Determining the uncertainty of the North Atlantic CO2 uptake using CMIP5 models

The North Atlantic Ocean is one of the strongest sinks for anthropogenic carbon dioxide (CO₂) on the planet. To predict the North Atlantic response to the on-going increase of atmospheric CO₂ concentration, having a clear understanding the CO₂ uptake variability is crucial. Although the amount of sea surface observations has increased by about a factor 5 since 2002, the observational space and time distributions are irregular, which leads to difficulties when investigating the whole basin variability of CO₂ uptake in the North Atlantic. To overcome the observational coverage issue, a Multi Linear Regression (MLR) mapping technique was used by Watson et al. (2009). Here we determine, using CMIP5 model data, the uncertainty in ΔfCO₂ predicted by the MLR technique developed by Watson et al. (2009). The basin-wide and monthly model data act as a reference for the MLR predicted field. The main advantage of the current analysis is therefore to provide uncertainty on the predicted ΔfCO₂ where no observations were recorded, which was lacking in the Watson et al. (2009) observational study. Overall, the choice of the spatial and temporal subdivisions is essential when estimating the CO₂ uptake in the entire North Atlantic, especially when considering periods with very few measurements (before 2002). The amount of data used to predict ΔfCO₂ is a key factor to reduce the uncertainty associated with the MLR mapping technique.