Investigating climatic drivers of snow phenology by considering key-substage heterogeneity

Investigating the main climatic drivers responsible for changes in snow cover phenology (SCP) is crucial for making scientific countermeasures to ensure water resources security in global mountainous regions. However, most studies have explored drivers of SCP changes using a fixed substage division scheme and correlation analysis (referred to as the traditional method), potentially limiting reliability and accuracy in mountainous areas with complex terrain and climate. Here, a novel method is developed to efficiently identify main climatic drivers of SCP changes. This method employs a flexible scheme to account for the spatial heterogeneity of the dominant sub-period (the sub-period with the major climatic effect) in combination with regression analysis. Using the arid region of China as a case study, the new method was applied to three SCP parameters including snow cover days, snow start date, and snow end date, based on a seamless snow cover dataset from 2002 to 2019. The method’s effectiveness was evaluated by comparing it with the traditional method. The results indicate significant spatial heterogeneity in the dominant sub-period(s), closely associated with local temperature and elevation. The traditional method failed to accurately identify the main drivers of SCP changes, as evidenced by sub-region and elevation zone analyses showing adjusted coefficient of determination (R²) of < 0.5 in most cases. This inadequacy is attributed to its fixed and inappropriate scheme of substage division. In contrast, the new method, with its flexible scheme, achieved much higher adjusted R² values (mostly > 0.5) and exhibited better performances in predicting SCP changes. Thanks to the new method, climatic causes of SCP changes were successfully identified in 12 hotspots (regions with significant SCP changes), all with adjusted R² > 0.5. The climatic causes were found to vary significantly across different hotspot regions and SCP parameters. The proposed method holds significant potential to enhance the reliability of analyses concerning main climatic drivers of SCP changes in mountainous regions globally.