According to the MEDALUS model, 36 percent of the study area now falls into the severe to very severe desertification classes. Two factors contributed most to this deterioration: the quality of management and policy, and erosion. These results identify where human decisions and natural processes interact to accelerate land degradation in a semi-arid environment.

To detect when the system shifted, the researchers applied breakpoint analysis using the BFAST software package, which scans long time series for abrupt changes in trend. The main breaking point emerged in 2015, coinciding with a sharp decline in vegetation cover, a drop in groundwater levels, and an increase in soil salinity. Beyond detecting the shift itself, the team examined whether statistical properties of the satellite time series could serve as early warnings.

They evaluated lag-1 autocorrelation, skewness, and standard deviation in R, testing whether these signals changed measurably before or during the transition. The analysis showed that NDVI and NDSI, combined with lag-1 autocorrelation, were the most effective tools for the early detection of ecosystem changes. Spatial analysis added a geographic dimension to the findings.

The northeastern, southeastern, and central parts of the study area emerged as the main desertification hotspots, characterized by significant changes in vegetation cover, soil properties, wind erosion, and groundwater depletion. The authors emphasize that combining remote sensing indicators with statistical methods is essential for building early warning systems and supporting sustainable natural resource management to control desertification. Desertification remains one of the major challenges facing drylands worldwide, with profound impacts on ecological sustainability, agriculture, and community livelihoods.

Subject of Research: Early warning signals of desertification detected with remote sensing time series in northeastern Iran

Article Title: Analysis of early warning signals of desertification hazard using time series of remotely sensed indicators in northeastern Iran

Article References: Boali, A., Asgari, H. R., Behbahani, A. M., Salmanmahiny, A., & Naimi, B. (2026). Analysis of early warning signals of desertification hazard using time series of remotely sensed indicators in northeastern Iran. Earth Science Informatics, 19(12), Article 230. https://doi.org/10.1007/s12145-026-02284-7

Image Credits: AI Generated

DOI: 10.1007/s12145-026-02284-7

Keywords: desertification, remote sensing, early warning signals, NDVI, soil salinity, groundwater, MEDALUS model, Landsat, northeastern Iran, breakpoint analysis, land degradation, drylands