Applying biomass-spectra relationships from Ivotuk, Alaska to six other sites in the Alaskan Arctic
Warming in the Arctic has resulted in a lengthening of the growing season and changes in the distribution and composition of tundra vegetation. Past studies have mapped tundra vegetation at relatively coarse spatial scales; however, vegetation changes in the Arctic are occurring at spatial scales within a few meters. This research relates plant functional types (PFTs) to hyperspectral remote sensing data from Ivotuk, Alaska (68.49°N, 155.74°W). Ivotuk is located on the North Slope of the Brooks Mountain range, and consists of the four vegetation communities of moist acidic tundra (MAT), moist nonacidic tundra (MNT), mossy tussock tundra (MT), and shrub tundra (ST). Hand-held hyperspectral data were collected during the 1999 growing season (5 June-27 August) at biweekly intervals using narrow, ~1.42 nm wavebands ranging from 400-1062 nm. Biomass data were collected from areas adjacent to these measurements, and sorted into PFTs including graminoids, forbs, mosses, horsetails, lichens, and deciduous and evergreen shrubs. Spectral bands were grouped into nine sections: Blue (400-450 nm), Green 1 (450-500 nm), Green 2 (500-550 nm), Yellow (550-580 nm), Orange (580-600 nm), Red (600-680 nm), Red Edge (680-725 nm), NIR (725-900 nm), and NIR plateau (900-1062 nm). Each section was averaged for a total of nine spectral variables, and PFTs were regressed against original and continuum-removed (CR) spectral variables using LASSO regression.
Preliminary results for peak growing season indicate no significant relationships for MAT or MNT using original spectra. Strong relationships (adjusted r2 = .31-.48) were found for three PFTs in ST. MT had one significant relationship for moss (adjusted r2 = .61). Continuum-removed spectra had no significant relationships for CR-MAT, CR-MNT, or CR-ST. CR-MNT had one significant relationship to deciduous shrub-dead foliar (adjusted r2 = .63). When data are analyzed without separation by community, there were significant relationships for original spectra and four PFTs (adjusted r2 = .32-0.45), and continuum-removed spectra and one PFT (adjusted r2 = .63). For biomass categories in which significant relationships were found, all data from Ivotuk will then be used as the training set, and tested against data from six other sites in the Alaskan Arctic. The overall goal of this analysis is to use remote sensing data to develop methods that can be applied to other sites in the Alaskan Arctic and used to estimate changes in vegetation composition and biomass quantity occurring with warming in the Arctic.