Predictability of heavy precipitation induced by mesoscale convective vortices

Recent work has demonstrated that mid-tropospheric vortices form within mesoscale convective systems and often persist well beyond the decay of the convection. In particular, analyses of radar and satellite image animations, as well as wind profiler and sounding data have established that 15-30 long-lived mesoscale convective vortices (MCVs) occur each warm season over the United States. In this talk, we first review the climatology of these events and examine how vertical motion induced by MCVs can initiate and modulate convection.

We then evaluate the detection and prediction of MCVs using the Rapid Update Cycle for the 1999 warm season. It is found that the RUC has a high degree of skill for MCV detection, but it is suggested that because of horizontal resolution limitations, the RUC is deficient in MCV prediction. We examine a single MCV case in detail using high resolution simulations from the PSU/NCAR mesoscale model. We show how an accurate 12-24 h forecast of heavy rainfall initiated by an MCV is possible. This work, in its totality, demonstrates that by virtue of being balanced circulations, MCVs may themselves be highly predictable on time scales of 12 h, thereby implying enhanced predictability of attendant precipitation.