Until recently the scientific literature has largely described the “greening” of the Arctic; from a remote sensing perspective, an increase in the Normalized Difference Vegetation Index (NDVI), or a similar satellite-based vegetation index. Vegetation increases have been heterogeneous throughout the Arctic, and were reported to be up to 25% in certain areas over a 30+-year timespan. However, more recently, arctic tundra vegetation dynamics have gotten more complex, with observations of tundra “browning” being reported. We used several remote sensing data products, including that from the Global Inventory Monitoring and Modeling System (GIMMS), as well as higher spatial resolution Landsat, and Corona/Gambit and commercial satellite data, to evaluate the patterns of arctic tundra vegetation dynamics (greening and browning) at circumpolar, regional, and landscape scales over the past 3-5 decades. At the circumpolar scale, we focus on the spatial heterogeneity (by tundra subzone and continent) of tundra dynamics over the past 35 years. Landsat time series and finer-resolution imagery allow us to evaluate the landscape-scale heterogeneity of tundra greening and browning for northern Alaska and Siberian Russia.
Results/Conclusions
The GIMMS circumpolar dataset revealed a general arctic tundra greening trend since 1982, however, the rate of greening decreased with increasing latitude. Although the greening rate was greatest in the southern tundra, the inter-annual variability in NDVI metrics was most strongly related to summer warmth in the middle latitudes of the tundra. Despite a general greening trend, the number of years of NDVI increase were similar to those for NDVI decrease, and the temporal autocorrelation suggests no evidence of acceleration. Landsat and higher resolution data indicate up to 25% areal increases of tall shrubs for Siberian Russia, with a range of 3.3-63.9% of area within landscapes greening, and 0-10.6% of areas browning. Multi-dataset analyses reveal that tundra greening and browning (i.e. increases or decreases in the NDVI respectively) are generated by different sets of processes. Tundra greening is largely a result of either climate warming, lengthening of the growing season, or responses to disturbances, such as fires, landslides, and freeze-thaw processes. Browning on the other hand tends to be more event-driven, such as the shorter-term decline in vegetation due to fire, insect defoliation, consumption by larger herbivores, or extreme weather events (e.g. winter warming or early summer frost damage). The continual use of multi-platform remote sensing data across spatial scales allows us to monitor these dynamics and elucidate their controls.