|Title||EURORISK/PREVIEW report on the technical quality, functional quality and forecast value of meteorological and hydrological forecasts|
|Publication Type||Technical memorandum|
|Secondary Title||ECMWF Technical Memorandum|
|Authors||Buizza, R, Asensio, H, Balint, G, Bartholmes, J, Bliefernicht, J, Bogner, K, Chavaux, F, de Roo, A, Donnadille, J, Ducrocq, V, Edlund, C, Kotroni, V, Krahe, P, K. Kunz, LM, Lelay, M, Marsigli, C, Milelli, M, Montani, A, Pappenberger, F, Rabufetti, D, Ramos, M-H, Ritter, B, Schipper, JW, Steiner, P, J. Del Pozzo, T-, Vincendon, B|
One component of the PREVIEW project is the analyses of issues linked to the validation of the overall forecast value of probabilistic meteorological and hydrological forecasts. This report is a contribution to this discussion, with attention focused to the case of probabilistic meteorological and hydrological predictions. Since a forecast is valuable only if it has a high technical and functional quality, the forecast value can be assessed only if both the technical and the functional qualities are assessed. Two frameworks are introduced to structure the problem of the assessment of the technical and functional quality: " The 'Technical Quality Framework', based on the assessment of four (technical) attributes: 'forecast characteristics', 'validation characteristics', 'metric' and 'significance test types' " The 'Functional Quality Framework', based on the assessment of four (functional) attributes: 'availability and means of distribution', 'content and format', support, maintenance and training' and 'communication of product's technical quality' Although the quantification of a forecast functional quality is not always possible, when this is the case the forecast value can be quantified by computing a 'generalized' product of the functional quality and technical quality scores. These frameworks are applied to six verification problems, to illustrate how they can be used to structure in a more appropriate way the problem of the validation of a forecasting system. Furthermore, some synthetic (idealized) examples where functional quality can be quantified are discussed, and the methodology that can be applied to assess the forecast value is presented.