About the job:


The Southern Ocean Carbon–Climate Observatory (SOCCO) at the CSIR has a vacancy for Research Scientist in large-scale ocean carbon cycle processes that can apply machine learning, artificial intelligence, and statistical modelling to advancing understanding of global CO fluxes and their direct policy relevance, positioning the researcher  at the forefront of international climate science. By linking Southern Ocean research to global carbon cycle assessments, the successful candidate will work within a dynamic, world-leading team and help shape future strategies for carbon monitoring and management. This position is based in Cape Town.


Key rsponsibilities:


Lead the development, refinement, and application of machine learning frameworks to improve surface ocean CO¿ flux estimates, building on SOCCO’s flagship ML-based CO¿ flux product (CSIR-ML6).
Lead R&D and innovation in the development of emerging AI technologies and techniques and application in the field of ocean-climate research
Evaluate the performance of Earth System Models (CMIP6/7) against the ML-based observational CO¿ flux product, focusing on: mean-state climatologies, spatial, temporal, and seasonal variability, and long-term trends.
Diagnose discrepancies between ocean-climate models and reconstruction products using Observing System Simulation Experiments (OSSEs) and advanced statistical approaches.
Conduct high-resolution OSSEs to optimize Southern Ocean CO¿ observing strategies and reduce uncertainty in sink estimates.
Supervise, train, and mentor postgraduate researchers, contributing to the growth of capacity in AI-driven climate science and advancing its integration into ocean-climate modelling
Actively contribute to international science-policy processes such as the Global Carbon Budget and State of the Climate reports.
Contribute actively to the development of competitive research proposals to secure external funding for long term sustainability of research activities.


Qualifications, skills and experience:


A Doctoral degree in Earth System Science, Oceanography, Applied Mathematics, Statistics, Computer Science, Data Science, or a related field; with at least five years’ proven expertise in machine learning, AI, and high-performance scientific computing;
At least five years’ experience applying supervised and unsupervised machine learning methods (e.g., neural networks, gradient boosting, clustering) to environmental or geophysical datasets, including development of approaches to improve predictive skill and model performance;
Demonstrated ability to work with large, sparse, and heterogeneous datasets, including gap-filling, noise reduction, spatiotemporal post-processing, and integration of observational and model outputs;
Ability to translate machine learning outputs into climate and environmental science applications;
Demonstrated track record of international collaboration and publications in leading peer-reviewed journals;
Excellent communication skills, including capacity-building and postgraduate supervision.


Closing Date


16/09/2025