Previous investigators have found the nocturnal low-level jet to result from inertial turning of mixed layer drag after evening surface decoupling and inertial turning of diurnal thermal circulations associated with pressure gradients over sloping terrain or over variations in surface conditions. This understanding is extended through examination of two-dimensional numerical experiments using the UW-NMS nonhydrostatic mesoscale model and a simple parcel model, and a real-data case is shown.
Spatial and temporal variations and relative magnitudes of terms from the momentum equation in the full-physics mesoscale model have been examined, as has the anticipated time of maximum wind based on inertial turning of the present-time ageostrophic wind (accounting for vertical vorticity in the inertial period). The oscillation sets up early in the day, with a nocturnal maximum already determined by late-morning mixed layer drag (acting through TKE in the model). Thus, the tendency for a nocturnal maximum long precedes evening decoupling and even precedes the afternoon local thermal circulation and associated pressure gradients. Evolution of the daytime pressure gradient gradually increases and delays the magnitude and time of maximum wind. Integrations were extended for several days, with strong sensitivity for the starting time of day and initial state on the first night only.
The basic dynamics were incorporated into a simple parcel model using specified functions for diurnal pressure gradient oscillation, mean synoptic pressure gradient, and time-dependent frictional term in the momentum equation. This allowed extensive exploration of parameter space, including type of drag law, separate starting and ending times of daytime drag and pressure gradient regimes, circular or straight flow, variations in Coriolis parameter, sensitivity to initial winds, and others. The parcel model is more sensitive to drag forcing than to the diurnal pressure gradient oscillation, while the full-physics model shows much stronger wind maximum for modest terrain slope than for no terrain slope (e.g. only drag forcing); this may be due to enhanced drag forcing (stronger subgeostrophic from surface mixing) in the afternoon when the local pressure gradient is enhanced. Most interestingly, increasing synoptic geostrophic wind steadily over time, simulating the effect of parcels moving into a synoptic weather system or a synoptic weather system moving into a low-level jet region, shows a large increase in ageostrophic response, enhancing the low-level jet. Curvature of mean geostrophic flow yields increased diurnal speed oscillation but slightly weaker sensitivity to diurnal pressure gradient forcing.
Examination of a real weather case, using profiler data to corroborate the operational UW-NMS prediction, shows the south high plains pressure gradient changes orientation at night, allowing parcels to zoom into the pressure gradient. After this initial accelaration, trajectories have a long stretch nearly normal to the pressure gradient, hence parcels maintain kinetic energy. The result is by morning, the low-level jet ends up a long distance from where it was forced by the diurnal slope flow. This may be critical to explaining the often-observed strong low-level jet over large parts of the midwest where the terrain slope is too small to account for strong nocturnal wind maxima.