Introduction

Vegetation cover and soil moisture are critical for ecosystem health, regulating water cycles and energy balance. Climate change rising temperatures, altered precipitation, and increased evapotranspiration directly impacts these variables, while non-climatic factors (lithology, topography, land use) modify responses. Remote sensing indices (NDVI, NDMI, and SMI) enable effective spatiotemporal monitoring. The Zakin-Mashari Watershed (Hormozgan, Iran) represents a semi-arid Mediterranean ecosystem with sparse vegetation and significant climatic variability. Despite its importance, no integrated study has simultaneously examined historical trends and future projections under climate scenarios alongside lithological and topographic controls. This study analyzes NDVI and soil moisture changes during 1994–2024, projects to 2055 under SSP2-4.5, quantifies relationships with precipitation and non-climatic factors, identifies dominant controls and decoupling mechanisms, and provides management recommendations.

Method

Study Area

The watershed covers 771.10 km², 64 km north of Bandar Abbas, bounded by 27°41′44″–27°55′50″ N and 56°8′35″–56°41′34″ E. Elevation ranges from 404 m to 3,245 m. Climate is Mediterranean with 253.71 mm annual precipitation, 28.87°C mean temperature, and 1.38 m/s wind speed. Geologically, the watershed lies within the Zagros Fold-Thrust Belt with limestone, dolomite, marl, shale, sandstone, conglomerate, and salt domes. Vegetation is predominantly shrubby, including Prosopis cineraria, Ziziphus spina-christi, Tamarix spp., and various Artemisia species.

Data Sources

Precipitation (1990–2020, 2021–2024 via SARIMA) from four gauges. Lithology from 1:100,000 maps refined to 1:25,000. Landsat 5,7,8 (1994–2024, January). Four GCMs (ACCESS-ESM1_5, CESM2, MIROC6, MPI-ESM1_2-LR) under SSP2-4.5. 30-m DEM for slope (8 classes) and elevation.

Data Processing

Landsat Level-1 geometric correction (GCPs, RMSE<0.5 pixels) and FLAASH atmospheric correction in ENVI 5.3. GCM data converted from kg/m²/s to mm/year. Scale Factor bias correction based on 1990–2024.

Indices (Rouse et al., 1974; Wilson & Sader, 2002; Carlson et al., 1994; USGS, 2023):

(1) NDVI=((NIR-Red))⁄((NIR+Red) )

(2) NDMI=((NIR-SWIR))⁄((NIR+SWIR))

(3) SMI=(LST-〖LST〗_min)⁄(〖LST〗_max-〖LST〗_min )

(4) LST=K_2/(〖ln(〗⁡〖K_1/L_λ 〗+1))-273.15

Analysis: Pearson correlation (Zar, 2010), linear regression (Draper & Smith, 2014), ARIMA (Tunnicliffe Wilson, 2016), hybrid Regression+ARIMA. Spatial: Zonal Statistics in ArcGIS 10.8 (ESRI, 2020), heatmaps in R/ggplot2 (Wickham, 2016).

Results

Current Status (2024)

NDVI ranged from -0.224 to 0.421, with highest values in northern/western high elevations and lowest in central/southern lowlands and salt domes. NDMI ranged from -0.210 to 0.308, with maximum moisture in northeastern marl-shale areas and minimum in limestone-dolomite regions. SMI ranged from -0.761 to 0.174.

Historical Trends (1994–2024)

Precipitation declined significantly (annual slope ≈ -3,215 mm.km²), from ~400,000 to <100,000 mm.km². NDVI_min improved from -0.30 to -0.10, while NDVI_max declined from 0.52 to 0.32; NDVI_mean remained stable (~0.10). NDMI_min improved from -0.485 to -0.182; NDMI_mean rose from -0.052 to +0.018.

Future Projections (2025–2055)

Precipitation increases from ~2,200,000 to 2,770,000 mm.km² (+570,000 units). Despite this, precipitation-NDVI (-0.032) and NDMI (-0.019) correlations remain weak (R²=0.102 for NDVI). NDVI_max remains 0.40–0.43 (flat); NDVI_mean 0.08–0.12 (flat); NDVI_min improves from -0.021 to -0.011. NDMI_max 0.30–0.40; NDMI_mean near zero; NDMI_min -0.04 to -0.05 (stable).

