Somebody's town floods. Within the hour the question arrives — was this climate change? — and for most of history the honest answer was a shrug about how you cannot attribute any single event. That answer is now obsolete. You compare the observed event against the same event simulated in a world without human emissions, and you produce a number: how much more likely, how much more intense. After the January 2025 Los Angeles wildfires, a rapid study found climate change made the extreme fire weather driving them roughly 35% more likely [survey_aggregator, World Weather Attribution 2025]. The World Weather Attribution service, set up in 2015, has now run more than 100 such studies across heatwaves, rainfall, drought, floods, wildfires and cold spells, in days rather than years [survey_aggregator, WWA 2025].
The distinct contribution is causation made visible on a human timescale. The palaeoclimatologist supplies the past, the modeller supplies the projections; you are the one who connects an abstract global trend to the specific water that is currently in somebody's living room. That is Revelation — the fingerprint was always there, and the professional act is producing it fast enough to matter.
The consequences travel further than the science. Attribution findings feed loss-and-damage negotiations, insurance and liability arguments, and a growing body of climate litigation, and there is evidence that attribution framing shifts public support for climate policy across countries [official, Nature Climate Change 2025]. The tooling is accelerating too — frameworks such as ClimaMeter now return initial attribution results within hours of an event [survey_aggregator, ClimaMeter / Frontiers in Climate 2025].
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 deadline is the discipline and it is brutal. You are doing statistical inference on a chaotic system, in four days, while people are still counting the dead, knowing the number you publish will be quoted for years and cannot be quietly corrected later. The field's answer is heavy methodological pre-registration — deciding how you will do the analysis before you know what the answer is — and that rigour is the only thing standing between rapid science and advocacy.
Sometimes the answer is no, and you have to say so. Not every disaster is made worse by climate change; some are driven by land use, poor infrastructure or bad luck. Publishing a null result into a public conversation that expects confirmation is professionally uncomfortable and is precisely what makes the positive results worth anything.
And this is one of the few sciences with a defendant. Your work turns up in courtrooms and in arguments about who owes whom for what. That is a very different relationship to your own output than most researchers have, and it attracts a particular kind of person — people who want their work to have teeth rather than citations.
Physics, maths, statistics, environmental or atmospheric science degree with genuinely strong quantitative and programming skills; extreme value statistics is the specific mathematics that distinguishes this work. A PhD in climate or atmospheric science is the standard route, though the field is young and unusually open to people arriving from statistics and data science. The centres of gravity are the World Weather Attribution collaboration, Imperial's Grantham Institute, KNMI in the Netherlands, and allied groups [survey_aggregator, WWA 2025]. Because the studies are open-method and published quickly, it is one of the few research areas a student can actually read end-to-end while still at school.
Speed does not remove the pre-registration discipline or the requirement to publish honest null results; a wrong fast answer is more damaging than a right slow one.
Rapid attribution becomes fully mainstream; value concentrates in methodological integrity and courtroom-adjacent testimony as litigation use grows.
People drawn to Climate Attribution Scientistare often drawn to these — in the order they're closest. The ones marked sit in a different field entirely.