements.

I’m not well versed in Bayes models, but is there a better way to compare these data and get some actual quantitative results? Or should I just stick to a more quantitative approach given the data’s limitations?