A sensitivity of a MAP IOP 15 event to the estimated uncertainties in the upper-level dynamical factors

Mesoscale mountain ranges, such as the Alps, are well known regions of frequent lee cyclogenesis. An event of a rapid Alpine lee cyclogenesis during MAP (IOP15) was associated with severe weather in the southeastern Alpine region: extreme Bora (with gusts over 50 ms^-1) developed over the northern Adriatic, while heavy precipitation (60 mm/24h) occurred over the northeastern Italy, Slovenia and continental Croatia.

A numerical analysis on the sensitivity of the event to the initial uncertainties in the upper-level dynamical factors was performed with the use of MM5 model. First, a statistics of the differences in the upper-level potential vorticity was calculated using ECMWF and NCEP analysis of the 21 event of the strongest Mediterranean storms. This statistics served as an estimate of the initial-analysis uncertainties in the upper-level trough. Thereupon, the statistics was used to deterministically scale the modifications in the initial conditions of the analyzed event (with a 90^th percentile of the derived error statistics), regarding the intensity and position of the upper-level trough. In this way, a consistent method was used to produce an ensemble of sensitivity simulations, with an estimate of the model uncertainty in the upper-level trough only.

The results indicate a notable variability in the intensity and position of the analyzed lee cyclone. The spread of cyclone intensities was the greatest during the lee cyclone generation phase (up to 50% of the cyclone intensity), while cyclone tracks diverged the most in the later phase of the lifecycle (750 km). In addition, the (in)existence of synoptic critical levels and span of the integral background Froude number over several flow regimes in the sensitivity simulations suggest that Bora was highly sensitive to the modifications in the upper-levels, resulting in the spread of intensities of ±30% of the wind speed in the control run. This suggests that the initial-analysis uncertainties in the upper-level dynamical factors have a strong potential to decrease the predictability of the Genoa lee cyclones as well as cyclonic Bora in the Adriatic, and that the improvement of forecast accuracy in the region could be achieved through the ensemble forecasting.