Adjusted warmth by rain-deficit intervals crossed zero for autumn timing in two northern US camera samples in 2008-2023

We tested whether low summer rain makes the link between summer warmth and autumn canopy timing more negative. Four eastern US region and period tests were planned. Two northern models used 129 camera-years in 2008-2015 and 2016-2023. Their interaction estimates were +1.27 and -2.93 days per degree C per local rain standard deviation. Both four-test adjusted intervals included zero. Year-sign tests gave p = 0.391 and 0.156. The two southern samples had only five and seven usable cameras, below the planned minimum of eight, and were not fit. Most model checks kept a negative later northern estimate but did not establish a repeated result. Changing the spatial groups and comparing current with next-year weather gave isolated adjusted intervals below zero in later North. These are associations in optical canopy timing, not causal climate effects or direct measures of leaf senescence. Few years, changing camera membership, selected single-cycle seasons and remaining observation bias limit inference. Pollution and daily extremes were not tested. The package includes code, source hashes, aggregate evidence, figures, reusable NOAA weather statistics and independent numerical checks. PhenoCam source records stay external.