Next-year spring temperature gave no clear link to current lilac bloom in four US samples in 1965-2008

We tested whether next-year March-April temperature predicted current lilac first bloom after current weather and plant and year effects were held fixed. Four US region and period models used 9,015 records from 1965 to 2008. Future slopes ranged from -0.182 to +0.571 days per C. All four adjusted intervals included zero. Current spring slopes remained negative. Planned sensitivity checks did not show a clear future signal. This result does not show that bias was absent. Current and future temperature had weak correlations after fixed effects, which limits the control. Unknown visits, selected long records, different lilac types by region, monthly grid weather, and few state clusters limit inference. We do not estimate causal warming effects, air pollution effects, or biodiversity loss. The package includes a raw-data rebuild, source hashes, the panel, all estimates, numerical checks, code, and a figure.

phenology · lilac · negative control · future exposure · observation bias · temperature · null result