Satellites, Servers, and Spatial Overlay: A multifaceted workflow for investigating spatial determinants of Coral Bleaching Trends across the Pacific

Rosanna Neuhausler, University of California Berkeley, Department of Geography, Berkeley, CA, United States, Laurel Larsen, University of California, Department of Geography, Department of Civil and Environmental Engineering, Berkeley, United States and C. Mark Eakin, NOAA/NESDIS/STAR Coral Reef Watch, College Park, MD, United States
Abstract:
The 2014-2017 coral bleaching event, triggered by rises in temperature, saw reductions of coral cover worldwide. Across the Pacific ocean, local coral losses were frequently as high as 50%, such as in regions in the Great Barrier Reef. While temperature and past bleaching temperature thresholds were able to predict a large amount of the bleaching that was observed, discrepancies exist; some areas with high temperature increases did not bleach, while others with lower temperatures did. Our main goal is to pinpoint spatially variable factors that can explain the variable levels of resistance to temperature changes of the previous bleaching event. To reduce potential confoundedness and omitted variable bias, we conduct an ocean-wide analysis and include a multitude of variables, including both those whose influence has been seen to weaken coral as well as factors that have not yet been considered at this scale. Due to the scope of this project, specifically the wide spread of data sources and potential gaps, our first step has been the development of a data pipeline. Our solution is fully open source, utilizing Apache NiFi for the processing pipeline, MinIO object storage for raw data, and local servers for metadata storage and url pointers to the MinIO server. It contains a multistep process for spatial overlay, following feature extraction from remotely sensed data within Google Earth Engine. We pull from a variety of data sources, such as Landsat-7, MODIS, governmental census and economic data, the RAM Legacy Stock Assessment Database, and existing local surveys, to tease out the general trends of these complex environments. Preliminary results from a subset of islands suggest improved prediction capabilities when including factors such as sediment plumes. Further outcomes of this analysis may illuminate best management practices for building or preserving reef resistance to temperature induced bleaching.
Back to: Using Satellite Remote Sensing and in Situ Data to Evaluate Marine Biodiversity and Ecosystems II