Same attribution-science method behind reports like World Weather Attribution and evidence in cases such as Urgenda v. Netherlands and Milieudefensie v. Shell, rendered here with three linked 2D panels instead of a 3D bar chart. Event intensity in each world is a Gaussian:
Factual: X₁ ~ N(μ₀ + ΔT·k, σ₀²·(1+v))
Counterfactual: X₀ ~ N(μ₀, σ₀²)
P(event) = P(X > threshold) = 1 − Φ((threshold − μ)/σ)
Risk Ratio RR = P₁ / P₀
Fraction of Attributable Risk FAR = 1 − P₀/P₁ = 1 − 1/RR
- Top panel — the two exact analytic bell curves, shaded past the threshold line: the area ratio of the shaded regions is the risk ratio.
- Middle panel — a live Monte Carlo dot-strip (beeswarm): every dot is one simulated model run, jittered vertically only for readability; its horizontal position is the sampled intensity. Orange/purple dots crossed the threshold.
- Bottom panel — Risk Ratio swept continuously across ΔT at the current threshold/variability; drag anywhere on it (or use the ΔT slider) to scrub warming and watch RR/FAR move together. Toggle log scale for small vs. extreme thresholds.
- Variability change — widens or narrows the factual bell independently of its mean shift, exactly as in the source model.
An FAR near 1 means the event was almost entirely attributable to human warming; an FAR near 0 (or negative) means it was about as likely, or less likely, in a world without greenhouse-gas emissions.