Amplified warming in the Circumpolar Arctic has led to shifts in vegetation composition, increased shrubification and northward expansion of tree line. Changes in Arctic vegetation cover and composition are likely to have a significant impact on the tundra ecosystem function through changes in hydrological regimes, energy exchange, and carbon and nitrogen cycling. Understanding the current composition and distribution of vegetation is essential for studying this sensitive ecosystem and predicting responses to environmental change. Mapping Arctic vegetation using coarse spectral and spatial resolution satellite remote sensing data is challenging because of poor image quality due to persistent cloud cover and polar darkness in addition to the high diversity and heterogeneity of vegetation. Our analysis is focused on the Seward Peninsula of Alaska, where the US DOE's NGEE-Arctic project is conducting its studies. Using airborne hyperspectral remote sensing data from NASA AVIRIS-NG and satellite remote sensing platforms, we are developing high resolution maps of vegetation community distribution and estimates of fractional plant functional type (PFT) distributions. We are developing and applying deep learning models trained using field-based vegetation community survey observations for remote sensing-based classification of vegetation communities. High resolution maps of vegetation for Seward Peninsula will be used to understand environmental controls on patterns of vegetation distribution on the landscape, while estimates of PFT distributions will inform and improve the representation of vegetation dynamics in Earth system models.