Correlation with Precipitation (2013–2024)

Pearson r = -0.0319 (NDVI) and -0.019 (NDMI), both non-significant (p>0.05). Precipitation explains only 10.2% of NDVI variability.

Influence of Non-climatic Factors and Decoupling

To systematically evaluate the influence of non-climatic factors on vegetation and soil moisture, and to identify potential decoupling between precipitation and ecosystem response, the key findings are summarized in Table 1. The table presents, for each factor (lithology, slope, aspect, and elevation), the classes/units with the highest and lowest NDVI and NDMI values, along with their corresponding Pearson correlation coefficients with precipitation. This structure allows for a rapid comparison of the relative importance of each factor and highlights where significant decoupling occurs. Notably, NDMI is most strongly correlated with slope (+0.431) and elevation (+0.344), while NDVI shows its strongest correlations with aspect (+0.175) and lithology (+0.162). Furthermore, the strong negative correlations in the limestone-anhydrite unit (-0.56) and at 1800–2000 m elevations (-0.67) indicate that groundwater management and irrigation have created a system where soil moisture no longer directly follows rainfall patterns.

Table 1. Summary of non-climatic factor influences on NDVI and NDMI, and decoupling correlations with precipitation.

Factors: Lithology Slope Aspect Elevation Decoupling

NDVI Highest (Value/Class) Limestone-Dolomite (0.0879); Limestone-Anhydrite (0.0874) 8–12% (0.0751);

5–8% (0.0738) South (0.0870); Southeast (0.0834) 2000–2200 m (0.0867);

2200–2400 m (0.0846) -

Lowest (Value/Class) Salt Dome (0.0361); Sandstone-Marl (0.0394) >60% (0.0652);

30–60% (0.0640) North (0.0323); Northwest (0.0294) 800–1000 m (0.0594);

657–800 m (0.0624) -

Correlation (r) with NDVI +0.162 -0.015 +0.175 -0.214 -

NDMI Highest (Value/Class) Shale-Marl

(-0.0371); Limestone

(-0.0381) >60%

(-0.0481);

30–60%

(-0.0496) Southeast

(-0.0666); South (-0.0665) >3200 m

(-0.0414);

2800–3000 m (-0.0336) -

Lowest (Value/Class) Alluvium (-0.0658); Limestone-Dolomite (-0.0660) 0–2%

(-0.0678) North (-0.0254); Northwest

(-0.0240) 657–800 m

(-0.0910) -

Correlation (r) with NDMI +0.264 +0.431 +0.261 +0.344 Limestone-Anhydrite:

-0.56;

1800–2000 m:

-0.67

Conclusions

This study demonstrates that non-climatic factors—lithology, slope, aspect, and elevation—exert stronger influence than precipitation on vegetation and soil moisture patterns in the Zakin-Mashari Watershed. While historical precipitation declined (1994–2024), projections indicate increasing trends (2025–2055), yet precipitation-ecological index relationships remain weak in both periods. Limestone-dolomite supports highest vegetation despite low soil moisture due to rapid infiltration; marl-shale retains moisture but supports less vegetation. South-facing "obsequent" slopes show higher vegetation and moisture than north-facing "consequent" slopes due to geological structure promoting weathering and soil development.

Strong negative correlations in specific lithologies and elevations reveal decoupling between rainfall and ecosystem response, attributable to groundwater pumping and irrigation. Projections suggest relative stability under SSP2-4.5, indicating ecosystem resilience mediated by local management and environmental heterogeneity. Priority management includes vegetation restoration on 8–12% slopes, sustainable groundwater use in 1800–2000 m elevations, and water harvesting on south-facing slopes. Future research should use higher-resolution imagery, machine learning, hydrological modeling (SWAT, WEAP), and investigate climate extremes and biodiversity implications. Effective arid-region management requires place-based approaches accounting for complex interactions among climatic, geological, and anthropogenic factors.