Consideration is given to the diagnosis and dynamics of synoptic and subsynoptic forecast error from a potential vorticity (PV) perspective. A depiction of the extratropical ''forecast minus analysis'' PV pattern on a cross-tropopause isentropic surface serves to illustrate characteristic features of the PV-error field, and these features relate both to the instigation, development, and breaking of Rossby waves at the tropopause, and to surface cyclones and anticyclones. An outline is provided of a three-component diagnostic approach for studying PV forecast error. The approach exploits the quintessential PV concepts of quasi conservation, inversion, and attribution, and its essence is illustrated qualitatively by reference to one particular synoptic sequence over the North Atlantic. It also provides a framework for assessing the dynamics of possible mechanisms for generating realized PV-error features. The approach offers a conceptually attractive and diagnostically useful method of analyzing, assessing, and understanding the dynamics of forecast error growth.

[1] An alternative approach is set out for the study of the difference between the forecast and contemporaneous analysis fields of a weather prediction system. Illustrations and interpretations are proffered of this difference when viewed in terms of potential vorticity (PV) for both a case study analysis and a single winter's climatology. The results indicate that the approach provides a compact and insightful description of the difference field's dynamics.

Consideration is given to the impact of both synoptic-scale flow features and different model configurations upon the performance of an Alpine-encompassing regional NWP model. The so-called regional LM forecast model is used to simulate two events selected from the Mesoscale Alpine Programme (MAP), and simulations are undertaken with: different domain sizes (at 7 km resolution); lateral boundary data supplied from two data sets (the ECMWF's operational and MAP Reanalysis fields); and an embedded domain (at 2 km resolution). Quantitative evidence is provided showing (a) the specification of the incident and evolving synoptic-scale flow can exert a major impact upon the quality of the resulting simulations; (b) the simulation of the low-level meso-α scale features of the flow is helped considerably by the refined MAP Reanalysis data set; and (c) the simulated meso-β scale precipitation distribution displays some skill but, at least for one of the two cases, major deficiencies are not offset by the use of Reanalysis data. Zusammenfassung Der Einfluss von synoptisch-skaligen Strömungsmerkmalen und unterschiedlichen Modellkonfigurationen auf die Güte eines regionalen numerischen Wettervorhersagemodells im Alpenraum wird untersucht. Das regionale Wettervorhersagemodell "Lokal-Modell" wird verwendet um zwei ausgewählte Fälle des "Mesoscale Alpine Programme (MAP)" zu simulieren. Es werden Simulationen durchgeführt mit: unterschiedlicher Gebietsgröße des Simulationsgebietes (für 7 km Auflösung); Randdaten, welche von zwei unterschiedlichen Datensätzen stammen (operationelle Analysen vom ECMWF und MAP Reanalyse Felder); und einem eingebetteten Simulationsgebiet für 2 km Auflösung. Es werden quantitative Anhaltspunkte dafür gegeben, dass (a) die Vorgabe der eintreffenden synoptisch-skaligen Strömung und deren zeitliche Entwicklung einen bedeutenden Einfluss auf die Qualität der resultierenden Simulationen ausübt, (b) der verfeinerte MAP Reanalyse-Datensatz die Simulation der bodennahen meso-α skaligen Strukturen verbessern kann und (c) die Simulation der meso-β skaligen Niederschlagsverteilung einige Vorhersagequalität hat, aber zumindest in einem Fall die hauptsächlichen Mängel durch die Verwendung der MAP-Reanalyse nicht beseitigt werden können.

No abstract