Mesoscale Applications for Microscale Model? Robert Plant (With thanks to Sue Gray) (Also K. Browning, C. Morcrette, A. Agusti- Panareda, M. Pizzamei, M. Valente) Microscale Model Meeting 16th December 2003

Blatant but Shameless Advertisement We are planning a UWERN meeting to review the current state of mesoscale modelling in the UK. The probable date is… Friday 13th February, 2004.

Example Meso/Microscale Projects Studies that might… benefit from the microscale model. require certain things from the microscale model. Boundary layer ventilation. Slantwise convection. The sting jet.

Boundary Layer Ventilation How much polluted air is transported to free troposphere by: warm and cold conveyor belts stacked slantwise convective circulations upright convection shear instability and turbulent mixing When and where are these processes important? Bader et al. (1995) There are several physical processes that can ventilate BL air into the free troposphere in frontal regions with different spatial and time scales: The large-scale flow constituted by warm and cold conveyor belts. Advection by the warm conveyor belt (ahead of the cold front) can transport pollutants from the (top) of the boundary layer (ie. About 800 mb) to the top of the free troposphere (ie. About 300 mb). Embedded in this large-scale flow, there is upright convection and Slantwise convection across the fronts which can be stacked in multiple layers as shown in this vertical cross section across a cold front by Browning et al (2001). Finally on the smaller scales, there is shear instability and turbulent mixing. Shear instability and turbulent mixing Height (km) Range (km) (Agusti-Panareda, Gray, Belcher) Browning et al. (2001)

UM domain-integrated free-tropospheric tracer (Tracer lifetime is 1hr) Advection + mixing Advection Advection + mixing + convection Advection + convection The contribution of the different mechanisms to the BL ventilation is very different when the tracer lifetime is 1 hour. These plots show also the evolution of the total tracer in the FT (red) and in the BL (black) for the different mechanisms of BL ventilation but in these plots all the tracers have a lifetime of 1 hour. We can see that advection is not transporting any significant amount of tracer in the BL, where as the amount of tracer in the BL increases because of the continuous source until it reaches a steady state. When the tracer is transported by advection and mixing the total amount of tracer in the FT increases very little, and the diurnal cycle becomes more pronounced with a maximum FT tracer over night when the BL collapses and a minimum during daytime when the tracer is reincorporated in the BL as its depth grows. When the tracer is transported by advection and convection there is also a diurnal cycle in the FT tracer but now the maximum is during the day when convection occurs (instead of night as in the previous tracer). Convection has the effect of exporting more tracer into the FT. The interesting thing is when all the mechanisms (advection, convection and mixing) can transport the tracer. The transport to the FT by mixing and convection together is much larger than the sum of the transports by the two separate mechanims. Once mixing has transported the tracer to the top of the BL then convection can export more tracer into the FT (and so the peak of the FT tracer is at mid-day). BL tracer FT tracer

Boundary Layer Ventilation Desirable features for modelling: code for tracers weak numerical diffusion good boundary layer scheme coupling to mesoscale model (UM?) to allow for representation of fronts, conveyor belts etc.

Upright convection  Driven by gravitational force.  Due to unstable temperature distribution.  Causes vertical motion. q Inertial convection  Driven by centrifugal force.  Due to unstable angular momentum distribution.  Causes radial motion. M

q high q low M Symmetric Instability (SI) (Bennetts and Hoskins, 1979) Parcel can be stable to both upright and inertial convection but unstable to slantwise convection. If parcel perturbed at an angle between the M and  lines, it will accelerate from its initial position on a slantwise trajectory. q high q low M

Slantwise Convection What are the interactions between upright and slantwise convection in frontal regions? What are the relative contributions of the release of conditional symmetric instability and M adjustment to the slantwise component? Difficult to model. Some success with idealized LEM experiments…

Slantwise convection simulated by LEM V (m/s) RH (%) (Pizzamei, Gray)

Slantwise Convection Desirable features for modelling: explicit treatment of upright convection good microphysics good vertical resolution coupling to mesoscale model (UM?) to allow for fully realistic simulations.

Sting Jet

(Browning, Valente, Gray) Sting Jet What are the dominant mechanisms leading to the intense winds in the dry slot region ahead of the hooked tail of the cloud head? What are the characteristics of cyclones exhibiting sting jets? What determines the locations of the regions of intense winds and their lifetime and propagation relative to the cyclone? (Browning, Valente, Gray)

Difficult to model in the UM. Sting Jet Desirable features for modelling: good microphysics good vertical resolution coupling to mesoscale model (UM?) to allow for fully realistic simulations. Difficult to model in the UM.

Use of Microscale Model Some progress is being made with existing high-resolution tools. (LEM/CRM, idealized UM) Will the microscale model be a better vehicle for studying these mesoscale problems? (Should it be?) If so, how will it be better?