How to judge the quality and value of weather forecast products

John Thornes, D Stephenson

Research output: Contribution to journalArticle

96 Citations (Scopus)

Abstract

In order to decide whether or not a weather service supplier is giving good value for money we need to monitor the quality of the forecasts and the use that is made of the forecasts to estimate their value. A number of verification statistics are examined to measure the quality of forecasts - including Miss Rate, False Alarm Rate, the Peirce Skill Score and the Odds Ratio Skill Score - and a means of testing the significance of these values is presented. In order to assess the economic value of the forecasts a value index is suggested that takes into account the cost-loss ratio and forecast errors. It is suggested that a combination of these quality and value statistics could be used by weather forecast customers to choose the best forecast provider and to set limits for performance related contracts.
Original languageEnglish
Pages (from-to)307-314
Number of pages8
JournalMeteorological Applications
Volume8
Issue number(3)
DOIs
Publication statusPublished - 1 Sept 2001

Fingerprint

Dive into the research topics of 'How to judge the quality and value of weather forecast products'. Together they form a unique fingerprint.

Cite this