A 13-month drift correction left no clear seasonal phase-agreement gap between two Great Lakes in 1948-2024

We tested whether Michigan-Huron had less agreement in seasonal peak timing than Superior after annual drift removal. We used coordinated monthly levels, version v2025, and paired annual harmonics for 1948-2024. The raw amplitude-weighted phase-agreement difference was -0.0853, with a conditional 95% five-year block interval of -0.1336 to -0.0386. A slope from months just outside each year, 13 months apart and corrected for the mean seasonal step, left a difference of -0.0001, with an interval of -0.0138 to +0.0133. The change in gap was +0.0852, with an interval of +0.0433 to +0.1281. A smoother 24-month correction retained a negative gap of -0.0439. Thus the cross-lake comparison depends on the drift method. External endpoints remain related to the annual observations and can remove genuine seasonal variation. These tests do not identify the true seasonal timing, a physical cause or forecast skill. This extends a prior campaign observation with a paired test and external drift estimates. Code, all numeric inputs, source hashes, annual fits, uncertainty draws, an audit and a figure are included.

Great Lakes · seasonal phase · annual drift · harmonic analysis · method sensitivity · water levels