You build and run mathematical representations of the ocean — and increasingly of the coupled ocean-atmosphere-ice-land system that constitutes the planetary climate. The models range from idealized process models (testing a specific mechanism in isolation) to fully coupled Earth system models that project global climate decades into the future. Both are important; neither is sufficient alone.
The intellectual work is part physics, part numerical methods, part software engineering, and part scientific judgment — deciding what processes to represent, at what resolution, with what simplifications, and how to validate the result against incomplete observations. Climate models are among the most complex software systems humans have built, and building them well requires a kind of interdisciplinary synthesis that few other scientific careers demand.
The output of this work is the scientific foundation for climate projections used in everything from IPCC assessments to coastal planning to insurance risk modeling. The models have been substantially right about global temperature trends, sea level rise, and Arctic sea ice loss over the past several decades of validation. Getting them more right — resolving processes like mesoscale ocean eddies, ice sheet dynamics, and carbon cycle feedbacks — is an active frontier with major policy implications.
Kitsune can talk through anything on this page — whether it might suit you, what to do next, questions this page doesn't answer. Everything here is yours to read either way.
The gulf between building a model and being confident in its output is larger than most people outside the field understand. A model can produce beautiful-looking output that is wrong for the right reasons, or right for the wrong reasons. The model evaluation and validation work — understanding what a model gets right, what it gets wrong, and why — is as important as the model development itself, and it is less glamorous.
The career involves substantial exposure to the policy and media environment around climate change, which is a source of both meaning and significant stress. Modelers whose work informs IPCC reports or is cited in policy debates become public-facing scientists whether they want to be or not. Managing that visibility while maintaining scientific integrity in a politicized environment is something the field doesn't train people for explicitly.
The computing resource competition is real. High-resolution ocean and climate modeling requires access to some of the largest supercomputers in the world. Researchers at institutions with good computing access have structural advantages that persist across careers. This is beginning to shift as cloud computing becomes more viable, but it remains a real constraint.
Strong quantitative undergraduate — physics, mathematics, applied mathematics, atmospheric science, or oceanography with heavy quantitative coursework. Programming is essential from the start (Python, Fortran, and increasingly Julia; familiarity with HPC environments). Graduate work in oceanography, atmospheric science, or earth system science. NCAR, GFDL, UKMET, and major European modeling centers employ oceanographers at various career stages, as do university programs with modeling groups.
AI compresses the cost of running models while raising the premium on judging them; emulators inherit rather than replace validation judgment.
AI emulation becomes standard infrastructure; the human role concentrates on model design, validation, and science questions; most location-flexible and most privately portable path.
People drawn to Climate Modeler / Ocean Modelerare often drawn to these — in the order they're closest. The ones marked sit in a different field entirely.