The early cyclogenesis of the March 12--14 1993 Storm of the Century (SOC93) is investigated using observation datasets and mesoscale numerical simulations using the Mesoscale Atmospheric Simulation System (MASS, version 5.12). The initial surface cyclone that forms in the western Gulf of Mexico about 1200 UTC 12 March is traced to a sequence of events that begins early on 11 March.

A mid-tropospheric baroclinic zone oriented along the Texas Gulf Coast provides a broad region of instability for deep, organized convection that is triggered early on 12 March, forming the initial warm-core low. Destabilization of the lower troposphere occurs over a large pool of cold air associated with the remnants of a surface front across central Texas; this low-level cold air is lifted and redistributed at mid-level by vigorous vertical mixing induced by the penetration of the mid-level warm, southwesterly mountain-plains solenoid (MPS) by a low-level (800 hPa) southeasterly flow carrying a hot airmass comprised of moist Gulf air and residual elevated mixed layer (EML) air. The upper-level adjustment to this mid-level cold air produces a meso-alpha scale jetlet that further conditions the environment for deep, organized convection.

Early on 12 March this cold pool moves south along the front range of the Sierra Madres in extreme northeastern Mexico. Daytime heating tightens the front along the terrain, thus accelerating the southwesterly MPS aloft and increasing the southeasterly moist onshore flow below. The MPS provides a strong, southerly ageostrophic signal north of Monterey (MTY), Mexico, while the exit region of the approaching subtropical jet (STJ) induces a northerly ageostrophic flow to the south. The resulting divergence induces a deep column of ascent, triggering moist convection. Convective outflow and acceleration over the latent heating-induced warm core of the storm act to extend and reinforce the upper-level jetlet formed on 11 March in response to the mid-level cold pool; this jetlet further organizes the developing storm. Early development of SOC93 takes place inland in the lower Rio Grande Valley with relatively weak vertical shear between 400 and 700 hPa in the EML, allowing a more direct coupling of the latent-heating induced warm core and surface pressure falls. The resulting cyclone at 1200 UTC 12 March has little westward tilt typically associated with extratropical (cold-core) systems.

These results depart significantly from previous studies in three important ways. First, the mid-level cold pool that provides the initial source of convective available potential energy originates in the upper Rio Grande Valley, and sensitivity studies indicate that the initiation of cyclogenesis is relatively insensitive to the sea-surface temperature field. Second, convection is not triggered primarily by adiabatically-induced warm advection in advance of an upper-level potential vorticity anomaly, but is instead triggered by strong upper-level divergence between the exit region of a jet streak and the thickness ridge from an MPS circulation. Finally, the transition to surface cyclogenesis is a warm-core process driven by deep, upright convection which is, in turn, supported by powerful divergence aloft and an organizing meso-alpha scale jet streak.