Code for the paper
Nwachukwu, P. N., Berti-Equille, L. How robust are satellite-detected links between forest greenness and climate variability? A sensitivity analysis of MODIS kNDVI in intact forests. Under review.
The paper relates monthly MODIS kNDVI anomalies (2000–2024) in 14 intact-forest zones to four climate modes (AO, ENSO, NAO, PNA). It uses rolling correlations, matrix-profile recurrence, motif and changepoint co-occurrence, critical-slowing-down indicators, a 0.5° grid analysis, a pixel bootstrap and an AVHRR comparison. Each result is tested under two deseasonalisations, with autocorrelation- corrected trend tests and false-discovery control.
forest_greenness_climate_robustness/
experiments/ analysis scripts (Python 3.12, Hydra)
conf/ configuration
data/raw/ CPC climate indices (2000–2025 and 1982–2024)
data/processed/ regional monthly kNDVI series
README.md setup, data sources and run order
Setup, data sources and the full run order are in
forest_greenness_climate_robustness/README.md.
Everything runs on a CPU; no GPU or cloud resources are needed.
The analysis outputs used in the paper and the 5 km pixel-level kNDVI series (1.6 GB) are on Zenodo: https://doi.org/10.5281/zenodo.20039952
CC BY 4.0.
European Horizon project eco2adapt (grant agreement 101059498).
Laure Berti-Equille, IRD, ESPACE-DEV, Montpellier, France (laure.berti@ird.fr)