A comparative assessment of the uncertainties of global surface ocean CO2 estimates using a machine-learning ensemble (CSIR-ML6 version 2019a) – have we hit the wall?

• Published in: Geoscientific Model Development Vol. 12; no. 12; pp. 5113 - 5136
• Main Authors: Luke, Gregor, Lebehot, Alice D, Kok, Schalk, Scheel Monteiro, Pedro M
• Format: Journal Article
• Language: English
• Published: Katlenburg-Lindau Copernicus GmbH 10.12.2019; Copernicus Publications
• Subjects: Artificial intelligence; Bias; Carbon dioxide; Carbon dioxide estimation; Carbon dioxide flux; Carbon dioxide measurements; Clustering; Confidence; Datasets; Drifters; Eddy kinetic energy; Estimates; Floats; Fluxes; Gliders; Industrial research; Interannual variability; Kinetic energy; Learning algorithms; Machine learning; Methods; Oceans; Optimization; Procedures; Regression analysis; Root-mean-square errors; Seasonal variability; Statistical analysis; Statistical inference; Training; Variability
• ISSN: 1991-962X; 1991-959X; 1991-962X; 1991-9603
• DOI: 10.5194/gmd-12-5113-2019

Over the last decade, advanced statistical inference and machine learning have been used to fill the gaps in sparse surface oceanCO2 measurements (Rödenbeck et al., 2015). The estimates from these methods have been used to constrain seasonal, interannual and decadal variability in sea–air CO2 fluxes and the drivers of these changes (Landschützer et al., 2015, 2016; Gregor et al., 2018). However, it is also becoming clear that these methods are converging towards a common bias and root mean square error (RMSE) boundary: “the wall”, which suggests that pCO2 estimates are now limited by both data gaps and scale-sensitive observations. Here, we analyse this problem by introducing a new gap-filling method, an ensemble average of six machine-learning models (CSIR-ML6 version 2019a, Council for Scientific and Industrial Research – Machine Learning ensemble with Six members), where each model is constructed with a two-step clustering-regression approach. The ensemble average is then statistically compared to well-established methods. The ensemble average, CSIR-ML6, has an RMSE of 17.16 µatm and bias of 0.89 µatm when compared to a test dataset kept separate from training procedures. However, when validating our estimates with independent datasets, we find that our method improves only incrementally on other gap-filling methods. We investigate the differences between the methods to understand the extent of the limitations of gap-filling estimates ofpCO2. We show that disagreement between methods in the South Atlantic, southeastern Pacific and parts of the Southern Ocean is too large to interpret the interannual variability with confidence. We conclude that improvements in surface ocean pCO2 estimates will likely be incremental with the optimisation of gap-filling methods by (1) the inclusion of additional clustering and regression variables (e.g. eddy kinetic energy), (2) increasing the sampling resolution and (3) successfully incorporatingpCO2 estimates from alternate platforms (e.g. floats, gliders) into existing machine-learning approaches.