A Comparison of two Mellor-Yamada-based PBL schemes in Simulating a Hybrid Barrier Jet

The coastal mountains of southeastern Alaska frequently produce intense barrier jets with strong turbulence, which can produce wind gusts sufficiently strong to cause damage to residential and commercial property. There have been few studies on the turbulence within barrier jets and even fewer assessments of the ability of contemporary planetary boundary layer (PBL) schemes to accurately simulate the structure of the turbulent kinetic energy (TKE) within barrier jets. A new PBL scheme, the Mellor-Yamada-Nakanishi-Niino (MYNN), has recently been integrated into the WRF-ARW. This scheme has potential to help reduce some of the common biases associated with the Mellor-Yamada-Janjic (MYJ) scheme, such as shallow PBL height and low TKE bias. This study investigates the performance of the WRF-ARW, with focus on the spatial and temporal structure of the TKE and the fluxes of heat and momentum within coastal barrier jets. The WRF-ARW was configured for three domains: 13, 4.33, and 1.44 km, using one-way nesting and 51 full sigma levels in the vertical. Two simulations using the MYJ and the MYNN PBL schemes are compared with measurements sampled by the Wyoming King-air research aircraft during the Southeastern Alaskan Regional Jets (SARJET) field experiment during September – October 2004.

The observed ambient winds were near coast-parallel early in the event, with cold air flowing over slightly warmer sea-surface temperatures. This produced an upward surface heat flux, resulting in buoyancy-generated TKE beneath 250 m ASL. This combined with shear-generated TKE in the surface layer to produce observed TKE of 0.5-2.0 m2 s-2 in the lowest flight leg (250 m ASL). Later in the event, the ambient winds became more southerly and a strong shear layer developed between the easterly gap outflow (at ~300 m ASL) and the weaker southwesterly flow aloft (~1000 m ASL). This resulted in a reduction of the local Richardson number to become less than 0.25. The observed maximum TKE increased to ~10 m2 s-2 and became elevated to the top of the gap outflow (400-500 m ASL). The TKE observed in the lowest flight leg remained significant and was still likely dominated by buoyancy production.

Both simulations (in the 1.44 km domain) produced a barrier jet that compared reasonably well to the flight observations. The simulated gap outflow south of the largest coastal mountains had wind speeds ~15 m s-1, while the maximum winds in the coastal jet were ~28 m s-1 at 400 m ASL (500 m ASL) for the MYJ (MYNN), which was about 1-2 m s-1 less than the observed. Early in the event, the simulated maximum TKE of 1-2 m2 s-2 (2-3 m2 s-2) in the MYJ (MYNN) was located at the surface beneath the gap outflow. As the ambient low-level winds rotated to become more southwesterly, the region of enhanced simulated TKE deepened due to increased shear-generated TKE at the top of the gap outflow, producing maximum simulated TKE at 500 m ASL of 2-3 m2 s-2 (8-12 m2 s-2) for the MYJ (MYNN).

Although both schemes performed well overall, the MYNN verified slightly better for temperature, wind direction, and much better for TKE. The structural differences will be detailed as well as the fundamental differences in the formulation of the two schemes. Finally, the effects of TKE advection will also be investigated and comparisons will be made to the simulations without TKE advection.