Skip to content
  • Natural hazards
  • Modelling and projections
  • Heat and heatwaves

Reading an attribution study

How science links a specific event to warming, what it establishes, and the three misreadings almost everyone makes of its results.

7 stepsabout 15 min read

Attribution studies have become the standard answer to the question asked after every heatwave. They are also among the most poorly reported, because their result is a probability and a probability is easily read as a certainty.

This path follows the order in which the notions hold together: first what a rare event is, then how a signal is separated from noise, finally what the method establishes. It ends with the language the results are written in, which is the last opportunity to get them wrong.

  1. Step 1 of 7Termabout 2 min read

    Return period

    The average interval between two occurrences of an event of given intensity; a hundred-year flood has a one in a hundred chance of happening in any year.

    Start here: nearly every attribution result is expressed as a change in frequency, and this is the notion most often read wrongly.

    Term
  2. Step 2 of 7Termabout 2 min read

    Internal variability

    Fluctuations of the climate produced by the system itself, with no change in external forcing.

    The background noise against which a signal is sought. Without it, variations that are atmospheric chance get attributed to climate.

    Term
  3. Step 3 of 7Termabout 2 min read

    Attribution

    The scientific practice of assessing how far human-caused climate change has altered the probability or intensity of an observed event.

    The formal definition, now that both ingredients are in place.

    Term
  4. Step 4 of 7Briefabout 3 min read

    Attributing an event to climate

    What an attribution study actually establishes, why it speaks of probabilities and never of a single cause, and how to read its results without making them say more than they do.

    The full file: the two-simulated-worlds method, what attributes well, and what still resists.

    Brief
  5. Step 5 of 7Studyabout 2 min read

    Anthropogenic contribution to global occurrence of heavy-precipitation and high-temperature extremes

    Erich Markus Fischer et al. · Nature Climate Change · 2015

    A founding paper, quantifying the share of heat and rainfall extremes already imputable to warming. Read the method before the figures.

    Study
  6. Step 6 of 7Studyabout 2 min read

    2021 North American heatwave amplified by climate change-driven nonlinear interactions

    Samuel Bartusek et al. · Nature Climate Change · 2022

    A concrete case, the 2021 heat dome, and the difficulty it poses: the event exceeded what simulations judged possible.

    Study
  7. Step 7 of 7Termabout 2 min read

    Confidence level

    The codified vocabulary by which the IPCC states how solid a conclusion is and how probable a result.

    Finish with the calibrated vocabulary. 'Likely' means 66 % in an IPCC report, and knowing it changes how everything above reads.

    Term

Path complete

You can carry on with the full directory or another path.