On the Scale and Magnitude of Precursors to Simulated Isolated Convective Initiation

Numerical forecasts of convective development remain hampered by an inability to robustly forecast the timing and location of convective initiation (CI). Numerous observational and numerical modeling studies have suggested that it is necessary to capture extremely small-scale variations in the pre-storm boundary layer to refine CI forecasts. This work aims to better quantify the scale of these CI precursor features, particularly with respect to the potential for capturing these features with a surface-based observation network. We examine the statistics of spatial scale, temporal longevity, and magnitude of perturbations in surface fields (e.g. temperature, winds, pressure, moisture) that precede isolated convective initiation in idealized numerical model simulations. Specifically, the CM1 cloud model is initialized with observed soundings from locations and times where there was widespread convective activity in the absence of strong large-scale forcing. Nascent cloud elements are tracked through their growth and maturity into precipitating storms, and the surface fields surrounding the developments of these storms are examined. Consistent, observable, perturbations in the surface temperature and wind fields are noted 30-60 minutes before the onset of precipitation, with less prominent perturbations in the pressure and moisture fields. The implications of these findings for the design of a CI-capable surface observation network for data assimilation are also discussed.