We are excited to announce that the repository will soon undergo an upgrade, featuring a new look and feel along with several enhanced features to improve your experience. Please be on the lookout for further updates and announcements regarding the launch date. We appreciate your support and look forward to unveiling the improved platform soon.
dc.contributor.author | Gregor, Luke![]() |
|
dc.contributor.author | Lebehot, Alice D.![]() |
|
dc.contributor.author | Kok, Schalk![]() |
|
dc.contributor.author | Scheel Monteiro, Pedro M.![]() |
|
dc.date.accessioned | 2020-03-25T06:36:08Z | |
dc.date.available | 2020-03-25T06:36:08Z | |
dc.date.issued | 2019-12-10 | |
dc.description.abstract | Over the last decade, advanced statistical inference and machine learning have been used to fill the gaps in sparse surface ocean CO2 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 gapfilling 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 twostep 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 of pCO2. 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 incorporating pCO2 estimates from alternate platforms (e.g. floats, gliders) into existing machinelearning approaches. | en_ZA |
dc.description.department | Mechanical and Aeronautical Engineering | en_ZA |
dc.description.librarian | am2020 | en_ZA |
dc.description.sponsorship | This work is part of a post-doctoral research fellowship funded by the CSIR Southern Ocean Carbon – Climate Observatory (SOCCO) through financial support from the Department of Science and Technology (DST) and the National Research Foundation (NRF) and hosted at the MaRe Institute at UCT. | en_ZA |
dc.description.sponsorship | This work received support from the European Space Agency (ESA)’s OCEANSODA – Ocean Acidification project (contract no. 4000125955/18/I-BG). | en_ZA |
dc.description.uri | https://www.geoscientific-model-development.net | en_ZA |
dc.identifier.citation | Gregor, L., Lebehot, A.D., Kok, S. et al. 2019, '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?', Geoscientific Model Development, vol. 12, no. 12, pp. 5113-5136. | en_ZA |
dc.identifier.issn | 1991-959X (print) | |
dc.identifier.issn | 1991-9603 (online) | |
dc.identifier.other | 10.5194/gmd-12-5113-2019 | |
dc.identifier.uri | http://hdl.handle.net/2263/73823 | |
dc.language.iso | en | en_ZA |
dc.publisher | European Geosciences Union | en_ZA |
dc.rights | © Author(s) 2019. This work is distributed under the Creative Commons Attribution 4.0 License. | en_ZA |
dc.subject | Machine learning | en_ZA |
dc.subject | Common bias | en_ZA |
dc.subject | Root mean square error (RMSE) | en_ZA |
dc.subject | Advanced statistical inference | en_ZA |
dc.subject | Sea–air CO2 fluxes | en_ZA |
dc.subject | Gapfilling method | en_ZA |
dc.subject | CSIR-ML6 | en_ZA |
dc.subject.other | Engineering, built environment and information technology articles SDG-13 | |
dc.subject.other | SDG-13: Climate action | |
dc.subject.other | Engineering, built environment and information technology articles SDG-14 | |
dc.subject.other | SDG-14: Life below water | |
dc.subject.other | Engineering, built environment and information technology articles SDG-09 | |
dc.subject.other | SDG-09: Industry, innovation and infrastructure | |
dc.subject.other | Engineering, built environment and information technology articles SDG-04 | |
dc.subject.other | SDG-04: Quality education | |
dc.title | 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? | en_ZA |
dc.type | Article | en_ZA |