Distributed and semi-distributed correlation analysis of structural and functional sediment connectivity index with soil erosion

Introduction
Soil erosion and sediment transport are among the most significant natural processes contributing to land degradation and the deterioration of soil and water resources within watershed systems. These processes are inherently dynamic and are influenced by a complex interplay of natural and anthropogenic factors, including topography, land use/land cover, rainfall variability, and vegetation conditions. Understanding the mechanisms governing sediment generation, transport, and connectivity within a watershed is therefore essential for effective soil and water conservation planning and management. In recent years, the concept of sediment connectivity has emerged as an efficient framework for describing the degree of linkage between sediment source areas and watershed outlets. The Index of Connectivity (IC) provides a quantitative measure of sediment transfer potential based primarily on topographic characteristics derived from Digital Elevation Models (DEMs). Nevertheless, recent studies have demonstrated that sediment connectivity is not solely controlled by topography, but is also affected by other factors such as surface roughness, land use patterns, and vegetation dynamics. Vegetation cover plays a critical role in reducing runoff velocity, mitigating raindrop impact energy, and regulating sediment transport processes. Consequently, incorporating dynamic vegetation indicators such as the Normalized Difference Vegetation Index (NDVI) can substantially improve the accuracy of monthly sediment modeling and connectivity assessments. Despite these advances, most existing studies have approached sediment connectivity from a structural or static perspective, with limited attention given to seasonal and monthly variability. At the same time, the Revised Universal Soil Loss Equation (RUSLE) has been widely applied for estimating soil erosion; however, its relationship with sediment connectivity at monthly temporal scales has received comparatively little attention. Accordingly, the present study aims to investigate the relationship between sediment connectivity and soil erosion at a monthly scale in the Kasilian Watershed, while also evaluating the effectiveness of incorporating NDVI into sediment connectivity analyses and monthly sediment estimation.
Materials and Methods
This study was conducted in the Kasilian Watershed, located in Mazandaran Province, northern Iran. The watershed is characterized by complex topography, a semi-humid to humid climate, and diverse land use/land cover types, including forestlands, rangelands, agricultural areas, residential zones, and rock outcrops. These characteristics make the region a suitable environment for investigating erosion and sediment transport processes. To evaluate sediment connectivity, the IC was employed. Initially, the annual IC was calculated for the year 2021 based on topographic attributes derived from the DEM. Subsequently, monthly IC values were estimated by incorporating monthly NDVI data as a weighting factor within the connectivity model. NDVI data for the Kasilian Watershed were extracted from MODIS satellite imagery and normalized to a range between 0 and 1 prior to model implementation. The relationship between sediment connectivity and monthly soil erosion was then assessed by comparing IC results with monthly soil erosion estimates derived from the Revised Universal Soil Loss Equation (RUSLE). Correlation analyses between monthly sediment connectivity and monthly soil erosion were performed at three analytical levels: land use/land cover classes, slope categories, and working units. Monthly mean values were extracted using GIS-based spatial analysis tools. After testing data normality using the Shapiro–Wilk test, Pearson or Spearman correlation coefficients were applied in SPSS software depending on the distribution characteristics of the variables.
Results and Discussion
The results revealed a significant relationship between sediment connectivity and soil erosion at the monthly scale. Incorporating monthly NDVI-based vegetation dynamics into the calculation of the sediment connectivity index substantially improved the correlation between sediment connectivity and soil erosion compared with the conventional annual IC approach. The correlations between the results of monthly and annual sediment connectivity calculations with monthly and annual erosion of the Kasilian watershed was investigated in slope, land use/land cover, and work unit classes showed the significant role of vegetation cover using the NDVI in the sediment connectivity index calculations to increase the correlation between sediment connectivity and soil erosion in the monthly time interval. This finding highlights the critical role of vegetation dynamics in regulating sediment transfer processes at monthly scale. The highest correlation between soil erosion and monthly sediment connectivity was observed in January and in slope classes (-0.85, sig. 0.02), while the lowest correlation was observed in October in land use/land cover (0.14, non-significant) and also in March in work units (0.17, non-significant). The results also showed that the combination of slope class and forest land use/land cover in two work units 2 and 4 caused a significant decrease in monthly soil erosion in these two units in April (0.01 t ha^-1). Also, working unit 4 has the lowest monthly sediment attachment rate in January (-5.64) and October (-5.54). The strongest correlations were observed during periods of dense vegetation cover across the watershed, particularly from May to November. In contrast, during periods characterized by sparse or absent vegetation cover, topographic conditions and land use/land cover factors exerted a stronger influence on both erosion estimates and sediment connectivity patterns, emphasizing the dominant role of these controls in the absence of sufficient vegetation protection.
Conclusion
The findings of this study demonstrated that soil erosion and sediment connectivity are interrelated and highly dynamic processes that vary considerably at the monthly temporal scale. Incorporating vegetation-cover information and dynamic remote sensing approaches can substantially enhance the accuracy of sediment connectivity modeling. The results further indicated that vegetation cover plays a key role in regulating the relationship between topography and sediment transport. Under conditions of sparse or absent vegetation cover, slope gradient and land use/land cover emerged as the dominant controlling factors, whereas under dense vegetation conditions, the protective effect of vegetation weakened the direct relationship between topography and sediment transfer processes. From a practical perspective, integrating remote sensing data with sediment connectivity models provides a more accurate and efficient framework for watershed management. This approach improves the identification of erosion-prone areas and supports more informed decision-making in soil and water resources management. Overall, the monthly sediment connectivity approach proposed in this study demonstrated clear advantages over the conventional annual topography-based method, providing a more realistic representation of erosion and sediment transport dynamics for watershed managers and environmental planners.