Winter warmth gave no clear change in March slopes of cherry bloom intervals in Japan in 1961-2024

We tested whether winter warmth changes the link between March temperature and the time from first cherry flowers to full bloom. We used 2,924 paired years at 46 Japanese stations in 1961-2024. Four models used separate northern and southern stations and two time periods. They controlled station and observation setting, year, and curved temperature responses. Interaction estimates were -0.307, +0.165, -1.046 and -0.278 days per C squared. All four intervals included zero after correction for four tests. Station trends, on-site records, equal station weights, omission of Kagoshima, and removal of curved terms did not show a common positive interaction. A conservative variance rule handled one negative secondary two-way variance estimate. These results do not establish no interaction or a chilling mechanism. Complete-pair selection, unknown visits and tree changes, coarse monthly weather, and fixed March windows limit inference. The package gives code, raw source hashes, a matched panel, a raw rebuild check, separate coefficient checks, all estimates and a figure.

phenology · cherry · winter chilling · temperature interaction · observation bias