seeks to enable the accurate, spatially scaled and temporally persistent measurement and validation of marine Carbon Dioxide Removal (mCDR)
The envisioned program aims to vastly expand our ability to measure carbon flux parameters in the ocean, enabling comprehensive Measurement, Reporting and Validation (MRV) of mCDR and the creation of a data-driven, model-based marine carbon accounting framework.
Such technologies could also ensure that the quantity and quality of removals are correctly valued in carbon markets and support any economic incentive to accelerate the adoption of mCDR to remove historic emissions.
Validation of sequestered carbon will promote the commercialization of mCDR techniques that are most effective and energy efficient in carbon removal rather than those that are merely easiest to implement but may not actually be as effective or may be so energy intensive themselves that they result in poor overall net lifetime carbon removal
Jesus:
ARPA-E would look to fund the development of sensors that can characterize these parameters at rates on the order of 150 km3/h when scaled to a one- gigaton industry, while matching the accuracy and precision of existing state-of-the-art sensors....
There will be an emphasis on wave-based inferential techniques that do not require co-location of the sensor with the sample of water being evaluated
Model development would target regional-scale, hypothetical but realistic near-future mCDR vignettes within the United States’ Exclusive Economic Zone (EEZ). Models will be developed to demonstrate viability using historical data, before being used as benchmark mCDR models that can be iterated, improved, and validated with future data collected via new sensor systems. The vignettes may include mCDR approaches described in the National Academy of Sciences report on Ocean CDR1, or other mCDR approaches not described but that could be reasoned as techno-economically feasible given a drawdown cost of approximately $100/ton CO2e at megaton to gigaton scale. ARPA-E would plan to fund one or more teams to develop these models. Models are expected to combine physical and biogeochemical ocean modeling, including comprehensive flux modeling of relevant parameters described in Table 1. Developed models would need to achieve a Root Mean Square Error (RMSE) value of no more than 0.1 over time and a temporal Anomaly Correlation Coefficient (ACC) of at least 0.7 when benchmarked against hold-out historical carbon parameter data, before estimates of mCDR effectiveness could be made. Model outputs will need to be developed in coordination with a carbon registry to create a data-based mCDR accounting framework that enables the assignment of credit quality and hence financial value to modeled mCDR events. Selected team(s) will be required to coordinate with a carbon registry to outline an MRV framework that can later be used as a foundation by mCDR project developers to author full protocols, which are peer-reviewed, market-ready methodologies that provide the basis for generating carbon credits. Such an assignment would require robust estimation of the quantity of CO2 drawn down beyond a given temporal threshold (i.e., 10-year or 100-year sequestration) and an indication of the probability that a given quantity of CO2 would remain out of the atmosphere or surface oceans for that duration.
The modeling team(s) would spend the first half of the program developing mCDR models in appropriate regional vignettes. Model performance would be evaluated using historical ambient data or pilot mCDR data, if available. The second half of the program would involve adapting these models for compatibility with data from the sensors under development and simulating the potential increase in MRV effectiveness that could be brought about using these sensors if they were matured and deployed at scale in a regional mCDR scenario. Models would also be used to develop an mCDR accounting framework using output data, evaluate potential improvements to statistical certainty, and assess preliminary implications for the techno-economic validity of the MRV approach. As such, the program would be designed such that modeling and sensor teams are required to coordinate and share data to enable this collaboration.
Language from teaming announcement:
Jesus: