Meteorology 5540
Mini-Lecture 8: Probability Forecasting II

I. Cost/Loss Ratios

  • The cost/loss ratio is merely cost of doing something (like protecting your rose bushes from frost) divided by the cost of the potential loss.
  • Sometimes is financial, other times it is merely a personal or emotional cost.
  • Probability density functions can be used to define the average expected loss for a given forecast. You need to sit down with your client and discuss estimated losses for certain meteorological cateogories. As a forecaster, your job is then to provide accurate probabilites that the event described by each category occurs.
  • If the cost/loss ratio determined by the client is exceded, then action should be taken.
  • Problems: a small chance of a catostrophic event occuring may not result in the cost/loss ration being exceeded, yet if that event occurs, they could be financially ruined! Be careful.

    II. Probability in worded forecasts

  • Numerical probability in a worded forecast can created confusion
  • Mostly cloudy with a 60% chance of showers and a few thunderstorms... what does this mean?
  • Seperate the numerical POP from the worded forecast to avoid confusion: Increasing clouds with possible rain later today. Chance of measurable rain 60% this afternoon.
  • Forecasts presented by the media in general do not include numerical probability. This conveys less information to the end user than a quantitative POP forecast.

    III. Glossary of words and phrases used in worded forecasts

  • This on-line reference provides definitions for forecast words and pharases that are used by northern and central California NWS offices.
  • Note that these are not necessarily strict definitions!

    IV. References and on-line links

  • Hughs, L. A., 1980: Probability forecasting. NOAA Technical Memorandum NWS FCST 24, NWS Central Region, Central Region, Kansas City, MO.
  • Murphy, A. H., and R. W. Katz, 1985: Probability, Statistics, and Decision Making in the Atmospheric Sciences. Westview Press, Boulder, Co. 545 pp.
  • Vislocky, R. L., J. M. Fritsch, and S. N. DiRienzo, 1995: Operational omission and misuse of numerical precipitation probability expressions. Bull. Amer. Meteor. Soc., 76, 49-52.
    Updated May 8, 1997