| Summary: | In comparison to other regions, the High Arctic is experiencing accelerated rates of warming (Meredith et al., 2019). Hyperspectral remote sensing may provide a way to monitor changes in productivity without having to make detailed ground-based measurements. During the 2017 field season researchers on Melville Island, Nunavut, collected in-situ hyperspectral data, plant nutrient concentrations, carbon dioxide gas exchange measurements, and various environmental parameters in a wet-sedge tundra environment. These data were processed with the overall objective of determining if spectral information may be used to quantify changes in productivity across the High Arctic. Using a random forest machine learning algorithm, wavelengths from the hyperspectral data were identified for use in nine vegetation indices (VIs) based on relationships to foliar nitrogen concentrations. Using linear regressions, these VIs were compared to the environmental parameters. Although none correlated significantly to foliar nitrogen, three VIs showed p-values < 0.05 (alpha = 0.05) consistently for the following variables: soil nitrate and ammonia concentrations, net ecosystem exchange (NEE), and gross primary productivity (GPP) values. This shows promise for the use of remote sensing techniques to aid in monitoring the High Arctic. Additional research within this field would help pave way towards increased certainty on the kinds of responses that are in store for these landscapes if warming is to continue at an accelerated rate. This may bring increased monitoring frequency and scale of environmental assessment across the High Arctic, granting communities influenced by warming additional tools to aid in safer regional navigation and improved emergency response preparedness. References Meredith, M., Sommerkorn, M., Cassotta, S., Derksen, C., Ekaykin, A., Hollowed, A., Kofinas, G., Mackintosh, A., Melbourne-Thomas, J., Muelbert, M. M. C. M. M. C., Ottersen, G., Pritchard, H., & Schuur, E. A. G. E. A. G. (2019). Polar Regions. In ...
|
|---|