These R scripts and datasets (CSV-files) were used to write the article 'Does drought advance the onset of autumn leaf senescence in temperate deciduous forests' published in the journal Biogeosciences (see https://doi.org/10.5194/bg-2020-337). Please cite the corresponding article from Mariën, B., Dox, I., De Boeck, H., Willems, P., Leys, S., Papadimitriou, D. and Campioli, M. 'Does heat stress and increased atmospheric aridity advance the onset of autumn leaf senescence in temperate deciduous forests' appropriately when using the R scripts or datasets (see https://doi.org/10.5194/bg-2020-337). The R scripts GAM_Nsite, GAM_Betula, GAM_Fagus and GAM_Quercus were used to run the generalized additive models (models 1 – 4) corresponding to the chlorophyll content index and canopy greenness, and to plot their respective smoothers and line plots (Fig. 4 – 7). The corresponding CSV-files are named CCI_Nsite_saplings, Coloration_Nsite_saplings, CCI_Leaf_Fall_mature_trees and Coloration_Leaf_Fall_mature_trees, respectively. The R script Meteo_plot returns the code required to plot the meteorological data from the CSV-files Meteo_Daily_averages_2017_2020 and Brasschaat_SWC (Fig. 2). The R script Plots_Nsite returns the code for plotting the relative humidity, temperature, soil water content and vapor pressure deficit in the glasshouses in Antwerp (Fig. 1). The corresponding CSV-files are named RH_Nsite_2018, Temp_Nsite_2018, SWC_Nsite_2018 and VPD_Nsite_2018, respectively. Plotting the breakpoint data and analyzing the results from the linear mixed models (model 5 – 6) was done using the R script Graphs_and_analyses_breakpoints and the CSV-files named Breakpoints_saplings, Breakpoints_mature_trees_CCI and Breakpoints_mature_trees_Col, respectively.