Spatio-Temporal Characterization of Arctic Landscapes Using Geospatial Analytics

Amplified warming in the Arctic has likely increased the rate of landscape change and disturbances in northern high latitude regions. Satellite remote sensing is a valuable tool for monitoring natural and anthropogenic changes occurring in remote, northern high latitude environments over multiple time scales. It offers the potential to characterize the vegetation, land cover, hydrology, geomorphology and permafrost characteristics of the Arctic landscape and improve and improve our understanding of changes these ecosystems are undergoing due to effect of natural and anthropogenic climate change and changing disturbance regimes. Combined with ground based observations of ecological processes, remote sensing offers opportunities for upscaling the ground based measurements to better understand the larger landscape.In this dissertation research I have developed 1) new techniques for integration of remote sensing data set from a range of platforms with different spatial and temporal resolutions; 2) computationally efficient statistical and machine learning techniques to get ecological insights from large volumes of high dimensional remote sensing data; 3) methods to characterize and map vegetation characteristics at NGEE Arctic field sites in Alaska; and 4) techniques for identification and attribution of disturbance regimes in Alaska. In a close partnership with field ecologist, geospatial and machine learning techniques I have developed in this research has led to new insights and high resolution datasets of Arctic vegetation processes.