Snow Cover Change Trends (NDSI) and the Impact of Regional and Global Teleconnection Patterns

Abstract

Introduction: Snow cover is among the most critical natural resources influencing water availability, surface runoff, agriculture, and tourism, with significant implications for both natural and human systems. Climatic signals often exert considerable influence on various meteorological elements, particularly snow cover during and after precipitation events. The impact of climate variability on snow accumulation and melt dynamics plays a pivotal role in the management of water resources in river basins that rely on snowmelt for their hydrological regimes. Accordingly, the present study aims to analyze the behavior of snow cover in the Karun River Basin and its relationship with climatic signals over a 22-year statistical period (2001–2022).

Materials and Methods: Initially, validated snow cover products derived from the MODIS sensor—specifically the Normalized Difference Snow Index (NDSI)—were retrieved within the Google Earth Engine (GEE) platform, spatially aligned with the Karun River Basin. Subsequently, time series data for 29 teleconnection patterns known to influence the climate of Iran and the target basin were compiled for the study period. To assess the trend and magnitude of snow cover changes, the non-parametric Mann–Kendall test and Sen’s slope estimator were employed. Furthermore, Pearson correlation analysis was conducted using SPSS software to evaluate the statistical relationships between snow cover data and the teleconnection indices.

Results and Discussion: The findings of this study indicate a declining trend in snow cover across the Karun River Basin during the cold months of the year, accompanied by statistically significant negative shifts. Specifically, during January, February, March, April, November, and December, the snow cover exhibited a consistent downward trend with abrupt reductions throughout the study period. The output of the Sen’s slope estimator further revealed a total decrease of approximately 2,794 square kilometers in snow-covered area. Correlation analysis between snow cover extent and teleconnection indices demonstrated that snow cover is significantly influenced by several atmospheric–oceanic patterns. At the 0.05 significance level, simultaneous negative correlations were observed with indices such as GMSST (February), EAWR (March), SOI, and RMM2 (December). During the cold season (November to May), snow cover also showed significant delayed correlations (one-month lag) with Solar Flux, EA, AAO, EAWR, RMM2, and SOI. Additionally, at the 0.01 significance level, a strong negative correlation was found with the AMO index in November.

Conversely, significant positive correlations at the 0.05 level were identified between snow cover and indices including EAWR, EPNP, SCA, PBO, OSI, NINO3.4, NINO4, ONI, and PNA. During the cold season (November to May), positive associations were also evident with SOI, MEIv2, ESPI, and EPO. At the 0.01 level, snow cover exhibited strong positive correlations with the SCA index (May) and ENSO-related indices (ONI, NINO3.4, NINO4, MEIv2, ESPI, and NINO3.1), particularly in December. Overall, the highest observed correlations were associated with various ENSO indices and teleconnection patterns such as EAWR, MEIv2, and the Southern Oscillation Index (SOI). These results suggest that both synchronous and lagged fluctuations in selected teleconnection patterns have statistically significant relationships with snow cover variability in the Karun Basin. Given the predictability of these large-scale climate signals and their strong correlations with snow cover dynamics, they offer valuable potential for improving future snow cover forecasts in the region.

Conclusion: Based on the results of this study during the 2001–2022 period, it was determined that snow cover in the Karun River Basin, located in the southern Zagros region, has been undergoing a declining trend. Moreover, fluctuations in several of the examined teleconnection patterns—across both their positive and negative phases—were found to exert synchronous and lagged influences on the extent of snow cover in the basin. These influences manifested as either reductions or increases in snow-covered area, depending on the phase and timing of the respective climate signals. Given the relative predictability of the examined teleconnection patterns, it is recommended that future studies incorporate advanced modeling approaches such as machine learning and deep neural networks. The application of these techniques can enhance the accuracy of estimating fluctuations in teleconnection indices and, consequently, improve the forecasting of snow cover variability in the Karun River Basin.