Mean gaps between first cherry flowering dates showed no clear decrease at 47 Japanese stations from 1961–1990 to 1996–2025

We tested whether first cherry flowering dates became closer across 47 Japanese weather stations. Each station has a regular-species record in every year from 1961 through 2025. The sample has 3,055 station-years. For each year, we calculated the mean absolute date gap across all 1,081 station pairs. The mean gap was 11.42 days in 1961–1990 and 11.09 days in 1996–2025. The change was −0.33 days, with a 95% interval of −1.54 to +0.91 days from 10,000 circular five-year block resamples of whole national years. The result does not establish a decrease or stability within one day. Mean flowering advanced by 5.88 days. The mean annual date–latitude correlation fell from 0.945 to 0.914, with a change interval of −0.050 to −0.013. This agrees in direction with prior work on the latitude link. A weaker latitude correlation does not itself establish closer dates. Checks that removed Hokkaido or Kagoshima, kept unchanged source remarks, or used full bloom also gave gap-change intervals that include zero. An independent raw-data calculation matched all 65 annual gaps. A clean rebuild matched 16 tables and panels. The package preserves raw inputs, provenance, code, draws, and figures. The intervals are conditional on the station network and block assumptions. Reference-tree and observation changes remain possible. The study does not identify a temperature cause or a response of wild populations.

phenology · cherry flowering · Japan · spatial date spread · fixed network · block bootstrap · reproducibility