dr jian zhong - modelling the neighbourhood-scale dispersion of ultrafine particles- dmug17

Modelling the neighbourhood-scale dispersion of ultrafine particles

using WRF-LES

Dr Jian Zhong, Dr Irina Nikolova, Dr Xiaoming Cai, Prof. A. Rob MacKenzie, and Prof. Roy M. Harrison (PI)

Dispersion Modellers User Group (DMUG) conference, 6 April 2017, London

School of Geography, Earth & Environmental Sciences, University of Birmingham

Modelling the neighbourhood-scale dispersion of ultrafine particles using WRF-LES

Introduction

Methodology

Results and Discussion

Future modelling work

Outline

Conclusions


Introduction

FASTER: Fundamental Studies of the Sources, Properties and Environmental Behaviour of Exhaust Nanoparticles from Road Vehicles


Introduction

Ultrafine particles (UFPs diameter < 100 nm, nanoparticles) are respirable particles. Due to their small sizes, UFPs can accumulate in the lungs and penetrate cells/tissue, and cause health effects.

Mühlfeld et al (2008)


Methodology

UFPs implementation in WRF-LES Number of nanoparticle compositions: 18 (1 non-volatile core

and 17 representative Semi-Volatile Organic Compounds, SVOC).

Number of sectional size bin: 15 (6 nm- 500 nm). UFPs mass (15x18) and gas concentrations (18) are tracked

(advection and diffusion solved by WRF-LES). Condensation and evaporation processes of UFPs are

coupled. UFPs number (15) is diagnosed/calculated based on UFPs

mass and the sectional diameters. 1 tracer for passive scalar (plume). Total number of tracers: 304 (=15x18+18+15+1).


Methodology

Condensation/evaporation: driven by the difference between the partial pressure of a gas and its saturation vapour pressure over a particle surface.

Jacobson (2005)

(a) Condensation (b) Evaporation


Methodology

Condensation/evaporation rate estimation: the flux between gas phase and particles for composition i in each size bin. 𝑑𝑚 𝑖

𝑑𝑡 =𝑎𝐹𝑆2𝜋 𝐷𝑝𝑀𝑖𝐷

𝑅𝑇 (𝑒𝑖∞−𝑎𝐾

𝑖 𝑒𝑖𝑣𝑎𝑝)

aFS: Fuchs-Sutugin correction factorDp: Particle diameterMi: Molar mass of component iD: Vapour diffusivity R: Universal gas constantT: Temperatureai

K : Kelvin effect term for component i e

i : Ambient partial pressure of component ievap

i: Saturation vapour pressure of component i


Methodology

Stiff system: various time scale of condensation/evaporation processes for each composition (N-Alkanes from C16 to C32).

Dynamical approach

Equilibrium approach

Published saturation vapour pressure data for the less volatile compounds range over almost five orders of magnitude.

Range of vapour pressures from C16 to C32 is up to 14 orders of magnitude, depending on the source of the data.


Methodology

Domain settings

Grid resolution: 50 m x 50 m in x and y direction; Stretched grids in z direction.

Grid number: 63 x 63 x 39. Domain size: 3.15 km x 3.15 km x 2 km. Time step: 0.5 s. Neutral atmospheric boundary layer. Mean wind speed (U): ~ 2 m/s. Periodic conditions for wind and open conditions for UFPs

in x and y directions.


Methodology

WRF-LES settings

Namelist configuration for WRF-LESNamelist option Setting (value)Time-integration scheme (rk_ord)

Runge-Kutta 3rd order (3)

Turbulence and mixing (diff_opt)

Mixing in physical space; full diffusion (2)

Eddy coefficient (km_opt) 3D 1.5 order TKE closure (2)Surface layer option (sf_sfclay_physics)

Monin-Obukhov scheme (1)

Surface heat and moisture fluxes (isfflx)

Use drag from sf_sfclay_physics and heat flux from tke_heat_flux (2)

Turn on UFPs tracers (tracer_opt)

User-defined logical parameter (2)



One-line (street) emission scenario

Horizontal slice from the S-N line emission at k=1 (30 mins)




Averaged vertical slice from the S-N line emission (30 mins)




Downwind particle number distribution from the S-N line emissions (at 30 min)



Idealised street canyon network

Plume




Horizontal slice at k=1 (30 mins)




Averaged vertical slice from 1st (left) and 2nd (right) S-N line emissions (30 mins)




Downwind particle number distribution from 1st (left) and 2nd (right) S-N line emissions (at 30 min)



Realistic street canyon network (potentially)


Future modelling work

Configuration of realistic street canyon network emission patterns into WRF-LES.

Test the dispersion of ultrafine particles under different type of atmospheric boundary layers (e.g. convective or stably stratified).

Develop inflow turbulence generator for WRF-LES under realistic wind conditions and fit for the comparison of ultrafine particle size distributions between the model and field measurements.


Conclusions

The multicomponent aerosol microphysics involving condensation and evaporation processes are stiff.

The WRF-LES model can be coupled with aerosol dynamics to simulate the fate of nanoparticles in urban air.

The dispersion of sized-resolved nanoparticles at the neighbourhood-scale can be revealed by WRF-LES.


Thank you!

Dispersion Modellers User Group (DMUG) conference, 6 April 2017, London