*2.5. Correlation Analysis*

To assess the relationship between snowpack and land surface phenology, we calculated the Pearson correlation coefficients between four snow season indicators (FSD, LSD, SSL and SCD) and three phenology indicators (SOS, EOS and LOS), and evaluated the significance of the correlation (*p*-value) by performing a *t*-test [36,65]. Based on the vegetation distribution, we focused on the correlation between snow season metrics and phenological metrics in six different vegetation types: forest, shrub, grass, alpine vegetation, cultivated vegetation and desert vegetation. In addition, an elevation gradient was set from 2500 m to 4500 m to study the vertical variation of the correlation between vegetation and snow seasonality. Correlations were calculated for each pixel and then further counted according to different elevations or vegetation types. The spatial patterns of snow seasonality and land surface phenology over the study area were characterized by the mean of the metrics.
