Impact of turbulent collisions on cloud development Ryo Onishi and Keiko Takahashi Earth Simulator Center (ESC), Japan Agency of Marine-Earth Science and Technology (JAMSTEC)
Turbulent collision kernel model
turbulent collision kernel model K c ( ) 2 r, r, l, u',re 2 π R w g( ) 1 2 η λ = r R R : collision radius (=r 1 +r 2 ) w r : radial relative velocity at contact g(r) : radial distribution function at contact St=0 St=0.2 St=1 St=4
DNS of particle suspended steady isotropic turbulence 2πL 0 4096 particles L 0 [m] Grid points 0.005 64 3 0.01 96 3 Air turbulence : Re λ =36~83 ω + ( U ) ω = ( w ) U + ν 2 ω+ F t periodic BC (quasi-spectral method) forcing Water droplets : St=0.1~200 dv p U ( x p ) v p = dt τ p (Lagrangian tracking method)
construction of g(r) model
Model for g(r) in monodisperse system 100 10 y 1 y 2 DNS y 1, y 2, y 3 with smoothing functions g(r) - 1 1 y 3 0.1 0.1 1 10 100 St
g(r)-1 (=R g ) model results
Collision kernel model for monodisperse particles K c 2 = 2π R w g( R) (< Wr > is from Wang. et al. 2000) r 150 100 DNS (Re λ =54) Wang et al. (Re λ =54) present model (Re λ =54) K c / λr 3 50 0 0.1 1 10 100 St
Collision kernel model for bi-disperse particles Turbulent contribution kernel g(r) for bi-disperse system g(r) for mono-disperse system in Zhou et al. (2001) Total kernel (turbulent & gravitational contributions)
Collision kernels for bi-disperse particles K c,total (r 1,r 2 ) / λ R 3 400 300 200 100 Re λ =44.3 DNS present model hydrodynamic model DNS results present model Hydrodynamic kerne 0 0 50 100 r 2 [μm] ( r 1 =30 [μm] )
Turbulent collision impact on cloud development
Multi-Scale Simulator for the Geoenvironment (MSSG) Applicable to global, regional and local scales seamlessly Ying-Yang grid for globe Consists of 3 modes; atmos. / ocean / coupled Highly optimized for the Earth Simulator (ES) global scale 1~10 km mesoscale 100m~5km urban scale 1m~100m
MSSG-Bin method (Hybrid-Bin method) Bulk method spectral Bin method for liquid water, and conventional Bulk method for solid water AP MSSG-Bin method 雲 water droplet 雨 Reisner et al. (1998) n p (r) mr [ kg/(m 3 unit ln r)] 0.02 0.01 size distributions 0.00 10 100 r [μm] mass doubling, 33 bins
Collision kernel models Hydrodynamic (Gravitational) Collision Kernel Model (No turbulent collisions) K c 2 ( r1, r2 ) = πr V ( r1 ) V ( r1 ) ( R : collision radius (=r 1 +r 2 ), V :settling velocity ) our Turbulent Collision Kernel Model (with turbulent collisions) K c 2 ( r1, r2, lη, u',reλ ) = 2πR wr g( R) w r : radial relative velocity at contact (Wang et al. 2000) g(r) : radial distribution function at contact (Onishi p.h.d thesis 2005)
Algorithm for Explicit turbulent collision calculation MSSG Dynamics compressive N-S eq. (static Smagorinsky model) etc Flow turbulent statistics ε ε = SGS ( Cs Δ) 2 S 3 u' Δ S / 2 l η = ( ν 3 / ε ) 1/ 4 15 l = ν λ u' ε u'lλ Re λ = ν Microphysics(Bin method) collision growth condensation growth etc K c ( r, r2, l η, u',re 1 λ )
RICO model intercomparison Initial data from Rain In Cumulus over the Ocean field campaign by GCSS Practical for investigating cloud microphysical processes white:q c >1e-5, blue:q r >3e-6 24-hour-simulation with MSSG-Bulk method 12.8 12.8 4km domain, Δ H =100m, Δ z =40m, Periodic B.C., 24 hours integration
extra 1-hour simulation for visualization (MSSG-Bulk) 1-hour-simulation with MSSG-Bulk method
RICO intercomparison participants model microphysics timestep [s] duration for 24h integration [h] MESO-NH 1st moment, Kessler 1 440 SAM 1st moment 2 JAMSTEC 1st moment 1 Utah 1st moment? EULAG 1st moment? 2DSAM 1st moment? DALES 2nd moment, Nc fixed 1 UCLA 2nd moment, Nc fixed ~1.5 WVU 2nd moment, Nc fixed ~1.65 COAMPS 2nd moment? 320 68??? 190 UKMO 2nd moment, Nc fixed ~0.7 500 RAMS 2nd moment+ccn 2 weeks 2DHARMA bin model (33) 3.5 10,000 RAMS@NOAA bin model (33) 2 5,400 SAMEX bin model?? MSSG-Bulk 1 st moment (2 nd for ice), Nc fixed ~2.2 250 MSSG-Bin bin model (33) ~2.2 1,750 80 72? *yellow: Bulk models * blue: Bin models 3 bin models & 12 bulk models Ref: Margreet van Zanten, GCSS-GPCI/BLCI-RICO Workshop, NY, 2006
Mixing Ratios of Liquid Water (q l ) and Rain Water (q r ) (fresh AP on ground surface. CCN recycling and break-up are not included.) MSSG-Bin with Turb. Kc graphs from GCSS-BLC wg, RICO<http://www.knmi.nl/samenw/rico/> Significant impact hydr&turb_case1-4.x33 of turbulent collisions is shown.
surface precipitation in time Surface Precipitation [W/m2 80 70 60 50 40 30 20 10 0 4hour averaging MSSG-Bin with KcHydr MSSG-Bin with KcTurb 0 5 10 15 20 25 Time [hour]
Concluding Remarks MSSG-Bin model with turbulent collision model reveals the importance of turbulent droplet collisions in cloud development. In RICO (shallow Cu) case, turbulent collisions significantly weaken the updraft, consequently decrease turbulence intensity due to quick removal of water lower the cloud height shorten the rain initiation time