Land Surface Phenology and Snow Seasonality in the Highland Pastures of Kyrgyzstan: What High Spatio-Temporal Resolution Data Reveals about the Landscape Heterogeneity of Vegetation

We are transitioning into an era of high spatial and high temporal resolution in Earth observation, but few sensors offer high spatio-temporal resolution. Yet, it is this combination that promises to reveal the most about the spatial heterogeneity of land surface phenology. Here we explore three growing seasons of surface reflectance data from the VENμS sensor observing highland pastures in Naryn rayon in central Kyrgyzstan. We build on our prior work that used 30m Landsat NDVI, 1km MODIS land surface temperature (LST), and 500m MODIS snow cover data to explore the linkages between snow seasonality and land surface phenology. Our prior work showed strong topographic controls on snow seasonality that, in turn, affect subsequent phenometrics. In general, a longer snow season or later snow melt date resulted in a higher NDVI peak during the growing season. However, there were significant slope and slope-aspect interactions modulating that linkage. Here we use multiple vegetation indices from 5m VENμS observations, 1km MODIS LST, and 20m Sentinel snow cover data to explore whether the patterns we observed using coarser spatio-temporal resolution data are also evident in the higher resolution image time series. Although the new suite of data does not include higher spatial resolution LST data, the landscape heterogeneity of vegetation is strongly associated with the higher spatial resolution snow cover data which links significantly with the higher spatio-temporal resolution vegetation indices. However, each vegetation index time series offers a different but complementary aspect of the vegetated land surface and, thus, the point in thermal time with the peak index value occurs differs among the indices.