OP2341 Data-driven prediction of landscape changes in permafrost coasts

NERC ONE Planet DTP

Coastline erosion due to climate change has been a major concern in the Mackenzie Delta region and is increasingly impacting its communities. We have excellent survey records of up to annual rates of change, but these mask the true drivers, which are continually influencing the nature of coastal change. This means we cannot relate specific seasonal trends, weather patterns, sea conditions, or extreme events to specific coastal changes and the impacts and implications they hold. These changes are directly affecting local infrastructure and potentially playing a significant role in the carbon releases that affect the earth systems processes. This research will explore exciting new quantitative vector tracking of time-lapse imagery of coastal type sites across the Mackenzie delta region and associate the continuous changes with environmental monitoring. These datasets will underpin the development and implementation of a new data-driven framework to predict the coastline changes.

The success of this project will result in: (i) a new process understanding of permafrost behaviour in the holding/releasing of carbon into the environment; (ii) a new predictive tool for both baseline and extreme environmental changes on coastal erosion, and (iii) open-source database and codes contributing to a further expansion of research in the field.

This project will use state-of-the-art monitoring approaches and machine-learning codes developed inhouse to train the model using available data. Participants of this project will have an opportunity to join fieldwork in the Arctic supported by Natural Resources Canada and support efforts to improve local community resilience.

Key Research Gaps and Questions:

• What are the main causes of Arctic coastline changes?
• Can continuous monitoring establish the relationships between environmental changes and coastal slope responses?
• Can response patterns from different permafrost coastal types be effectively applied to predict widescale patterns of change?

Prerequisites:

Experience in environmental/weather monitoring, geomorphology, remote sensing, programming language (e.g. Python) For more information, please contact Hoang Nguyen ([Email Address Removed])