A 45-day gap limit changed monthly temperature slopes by up to 0.70 days per C in four US lilac samples in 1961-2004

We tested upper-gap exclusion in four US lilac region-period samples from 1961 to 2004. A 45-day limit removed 477 of 5,482 selected site-type-program-year records. It moved all eight March and April temperature slopes toward zero in size. The largest change was 0.70 days per C. Clipping instead of deletion gave smaller changes. A 60-day limit changed these slopes by at most 0.155 days per C; a 90-day limit removed no records. Models included unit and year effects and winter temperature. These are fixed-sample sensitivity results, not estimates of true observation bias or evidence that a cap is valid. All 45-day deletion intervals for March and April included zero. The data come from selected planted shrubs with no historical visit schedules. The package preserves source events, a weather-linked panel, source hashes, all estimates, code, independent numerical checks and a figure.

phenology · lilac · observation bias · outcome truncation · monthly temperature · sensitivity analysis