Forecast verification: current status and future directions

• DOI: 10.1002/met.52
• Publication Date: 2008-03-20T18:17:34Z
• Authors: B. Casati; L. J. Wilson; D. B. Stephenson; P. Nurmi; A. Ghelli; M. Pocernich; U. Damrath; E. E. Ebert; B. G. Brown; S. Mason
• Keywords: Forecast verification; Functional verification; Data verification; Software verification

Abstract Research and development of new verification strategies reassessment traditional forecast methods has received a great deal attention from the scientific community in last decade. This effort arisen need to respond changes encompassing several aspects process, such as evolution forecasting systems, or desire for more meaningful approaches that address specific user requirements. Verification techniques account spatial structure presence features fields, which are designed specifically high‐resolution forecasts have been developed. The advent ensemble motivated re‐evaluation some scores probability forecasts. expected climatological increase extreme events their potential socio‐economical impacts revitalized research studies addressing challenges concerning event verification. issues encountered operational environment widely discussed, needs different communities identified, models assess value users proposed. Proper practice correct interpretation statistics extensively promoted with recent publications books, tutorials workshops, open‐source software tools. paper addresses current verification, reviews most recently developed techniques, provides recommendations future research. Copyright © 2008 Royal Meteorological Society Crown right Canada.