Large-eddy simulations of diurnally heated airflow over idealized terrain

Orography disturbs impinging airflow by acting both as an obstacle in its path (mechanical forcing) and as an elevated heat source (thermal forcing). While the basic dynamical responses to each forcing type are broadly understood, the combined response to the two forcings is not, particularly in nonlinear, time-varying flows where the two forcings may interact. In this study, large-eddy simulations of diurnally heated airflow past an elongated mountain ridge are conducted to investigate these interactions. The analysis focuses on the sensitivity of daytime boundary-layer horizontal and vertical mass fluxes to key environmental and terrain-related parameters. To separate the flow dynamics into relevant regimes, a modified nondimensional mountain height (M_m, or inverse Froude number) is derived that takes into account both mechanical and thermal forcing. In general, thermal forcing acts to reduce M_m and enable otherwise blocked flow to surmount the terrain. When M_m is sub-unity, the impinging flow easily surmounts the terrain, with thermal forcing acting to accelerate (decelerate) it over the windward (lee) slope. Despite the strong windward ascent in this regime, the mountain-scale mass convergence is nearly zero and little mass is transported across the boundary-layer top. As M_m exceeds unity, flow separation occurs in the lee and the thermal forcing draws upslope motions over both mountain faces, which greatly increases the mountain-scale mass convergence. These upslope flows collide in a narrow flow-separation zone over the lee slope, giving rise to an intense and deep updraft that rapidly ventilates boundary-layer air into the free troposphere. This updraft is further enhanced by its superposition with elevated gravity-wave ascent over the lee slope. The combined response to mechanical and thermal forcing is interpreted through a new pressure-based scaling that predicts the relative contributions of each effect. Implications of these results for the parameterization of orographic convection and pollutant dispersion in large-scale models are discussed.