Meeting Documents
Developing and Validating a Model of the Salish Sea for Simulating Ocean Alkalinity Enhancement on GPUs with Oceananigans + OceanBioME
Presented at: AGU Annual Meeting 2024
Abstract
Novel open source ocean model code, Oceananigans, has been developed in Julia by the Climate Modeling Alliance (CliMA) providing greatly enhanced user-friendliness and modularity, and leveraging GPU acceleration for unprecedented computational speed. Biogeochemical models are being developed in parallel through OceanBioME, offering the flexibility to use biogeochemical models of varying complexity and easily couple individual-based models and sediment models. This new modeling ecosystem holds significant potential for research and operational measurement, as well as reporting and verification of marine Carbon Dioxide Removal. Here we present initial results from the development and validation of a regional configuration of Oceananigans + OceanBioME for the Salish Sea, aiming to simulate an ocean alkalinity enhancement (OAE) pilot project located in Port Angeles Harbor. We employ the LOBSTER medium complexity biogeochemical model, incorporating carbonate chemistry, and force the model using realistic historic atmospheric reanalysis data. Additionally, we apply ECCO-derived restoring of temperature and salinity at the open boundary, as boundary inflows in Oceananigans are still under development. We validate the model baseline physics and biogeochemistry with available observations and compare to existing output from the ROMS-based LiveOcean model currently being used to model the OAE pilot project site. We will conduct a series of 1 month duration continuous alkalinity point source releases at the Port Angeles site during different months of the year to assess seasonal variability in the plume trajectory.
View Document (AGU) »