📍Trajectory Modeling
From a modeling perspective, the first key question to solve when deploying passively drifting material into the open ocean was "Where does it go?" In the chaotic, high-energy open-ocean environment, answering this question was not trivial. The Ocean Modeling team spent most of our time researching, improving, and automating trajectory models so they could be tuned with in-situ data, run as part of a Monte Carlo ensemble, and driven with different environmental datasets to produce accurate results.
Running Tide's Software Engineering team and Ocean Modeling collaborated to build a custom wrapper around Ocean Parcels, a popular Lagrangian Simulator. Then we were able to model horizontal surface advection of deployed sensor buoys by incorporating ocean currents, waves, and windage, and using a gradient descent algorithm to tune the model's fit to observed trajectories. This approach allowed us to determine the probable sinking location of biomass statistically and even predict trajectories for the purpose of deployment site selection in the days and hours leading up to vessel departure from port.
We presented our modeling work at AGU Ocean Sciences Meeting in 2024 and published a whitepaper on our approach. Both are linked below and in the Document Repository section of this website.
Before modeling actual deployments, we had three opportunities to deploy GPS sensors alone in the target deployment region near Iceland in the North Atlantic. These "Sensor Only" pilots were critical to developing our modeling tools in advance of biomass deployments.
Each deployment in 2023 was unique; trajectories were highly variable based on weather and windage. While we were able to confidently select deployment sites to avoid risk of beaching within the expected float times of the wood chips, it would be difficult to model multi-month float trajectories for future buoys supporting macroalgae growth to maturity. Deploying sensors in the North Atlantic in 2023 was a good start, but significantly more work is needed to de-risk beaching of long-duration drift trajectories.
Expansion on this work could include horizontal advection during sinking. During rapid sinking, horizontal advection is minimal. But this effect would be important to understand for buoys with macroalgae growth, which would likely sink more slowly and with greater drag than small wood chips. Some preliminary sink rate data about the rates of wood chips sinking through the water column can be found in the report below:
Last updated
