Climate Science

Almost nothing downstream of that claim is settled. How much warming for a given amount of carbon. What happens to the West Antarctic ice sheet, and when. Whether the Atlantic circulation has a tipping point and where it is. What August looks like in Lisbon in 2070. What today's flood in your town has to do with any of it. Those are open, hard, and being worked on right now by people who are currently about thirty.

The structural pull is Discovery, and the honest version of that is uncomfortable: this field exists because we are running an experiment on a planet we do not fully understand, and the job is to find out what we have done before it finishes happening. That is Discovery under a deadline, which is an unusual combination — most sciences get to be patient. Explanation sits immediately underneath and is not optional here in the way it is in most research fields; a climate finding that nobody outside the field understands has failed at its actual purpose, and this is one of the few sciences where the ability to explain is a core professional competence rather than a nice extra. Revelation runs alongside, because much of the system is hidden — the past is buried in ice and sediment, and the human fingerprint on a specific event is invisible until somebody does the analysis that reveals it. Preservation appears wherever the work is about what survives the transition.

It is worth being clear about how this field differs from the two next door on this map. Meteorology asks what the atmosphere does next week. Oceanography asks what the ocean is doing. Climate science asks what the whole coupled system does over decades to millennia, and it has increasingly become the discipline where the answer gets handed to somebody who has to act on it. That last part is the newest and least-known thing about it: a substantial and fast-growing share of climate science now happens inside councils, insurers, banks, engineering firms and infrastructure operators rather than universities, because those organisations are now legally and financially obliged to know what the climate is going to do to them.

The daily texture is a terminal, a very large dataset, and a long argument with your own uncertainty. Python, high-performance computing, statistics, and version control are the working tools of most people in this field most days. The exceptions are memorable — a season on the Antarctic plateau, a flask sampling site on a windy hill, a committee room in Westminster — and they are the minority. There is also a peculiar emotional structure to the job that nobody prepares you for: you spend your career becoming precise about something you would rather were not true, and then explaining it to people who would also rather it were not true.

What nobody tells you is that the field has quietly split into two career economies with very different weather. The research end is competitive, postdoc-heavy and slow: the Met Office advertises senior scientist packages up to around £58,400 and a postdoctoral position at around £54,500 with substantial Civil Service pension contribution, and an Ocean Climate Scientific Manager package up to roughly £75,900, while its general scientist average sits nearer £33,000 [official / from salary and job-posting data, Met Office careers + Indeed 2025-26 — note the mix of "total reward package" and base-salary figures across these sources, which are not directly comparable]. The applied end is a hiring market: climate risk analysts in England average around £58,000, roughly 22% above the national average, with the 25th–75th percentile band about £41,100–£66,400 and top earners near £76,300 [from salary and job-posting data, Glassdoor 2025]. There are roughly 355 climate adaptation roles and nearly 3,000 climate-change roles advertised in the UK at a given time [from salary and job-posting data, LinkedIn / Adzuna 2025]. Meanwhile the Climate Change Committee's 2025 report to Parliament found UK adaptation "inadequate" across every sector, with progress too slow, stalled, or heading the wrong way [CCC Progress in adapting to climate change, 2025]. Read that as a seventeen-year-old and the message is that the country has publicly admitted it is not ready and has started paying people to fix it.

Entry is a numerate degree and then a choice about which economy you want. Physics, maths, environmental science, geography, chemistry, earth science and engineering all lead in, and programming is the actual gate rather than a bonus — the field runs on code. The research route means a PhD and usually postdocs; major groups sit at the Met Office Hadley Centre, NCAS and UK universities, at MPI-M in Germany, IPSL in France, and NCAR/GFDL/NASA in the US. The applied route often does not need a PhD at all, and values a master's plus modelling and communication skill, with certifications such as the GARP Sustainability and Climate Risk credential valued in finance [from salary and job-posting data, sector recruitment 2025]. In Portugal, IPMA runs a Divisão de Clima e Alterações Climáticas producing the national climate bulletins and the Portuguese climate scenarios and portal work with Instituto Dom Luiz at the University of Lisbon under the ADAPT programme, and the university route runs through física, engenharia do ambiente, geografia and meteorology-adjacent programmes at Lisboa, Porto and Aveiro [IPMA / university admissions 2025].

People and jobs. Research economy: competitive, postdoc-heavy academic market (Mode 1), a long pre-AI pattern. Applied economy: Climate Change Committee's 2025 statutory report found UK adaptation 'inadequate... across every sector' [CCC 2025]; ~355 UK climate-adaptation and ~3,000 climate-change roles advertised at a given time [LinkedIn/Adzuna 2025]; US climate-change-analyst average pay $97,659 (July 2026) [ZipRecruiter 2026].

The main limit on getting in. Research entry is gated by PhD/postdoc funding availability (a credential/funding gate, not AI displacement). Applied entry is gated more loosely — Mode 1 is explicit a PhD is 'not required' for Climate Services or Climate Risk roles, and the binding constraint there is closer to awareness/talent-pipeline (students not knowing these applied roles exist) than to a hard credential floor.

Entry-level jobs. Aggregate entry-level demand is growing, driven almost entirely by the applied economy's statutory-adaptation and disclosure-regulation-fuelled hiring; this masks the research economy's real, AI-independent competitiveness, which the segmented derived-metrics block (Section 8) surfaces explicitly.

Cost of training. Research-track PhDs carry real opportunity-cost debt risk against a competitive academic market; applied-track master's-level entry into climate services or climate risk carries less debt risk given above-average applied-economy pay.

Because the applied economy's binding constraint is chronic under-capacity against statutory and regulatory obligation, AI that accelerates modelling, detection, and attribution reads as capacity relief for an admittedly under-resourced sector (per the CCC's own words) rather than as displacement — while the research economy's AI-independent funding competitiveness should not be conflated with an AI effect.