Assessing rangeland rehydration through remotely sensed vegetation response

Detecting ecosystem responses to rehydration in extensive grazing landscapes is challenging due to their large spatial scale and seasonal variability. We evaluated whether high-resolution Sentinel-2 multispectral imagery, combined with factorial statistical analysis, can support adaptive monitoring of rehydration interventions in arid and semi-arid pastoral systems of northern Australia. Using a replicated before–after, control–impact (BACI) design, we monitored three 75-ha treatment plots modified with water-spreading banks designed to slow overland flow and enhance infiltration, alongside three matched control plots located ~1.6 km away. From monthly time series, we derived total standing dry matter (TSDM), green and non-green biomass, and surface moisture/soil exposure indices—including the Normalised Difference Moisture Index (NDMI), Normalised Difference Water Index (NDWI), and Normalised Bare Soil Index (NBSI). Analyses focused on three periods: late dry season pre-construction (August–October 2022), wet season (November 2022–April 2023), and construction phase (May–July 2023). Two-way factorial ANOVA (treatment × period) revealed significant interactions, with the strongest responses observed in treatment plots adjacent to the constructed banks. Notably, green biomass and NDMI increased during and immediately after construction, indicating that landscape-scale redistribution and retention of surface water can enhance pasture productivity. This integrated remote sensing and statistical framework offers a scalable, cost-effective approach for evaluating rangeland restoration impacts in data-scarce, climate-vulnerable landscapes.