Estimate of relationship between state of subarctic vegetation and climatic parameters

In this study we examined relationships between vegetation phenology and the physical environment across the midlatitudes of the Northern Hemisphere between 1982 and 2007. Spatial correlation, principal component analysis, and canonical correlation analisys—all linear techniques—were applied to monthly time series of the Normalized Difference Vegetation Index (NDVI), derived from an AVHRR sensor, sea ice distribution (SSMR-SSM/I), and air temperature (NCEP/NCAR Reanalysis Atlas). A negative correlation between June NDVI of the tundra belt and sea ice conservation in the coastal zone of Arctic Ocean during summer is possibly forced by broad-scale atmospheric circulation. We developed a model with a single predictor and implemented historical-reconstruction NDVI for a long-term period, including the presatellite era. According to this model, the regional patterns of summer vegetation can be explained by peculiarity of atmospheric circulation in spring.