RII Track-4: PermaSense: Investigating Permafrost Landscapes in Transition Using Multidimensional Remote Sensing, Data Fusion, and Machine Learning Techniques

University Of Alaska Fairbanks Campus, Fairbanks AK

Investigators

Abstract

Permafrost underlies around 25% of the Northern Hemisphere terrestrial landscape. Its degradation is impacting northern landscapes and societies. With an increase in the number of remote sensing platforms since the early 2000s, the ability to detect and measure permafrost-region disturbances over large areas has become more feasible. This research, a project we call PermaSense, will provide funding to increase the research capacity of a non-tenure track Research Assistant Faculty member and an early career post-doctoral researcher through extended visits, training, and collaboration at the University of Connecticut (UConn) Department of Natural Resources & the Environment (DNRE). PermaSense will allow the PI and postdoc to build upon their permafrost-region field and remote sensing research program by acquiring new data fusion and machine learning techniques. PermaSense products are directly relevant to the State of Alaska and the nation. The skills and knowledge transfer gained during this project will increase the capacity of permafrost research and remote sensing at the University of Alaska Fairbanks. The research addresses several of the Interagency Arctic Research Policy Committee (IARPC) performance elements related to permafrost, terrestrial ecosystems, coastal resilience, and environmental intelligence as specified in the FY2017-2021 Arctic Research Plan. Permafrost is defined as ground that remains at or below 0 degrees Celsius for at least two consecutive years. Disturbance and warming of near-surface permafrost may lead to widespread terrain instability in ice-rich permafrost regions, impacting ecosystems, hydrology, infrastructure, society, and soil-carbon dynamics. Remote sensing is an important resource for observing, documenting, and better understanding landscape change from local to pan-Arctic scales. However, no one remote sensing tool is particularly suited for detecting and observing the suite of landscape change scenarios associated with transitioning permafrost. PermaSense will investigate myriad land surface changes occurring in permafrost regions using multidimensional remote sensing, data fusion, and machine learning techniques. PermaSense will enhance methodological developments and adaptations to unseal faster, deeper and more accurately analyze large volumes of multidimensional remote-sensing data to address the guiding research question: How extensive is contemporary permafrost degradation in the Arctic and Subarctic? We will conduct an analysis of multidimensional remote sensing observations at four representative permafrost-region study areas that capture the variability in the lateral extent of permafrost. The inherent differences in ecology, climate, landscape history, and their role in transitioning permafrost regions will be tested using common approaches across all sites as well as value-added products available for particular regions. PermaSense will develop an online resource and toolset that will provide spatially and temporally scalable information on permafrost region disturbances and a tool for local planning activities, the scientific community, and regional decision-makers tasked with responding to permafrost regions in transition. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.