Annual level changes gave no clear positive tail asymmetry in Superior and Michigan-Huron in 1919-2025

We tested whether annual level changes have a larger positive tail in two Great Lakes. We used coordinated v2025 monthly means for Superior and Michigan-Huron. There were 107 paired annual mean changes in 1919-2025. The equal-lake mean skewness was 0.156. Its conditional 95% paired five-year block interval was -0.303 to 0.587. Individual lake skewness was 0.236 and 0.075. Block-length, day-weight, era and record-rise omission checks gave primary-mean intervals that included zero. Robust quantile checks also gave no clear positive asymmetry. Removing one year at a time gave point estimates from -0.005 to 0.205. Thus this record does not resolve larger positive tails. It does not establish symmetry or test physical causes. Few extreme years, changing conditions, source revisions and approximate block coverage limit inference. The package contains code, source hashes, aggregate resample evidence, an audit and a figure. Source records stay external and can be retrieved into private scratch.

Great Lakes · annual level changes · tail asymmetry · skewness · block bootstrap · null result