Dr Henry Houskeeper1, Dr Tom Bell2, Dr Alejandra Mora-Soto3, Prof Kyle Cavanaugh1
1University of California Los Angeles, Los Angeles, United States, 2Woods Hole Oceanographic Institution, Woods Hole, United States, 3University of Victoria, Victoria, Canada
Giant kelp provides three-dimensional structure to subtidal communities that supports productive and diverse coastal ecosystems, but this species is temporally dynamic, with regional populations varying on seasonal to decadal time scales. Satellite image analysis enables high spatial and temporal coverage for mapping existing and historical giant kelp populations, but extracting coverage of giant kelp canopy from large satellite datasets requires automated approaches for kelp classification that are robust across dissimilar and dynamic coastal environments. Presently, long-term datasets are limited in many regions due to insufficient resources to support large-scale manual image classification, combined with incomplete automation of robust satellite image classification tools. Recently, temporal and geographical coverage of satellite kelp canopy datasets has been expanded based on parallel collaborations to advance automation of image classification, including through approaches based on citizen science, cloud computing, and artificial intelligence. We present preliminary results of historical (spanning over 40 years) giant kelp forest dynamics along temperate coastlines of North and South America using updated methods for automated satellite (Landsat) classification. We present an assessment of inter-regional differences in method performance, and we investigate long-term variability in regional kelp canopy extent. We investigate inter-regional patterns in kelp forest resilience and resistance, and we compare broad patterns in coverage and latitudinal extent between the northern and southern latitude domains.
Presentation Slides – Henry Houskeeper
Biography:
Henry F. Houskeeper is a post-doctoral scholar in the Department of Geography at the University of California Los Angeles working with the NASA Floating Forests Project, and is a 2022 Google Cloud Research Innovator. Henry received his Ph.D. in Ocean Sciences from the University of California Santa Cruz with Raphael Kudela, where he also served during two years as an ocean science remote sensing mentor for the NASA Student Airborne Research Program (SARP). His areas of research include environmental remote sensing of coastal waters and automation of satellite image classification.
