Our research focuses on analyses of Saint Lawrence Valley heavy precipitation events occurring in the cold season. Although synoptic-scale and planetary-scale circulation anomaly precursors to these events may be identified several days prior to these events, predictions of these extreme events are often poor. We suggest that the terrain-induced frontogenesis is responsible for the enhancement of already-heavy precipitation amounts. We present results from analyses of two cases of extreme precipitation to verify our suggestion: 1) The rainfall of 9-10 November 1996 in which a record-breaking 24-h amount of 135 mm was recorded in Montreal, QC and 2) The ice storm of 5-9 January 1998 in which freezing rain amounts exceeded 110 mm in several locations throughout the Saint Lawrence valley.

For each case, we have found excellent analogues in terms of its sea-level pressure and 1000-500 hPa thickness field. Despite the similarity in the synoptic-scale circulations fields, each analogue was associated with only 20 to 40% of the observed precipitation values recorded in the extreme events. Synoptic-scale analyses of these cases yields the following key differences between the analogue and its base case 1) The extreme cases are characterized by optimum placements of surface cyclone and geostrophic frontogenesis zones 2) Precipitable water values, tropospherically-integrated water-vapor transports, and lower-tropospheric equivalent potential temperatures were each substantially larger in the extreme case than in the corresponding analogue case.

Our mesoscale analyses, supported by high-resolution numerical simulations, reveal more robust surface frontogenesis and strong associated vertical circulations in the extreme cases than in the analogues. Our sensitivity experiments reveal that some of the mesoscale frontogenesis is produced by orographic channeling of cold air towards either a stationary or northward-moving warm air mass.