Idealized large-eddy simulations of nocturnal low-level jets over subtropical desert regions and implications for dust-generating winds

摘要

Nocturnal low-level jets (LLJs) are maxima in the wind profile, which often form above the stable nocturnal boundary layer. Over the Sahara, the world's largest source of mineral dust, this phenomenon is of particular importance to the emission and transport of desert aerosol. We present the first ever detailed large-eddy simulations of dust-generating LLJs. Using sensitivity studies with the UK Met Office large-eddy model (LEM), two key controls of the nocturnal LLJ are investigated: surface roughness and the Coriolis force. Functional relationships derived from the LEM results help to identify optimal latitude-roughness configurations for a maximum LLJ enhancement. Ideal conditions are found in regions between 20 and 27°N with roughness lengths >0.0001 m providing long oscillation periods and large jet amplitudes. Typical LLJ enhancements reach up to 3.5 m s for geostrophic winds of 10 m s. The findings are largely consistent with results from a theoretical LLJ model applied for comparison. The results demonstrate the importance of latitude and roughness in creating regional patterns of LLJ influence. Combining the functional relationships with high-resolution roughness data over northern Africa gives good agreement with the location of morning dust uplift in satellite observations. It is shown that shear-induced mixing plays an important role for the LLJ evolution and surface gustiness. With decreasing latitude the LLJ oscillation period is longer and, thus, shear-induced mixing is weaker, allowing a more stable nocturnal stratification to develop. This causes a later and more abrupt LLJ breakdown in the morning with stronger gusts, which can compensate for the slower LLJ evolution that leads to a weaker jet maximum. The findings presented here can serve as the first step towards a parametrization to improve the representation of the effects of nocturnal LLJs on dust emission in coarser-resolution models.