Misforecasts of Synoptic Systems: Diagnosis via PV Retrodiction

Abstract

A strategy is outlined to explore the contribution of errors in the specification of tropopause-level flow features to the misforecast of individual synoptic and subsynoptic weather systems. The approach is founded upon a potential vorticity (PV) perspective of the flow. It entails the identification of key upper-level PV error elements at the forecast time, their Lagrangian retrodiction back to the initial analysis time in a process incorporating intermittent PV inversion, and a further forward integration starting from the revised initial state. The procedure can yield direct information on the role and strength of upper-level effects, and if the revised ?forecast? verifies successfully it can indicate the location and nature of the analysis error. To illustrate the approach a case study example is provided of a significant misforecast of rapid surface frontal-wave cyclogenesis. A conventional 24-h forecast with a limited-area NWP model failed to capture the low-level cyclone, and the forecast exhibited significant errors in both its low-level and upper-level components of the PV distributions. The revised simulation shadows the analyzed development much more successfully. The result is discussed in the context of the need to improve the initial analysis fields and to devise alternative forecasting strategies.