Submarine Scientific awarded NSF SBIR Phase I grant to develop GPU-optimized ocean carbon modeling platform
Submarine Scientific has been awarded a $294,759 NSF Small Business Innovation Research (SBIR) Phase I grant to develop an advanced, automatically configurable ocean modeling platform for marine carbon dioxide removal (mCDR) MRV.
The project, led by co-founder Dr. Veronica Tamsitt, will integrate GPU-optimized ocean modeling software with AI-driven automation to streamline the configuration, validation, and analysis of regional ocean models; making high-resolution, scientifically rigorous modeling accessible to non-expert users. At the core of the platform is Oceananigans, a cutting-edge GPU-native ocean simulation framework, combined with tools for automated data retrieval, model validation, and uncertainty quantification.
Key research objectives include:
Automating ocean data processing workflows
Exploring the use of large language models (LLMs) to simplify model setup
Developing gradient-based methods for optimizing mCDR project parameters.
Test real-time adaptive domain boundaries to improve computational efficiency.
If successful, the platform will significantly reduce the cost and complexity of ocean modeling for mCDR verification and reporting — lowering barriers to entry across the sector and helping generate the high-integrity carbon credits needed to build a credible marine carbon removal market.
