Climate-adjusted time-series analysis of vegetation dynamics following pipeline disturbance and restoration: A proof-of-concept framework

Pipeline infrastructure causes abrupt vegetation disturbance, yet restoration effectiveness remains poorly quantified because climate variability and lagged vegetation responses confound observational studies. This proof-of-concept study developed a climate-adjusted interrupted time-series (ITS) framework to evaluate vegetation dynamics following pipeline clearing (November 2020) and replanting (March 2023). The analysis used 65 monthly observations per polygon (3 polygons, 0.42 ha; 195 polygon-month observations) at a study site in Italy. Monthly Sentinel-2 vegetation indices (NDVI, NDRE) and ERA5-Land climate variables (January 2020–May 2025) were analyzed using interrupted time-series models with fixed breakpoints, Fourier seasonality, standardized climate covariates, and polygon fixed effects. UAV multispectral surveys (May 2025, 2 cm resolution) and field measurements validated that fine-scale vegetation establishment was occurring below Sentinel-2's 10 m detection threshold. The framework successfully detected large disturbances: pipeline clearing caused immediate vegetation loss (NDVI -0.536, NDRE −0.368; both p < 0.001) followed by significant natural recovery during 2021–2022 (NDVI +0.074 yr−1, NDRE +0.049 yr−1; both p = 0.015). After accounting for climate and seasonality, replanting showed no statistically significant site-wide effect (NDVI +0.019, 95% CI: −0.064 to +0.103, p = 0.650). UAV validation confirmed localized vegetation establishment (NDVI r = 0.918, NDRE r = 0.841 with field survival), demonstrating that recovery was occurring at sub-pixel scales but remained below Sentinel-2's detection threshold due to spatial averaging. The climate-adjusted ITS framework provides a replicable, transferable approach for disentangling anthropogenic and environmental drivers where traditional control designs are infeasible. While detecting subtle restoration effects requires larger samples, extended monitoring, or higher-resolution imagery, this demonstrates a methodological advance for climate-aware evaluation of small-scale restoration interventions.