Investigating the effect of temporal and spatial changes of precipitation and temperature on vegetation index (Case study: Yazd province)

Vegetation cover is affected by the interplay of precipitation and temperature, as the amount and timing of precipitation, along with temperature patterns, directly influence plant growth, distribution, and overall health. The Normalized Difference Vegetation Index (NDVI) serves as an indicator of surface vegetation condition and effectively reflects the impacts of environmental changes. Precipitation and temperature are key controlling factors in NDVI variations. This study aims to examine the effects of precipitation and temperature changes on NDVI values derived from MODIS data in Yazd Province over a 24-year period (2000 to 2023).

This study utilized NDVI data extracted from the MODIS sensor aboard the TERRA satellite. The NDVI index with a spatial resolution of 250 meters was generated every 16 days. These images, which include vegetation indices, are used for global vegetation monitoring, land cover representation, and its changes. The analysis of NDVI changes was performed in Google Earth Engine (GEE), and spatial analysis of the images was conducted using ArcGIS 10.8.3. Due to the lack of long-term precipitation and temperature data and the absence of suitably distributed meteorological stations in Yazd Province during the study period, ERA5-Land reanalysis data for the 24-year period (2000 to 2023) were utilized. The spatial resolution of ERA5-Land is 9 km, and its temporal resolution is hourly. Data processing related to the calculation of annual, monthly, and seasonal averages of NDVI, precipitation, and temperature was carried out, followed by trend analysis using the Mann-Kendall test. The correlation between NDVI and climatic variables (precipitation and temperature) was examined at the annual, monthly, and seasonal scales.

The time series of NDVI, precipitation, and temperature at the annual scale for the 23-year study period showed an insignificant upward trend. Correlation analysis between NDVI and precipitation and temperature revealed that at the annual scale, the correlation of NDVI with precipitation was 0.31 (r = 0.31, p-value = 0.05) and with temperature was 0.15 (non-significant). At the seasonal scale, the highest correlations between NDVI and temperature occurred in autumn (r = 0.14), while the highest correlation with precipitation occurred in spring (r = 0.75). At the monthly scale, the strongest positive correlation with temperature was observed in July (r = 0.31) and the weakest in February (r = 0.08), while the strongest positive correlation with precipitation occurred in March (r = 0.38) and the weakest in September (r = 0.04). Based on the average monthly correlation coefficients, the effect of precipitation on vegetation changes was found to be stronger than that of temperature. A differential map of NDVI between the start (2000) and end (2023) years indicated that areas with increasing and decreasing vegetation changes were scattered irregularly and did not follow any specific spatial pattern. Vegetation changes in different areas were observed with considerable variation in magnitude and form.

The use of advanced remote sensing techniques, such as time series analysis of satellite images, the integration of multi-source data, and machine learning algorithms, has become a focus for improving NDVI estimation and other climatic variables, including temperature, precipitation, and evapotranspiration. The results indicate that both precipitation and temperature have a positive effect on NDVI, with precipitation having a greater impact. However, the intensity of this effect varies depending on the time scale. At the annual scale, NDVI showed the highest positive correlation with both precipitation and temperature. In spring, the highest correlation between NDVI and precipitation was observed, while in summer, the highest correlation occurred between NDVI and temperature. The findings of this study show that vegetation changes, in addition to the direct impacts of climatic variables, are also indirectly influenced by human activities such as land-use change, water resource management, and land exploitation in Yazd Province. These findings contribute to a deeper understanding of the interactions between vegetation cover, climatic variables, and human activities.