Aim: The southern part of Yazd Province is more exposed to desertification due to limited rainfall, rising temperatures, and human pressure. This study aims to analyze the spatio-temporal trends of desertification and its influencing climatic factors from 2000 to 2022 using remote sensing and climatic variables.
Material & Method: The study examined vegetation cover index (NDVI), Palmer Drought Severity Index (PDSI), precipitation, runoff, and minimum/maximum temperatures. Data were extracted from MODIS, TerraClimate, and WorldClim satellite products and processed using ArcGIS, R programming, and Google Earth Engine. The non-parametric Mann-Kendall test was applied to determine trends, and the Pearson correlation coefficient was calculated to assess relationships between NDVI and other variables.
Finding: Results indicated that NDVI showed a significant decreasing trend in over 85% of the study area (with -0.026 per month). This decline was strongly linked to reduced precipitation in 65% of pixels, decreased runoff in 60% of areas, and negative PDSI values in 70% of pixels. In low-lying and low-rainfall areas, NDVI decline was more severe (up to 90% of pixels). Minimum and maximum temperatures exhibited increasing trends in over 80% of the area, intensifying evapotranspiration and soil moisture loss. PDSI showed negative values in most years, indicating prolonged droughts. Time series statistical analysis confirmed significant decreases in NDVI.
Conclusion: The findings reveal that desertification in southern Yazd Province results from the synergy of reduced precipitation (up to 65% area with decreasing trends), runoff decline (60% pixels), temperature rise (80% area with increasing trends), and consecutive droughts. Continued trends exacerbate desertification risks and threaten ecosystem stability, with an average NDVI reduction rate potentially leading to over 30% vegetation loss in the next decade.
Innovation: The innovation lies in the integrated use of long-term satellite data, trend analysis, and correlation assessment of these parameters with NDVI.