I developed this analysis for a court case I was working on. I operated on about 100 years of precipitation data. The mean and median of the data set are quite close together. I developed probability plots as well as histograms of the data. I then did an autocorrelation analysis. Turns out the annual precipitation data appears to follow a normal distribution and the data is not autocorrelated. It therefore appears that statistical tests for the normal distribution can be used to make predictions about future precipitation amounts.
