| Summary: | Globally, kelp forests provide the foundation of temperate coastal ecosystems through the creation of three-dimensional habitat which supports many significant economic, cultural, and ecological species. With the increasing threat of local and global anthropogenic stressors including climate change, the long-term, large-scale monitoring of kelp forests is crucial in understanding the response of these foundation ecosystems to spatio-temporal drivers in a time of rapid global change. On the coast of British Columbia, Canada, kelp forests of Macrocystis pyrifera and Nereocystis luetkeana show highly variable patterns of change, however, lack sufficient time series data to truly understand trends, threats, and drivers. In particular, Haida Gwaii, an archipelago off the main coast of British Columbia, supports some of the most expansive kelp forests in the province. However, local Haida people have observed drastic declines at traditional harvesting sites leading to the prioritization of the management, monitoring and protection of these crucial ecosystems throughout their territory. Remote sensing technology now provides an effective way to track changes and trends in remote kelp forest ecosystems at large scales like that of the Haida Gwaii coastline. Specifically, Earth Observation satellite imagery of a variety of spatial resolutions exists back to the 1970s which can be leveraged for mapping kelp forest canopies through time. This research aims to quantify the distribution, variability, and drivers of change of Haida Gwaii kelp forests with the use of satellite imagery and historical data. Specifically, to address this goal, (i) we develop a methodological framework that enables the creation of a long-term dataset of kelp canopy area using archived multispectral satellite imagery from multiple satellite sensors that vary in their spatial resolution (0.5 m – 60 m) and temporal coverage (1973-2021). To do this, we combine a workflow of standardized remote sensing practices and an adaptable image-to-image ...
|
|---|