August 16, 2012

Probabilistic weather forecasts

For the Olympics, the British Meterology Office was producing animated probabilistic forecast maps, showing the estimated probability of various amounts of rain or strengths of wind at a fine grid of locations over Britain.  These are a great improvement over the usual much more vague and holistic predictions, and they were made possible by a new and experimental high-resolution ensemble forecasting system.  (via)

I will quibble slightly about the probabilities in the forecast, though.  The Met Office generates a set of predictions spanning a reasonable range of weather models and input uncertainties, and then says “80% change of rain” if 80% of the predictions have rain at that location.   That is, 80% means an 80% chance that a randomly chosen prediction will say “rain”, it doesn’t necessarily mean that “out of locations and hours with 80% forecast probability, 80% of them will actually get rain”.

It’s possible to improve the calibration of the probabilities by feeding the ensemble of predictions into a statistical model, and researchers at the University of  Washington have been working on this.  Their ProbCast page gives probabilistic rain and temperature forecasts for the state of Washington that are based on a statistical model for the relationship between actual weather and the ensemble of forecasts, and this does give more accurate uncertainty numbers.

avatar

Thomas Lumley (@tslumley) is Professor of Biostatistics at the University of Auckland. His research interests include semiparametric models, survey sampling, statistical computing, foundations of statistics, and whatever methodological problems his medical collaborators come up with. He also blogs at Biased and Inefficient See all posts by Thomas Lumley »

Comments

  • avatar

    This is neat! I hope more forecasters adopt something along these lines. Surely they must have probabilities in-house, but the forecast itself is usually a point estimate.

    Regarding the middle paragraph, probabilities don’t usually correspond to frequencies like you suggest they should (that’s called coverage right?) But I think it’s cool that if their physics model is not quite right (say, it underpredicts rain) then you can say the forecast should have been *some transformed version* of the output of their simulations, and then infer what the transformation should be.

    Great stuff.

    12 years ago

    • avatar
      Thomas Lumley

      It’s called coverage if you’re a frequentist, and calibration if you’re a Bayesian: everyone agrees it’s something that you want.

      But,yes, as you say, cool (with isolated showers turning to rain, given UK weather)

      12 years ago