What statistical test can help me resolve this correlation?

.

My aim is to define a region based on the proportion of variances (within the yearly classification) over which it will be un-necessary to go beyond a particular correlation, as the results are sufficiently similar.