Arctic plants are small in stature and spectrally diverse, which presents challenges to current NASA missions to visualize effects of disturbance or directional vegetation change via mapping. Remotely sensed data having fine spatial (ca. 10 cm pixels) and spectral grain (eg. “hyperspectral”) will therefore help resolve patches of many arctic plant groups, such as dwarf shrubs, bryophytes and lichens and separate them from litter, wood or rock/soil. To address these challenges, in summer 2018 we sampled vegetation at 15 different sites around Fairbanks, Alaska using ground-based and airborne hyperspectral sensors under eight different AVIRIS ng flight lines next gen flight lines (circa 2017-2018). At each AVIRIS flight line, we estimated percent cover of plant functional types in eleven 1m2 quadrats every 10 m along a 100m transect. We then flew our UAV and imaging spectrometer (Headwall Micro A-series VNIR, 400-1000 nm, 330 bands, 10 cm pixels). Spectral signatures of any surfaces were sampled using a field spectroradiometer (PSR+ Spectral Evolution, 400-2500 nm, 1nm bands). We collected 600+ georeferenced scans from 70+ species/plant functional types at 25+ different sites around Alaska. Spectral profiles showed many different plant species have similar to indistinguishable signatures (eg. Paper birch and Alder) while many plant functional types that have been grouped together (eg. Moss) were very spectrally heterogenous. UAV-based hyperspectral imagery (ca. 4-10 cm pixels) resolved pure pixels of many artic plants. Our approach resolves fine grained ecological features, such as networks of circular patches (mostly lichens, brophytes and mineral soil) over very large areas (ca. 10,000 m2), created by small cryoturbation features (frost boils). We explore spectral unmixing and other statistical approaches to compare mapping results using our spectral library with AVIRIS ng (4 m pixels) and our UAV-based VNIR hyperspectral imagery.