Answers are grounded in this paper and cite their sources.
Idealized large-eddy simulations are conducted to investigate the sensitivity of deep-convection initiation (DCI) to the properties of subcloud updrafts. To this end, a simulated mesoscale convergence line is prescribed and varied across four experiments, the first a control simulation (CTL) informed by observations. Three sensitivity experiments are conducted with doubled convergence amplitude (STRONG) or width (WIDE) and tripled convergence duration (LONG). While CTL generates only congestus clouds reaching heights of 6 km, the other three undergo DCI (cloud tops reaching 10 km) at varying times. The DCI failure in CTL stems from the narrowing and decay of the subcloud updraft between the top of the prescribed convergence layer and the level of free convection (LFC) (a layer with weak convective inhibition, or CIN), which produces only small, rapidly detraining clouds. In contrast, the subcloud updraft in STRONG reaches the LFC with greater width and momentum, producing stronger and deeper cumuli with greatly reduced midlevel detrainment. WIDE behaves similarly to STRONG except that its clouds are further invigorated by (i) increased buoyancy owing to weaker fractional entrainment and (ii) enhanced low-to-midlevel vertical acceleration due to a stronger and wider midlevel buoyancy anomaly. LONG ultimately undergoes DCI primarily due to the erosion of CIN and the development of a strong midlevel buoyancy anomaly, both of which aid ascent within the CIN layer. These findings highlight subtle interactions between midlevel buoyancy and pressure anomalies and cumulus entrainment and detrainment that help to regulate DCI.
No takes yet. Share an insight, caveat, or question.
Kirshbaum et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: