Abstract
The origin of 653 convective storms occurring over a 5000 km2 area immediately east of the Colorado Rocky Mountains from 18 May to 15 August 1984 was examined. Seventy-nine percent of the 418 storms that initiated within the study area occurred in close proximity to radar-observed boundary-layer convergence lines. This percentage increased to 95% when only the more intense storms (≥60 dBZe) were considered. Colliding convergence lines initiated new storms or intensified existing storms in 71% of the cases. A new storm took a median time of 24 min to grow to 30 dBZ following line collision.
The convergence lines ranged in length between ten and several hundred kilometers. Both radar and mesonet stations indicated that the primary convergence was concentrated in a zone 0.5 to 5 km in width. These lines were characterized on Doppler radar as thin lines of enhanced reflectivity between 0 and 20 dBZe and as a line of strong radial or azimuthal gradient in Doppler velocity. These lines were observed even in clear air in the absence of any clouds. The origin of many of the convergence lines was unknown and requires further study. The most common identified origin was from convective storm outflows. Other origins were believed to be topography and differential heating.
This study, which utilized radar data, supports the findings of Purdom (1982), who utilized satellite data, which indicated that mesoscale boundary-layer convergence lines play a major role in determining where and when storms will form. These results suggest that what often appears as random thunderstorm formation (air mass thunderstorms) is usually deterministic.
Major advances now appear possible in the 0–2 h time-specific forecasts of thunderstorms. Realization of this potential will require the integration of Doppler radar to detect and monitor convergence lines, high resolution satellite data to monitor cloud growth, and surface and sounding data to estimate atmospheric susceptibility to deep convection.