A Forecast and Analyzed Cyclogenesis Event Diagnosed with Potential vorticity

Abstract

A cyclone that developed over the eastern United States during December 1992 is investigated using a potential vorticity (PV) framework. Upon partitioning the perturbation PV field (which includes the near-surface potential temperature distribution) into upper-level (UL), low-level (LL), and lower boundary (LB) components, the extent to which particular PV anomalies contribute to the cyclone development is quantified by inverting the PV distribution associated with each component. In addition, the accuracy of a 48?84-h forecast, produced by the CCM2 version of NCAR's Community Climate Model, is assessed. The assessment primarily concerns the 36-h geopotential height tendencies forced at 850 mb by the individual components of the analyzed and forecast PV tendency fields. It is found that the lower boundary (923 mb) thermal wave accompanying the storm is amplified mostly by the winds associated with an upper-level disturbance and that latent heating plays an important, but secondary, role in near-surface development. The upper-level disturbance, which is accompanied by the formation of a trough of low-? air at the tropopause, existed in an amplified form prior to the onset of rapid surface deepening. The winds associated with the upper-level perturbation PV initiate the growth of this trough, whereas the winds associated with the LL and LB components further amplify it. In general, the development is characterized at first by the mutual amplification of PV anomalies (i.e., baroclinic instability more important) and later by the superposition of anomalies. Although the total height tendency fields forced by the analyzed and forecast data were quite similar, the fields forced by the individual components (which sum to the total) showed substantial differences. This suggests that the model may have been right for the wrong reason. In particular, the model overforecast the upper-level disturbance. This was manifested in the development of a tropopause-? anomaly in the model that was 4 K too cold. The strong circulation induced by this anomaly undoubtedly contributed to the overamplification of the lower boundary wave in the model. On the other hand, the model underestimated a low-level height fall center, apparently of diabatic origin, and displaced it to the northwest of the analyzed feature. These findings point to the importance, in numerical weather prediction, of accurately resolving the near-discontinuity in PV values at the tropopause and of properly handling the release of latent heat throughout the troposphere.