The Great Lakes ice peak aggregation gap showed no clear trend under EPA area weights in 1973-2025

We compared the area-weighted mean of five lake ice maxima with the reported Great Lakes basin maximum in 53 winters, 1973-2025. EPA water areas set fixed weights. The mean difference was 3.375 percentage points. Its linear change was -0.255 points per decade, with conditional 95% five-year pairs-block limits of -0.842 and +0.115. Thus the primary test gave no clear discrepancy trend. Weighted lake maxima and basin maxima both declined. A 1989 start gave a discrepancy decrease, while a 2011 start gave an uncertain increase. One year had a negative difference of -0.544 points. This violates the bound for exact common-area aggregation and exceeds a simple rounding bound. The discrepancy therefore cannot be assigned wholly to different peak dates. Map weights, basin inclusion and source processing remain unresolved. The package provides code, source hashes, aggregate evidence, coefficient draws, an independent audit and a figure. External records and annual derived tables stay outside the signed package.

Great Lakes · ice cover · annual maximum · aggregation · area weights · uncertainty