Data and R scripts for: Climate-sensitive periods of tree growth advance under cold, delay under heat, and widen under drought

This Zenodo deposit contains the R scripts and processed datasets needed to reproduce the figures and supplementary tables of the manuscript "Climate-sensitive periods of tree growth advance under cold, delay under heat, and widen under drought." Thirteen scripts cover the main mixed-effects models, the figures, and the nine sensitivity analyses added during revision. The deposit starts from processed analysis tables rather than from raw tree-ring and climate series. Extracting the climate-sensitive periods from daily ERA5-Land data takes roughly three months of computation for about 3,000 sites, so that step is documented in the README rather than bundled here. The raw correlation surfaces and the scripts that reduce them are available from the corresponding author on request. Processed inputs include the analysis table (modelling_data_r1.rds), its counterparts built from SPEI and vapour-pressure deficit, the window coordinates under all seven extraction rules, and site-level metadata for the 2,926 chronologies in the analysed network (meta_data.csv). A complete output/ folder from the run behind the manuscript is included, so every figure and table can be inspected without running anything. Because Zenodo stores files as a flat list, the two directories are deposited as data.zip and output.zip; extracting both at the deposit root, alongside the scripts, rebuilds the layout the scripts expect. The README describes how to run each script, which manuscript item it produces, and how the derived data files were built. Supplementary Data 1 reports the per-group trend estimates and diagnostics obtained under each of the seven window-extraction rules; Supplementary Tables 4 and 5 are drawn from it and are also provided as separate files. Supplementary Data 2 and 3 provide the full model outputs for the analyses of CSP shifts. Supplementary Data 2 reports models fitted at the climate zone × clade level, Supplementary Data 3 those fitted at the genus × climate zone level. Each includes complete fixed-effect estimates (intercept and year terms), chronology-level random-intercept estimates, variance components for the random intercept and residual, and key model fit and diagnostic metrics: log-likelihood, AIC, BIC, residual σ, AR(1) ρ, convergence information, and summary residual diagnostics. Relative to the previous version of this record, the only change is the numbering of four supplementary figures: what were Supplementary Figs. 7, 8, 9 and 10 are now 8, 9, 10 and 7 respectively. The code, data and results are otherwise identical.