First-leaf updates gave no clear first-bloom forecast gain in four US shrub samples in 1980-2008

We tested a first-bloom date update from the first-leaf date of the same shrub. Each test used only earlier paired years for that plant. The baseline was the past mean bloom date, bounded below by current first leaf. Four East and West samples in 1980-2008 used 4,815 pairs. Mean error reductions ranged from -1.45 to 0.54 days. All four adjusted intervals included zero. Update errors ranged from 6.33 to 9.42 days. Missing visits and report times can link recorded dates and reduce useful lead. These are past-only archive tests, not live forecasts or causal climate effects. The package has a fixed data panel, all predictions, source hashes, code, checks and a figure.

phenology · first leaf · first bloom · forecast · warm springs · observation bias · lilac · honeysuckle