The student will know the measures that are used for verification of probability forecasts and forecast probability distributions from ensembles

The student will be able to interpret results of verification of probability forecasts

The student will be able to select the appropriate verification measure for probability forecasts according to the attributes he/she wishes to measure

Module description: This module covers the verification methods commonly used to assess the quality of probability forecasts, along with the attributes of the forecast that they measure, and their interpretation. The attributes are defined in the introductory unit, and are included in the titles of the following units so that you can choose those which interest you. The units of the module are also separated into two tracks, according to the definition of the probability forecast. Probability forecasts of dichotomous (yes-no) variables are the most common in practice and the bulk of the material is devoted to this type of variable. Verification methods for probability forecasts of multi-category variables and full probability distributions such as those obtained from ensemble forecasts are covered in separate but parallel units.

Knowledge of forecast types (Module 1: Introduction)