Winter warmth had no clear link to ten-day cherry bloom intervals in four Japan samples from 1953 to 2021

We tested whether warmer winters were linked to a higher chance of at least ten days from first to full cherry bloom. We reused 4,436 paired JMA station-years at 71 stations in northern and southern Japan in two periods. Models controlled station effects, year effects and spring temperature. Winter slopes ranged from -4.43 to +6.87 percentage points per degree C. All four adjusted intervals included zero. Other interval thresholds, station trends, equal station weights and stable observation settings also gave uncertain links. These results do not establish a common increase or a zero effect. Few long intervals in the North, selected complete pairs, unknown visits and tree changes limit inference. Seasonal temperature does not measure daily chilling. This is a new tail test on a published panel, not an independent data replication. The package includes the panel, source hashes, code, all results, date checks, numerical checks and a figure.

phenology · cherry · winter temperature · tail risk · observation bias · JMA