Premium
On the sensitivity of deep‐convection initiation to horizontal grid resolution
Author(s) -
Tang S. L.,
Kirshbaum D. J.
Publication year - 2020
Publication title -
quarterly journal of the royal meteorological society
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.744
H-Index - 143
eISSN - 1477-870X
pISSN - 0035-9009
DOI - 10.1002/qj.3726
Subject(s) - buoyancy , entrainment (biomusicology) , convection , hydrostatic equilibrium , deep convection , cloud top , environmental science , meteorology , atmospheric sciences , mechanics , cloud computing , physics , computer science , quantum mechanics , rhythm , acoustics , operating system
Abstract Idealized numerical simulations are used to study the sensitivity of deep‐convection initiation to horizontal grid spacing (Δ h). In a conditionally unstable but strongly inhibited environment, a localized surface heating function gives rise to low‐level convergence and a vigorous subcloud updraught that breaches the level of free convection (LFC). The vertical cloud development is sensitive toΔ h: the clouds reach 8–9 km forΔ h = 250 m (CTRL) andΔ h = 500 m (DX500) but only 6–7 km forΔ h = 125 m (DX125). This trend is not associated with major differences in midlevel cloud vigour (e.g., buoyancy and vertical velocity), but can be explained on the basis of cloud‐core mass flux (M c). TheM cprofile is regulated by its value at the LFC, largely set by subcloud processes, and its core‐layer gradient, set by entrainment and detrainment. The former is the largest in CTRL and weakens as the grid is coarsened (DX500) or refined (DX125), both due to a widening and weakening of the subcloud updraught. Whereas in DX500 this widening stems from a failure to adequately resolve the 1‐km‐wide updraught on the model grid, in DX125 it results from stronger total (subgrid plus resolved) turbulent mixing, which increases monotonically asΔ his decreased. The wider and more diffuse updraughts in DX125 and DX500 are also more hydrostatic and generate weaker buoyancy‐driven accelerations. Within the cloud layer, the entrainment is very similar in the three cases, as is the detrainment for the higher‐resolution cases (DX125 and CTRL). However, the midlevel detrainment is relatively weak in DX500, which facilitates slightly deeper ascent than in CTRL. By contrast, the small core‐baseM cand strong midlevel detrainment yields the shallowest clouds in DX125.