January-February warmth slopes for tree pollen peak dates were uncertain in four eastern US samples in 2003-2024

We tested winter warmth and tree pollen peak dates in 656 station-years at 52 selected eastern US stations. Separate models covered two regions and two periods in 2003-2024. Models controlled site, year and seasonal observation count. All four January-February temperature slopes were negative, from -4.74 to -0.54 days per C. All four corrected intervals included zero. Count adjustment changed slopes by less than 0.43 day per C. Higher observation counts, site trends and two-way county/year uncertainty also left all four peak intervals across zero. A January-March window gave intervals below zero in two samples, but overlaps some outcomes. These results do not establish a common winter warmth response or its absence. Mixed tree taxa, missing visits, selected stations, coarse weather and exposure timing limit inference. The package gives code, aggregate results, a figure and exact source hashes. External source records and merged tables remain outside the publication.

pollen · phenology · winter temperature · observation bias · regional replication · null result