Post on 29-Jan-2016
description
Stephan de Roode (KNMI)
Entrainment in stratocumulus clouds
4 6 8 10
total specific humidity [g/kg]
0 0.5
liquid water content [g/kg]
284 288 292 2960
100
200
300
400
500
600
700
800
temperature [K]
cloud top
cloud base
stratocumulus vertical structure
290 295 300
virtual potential temperature [K]
θv =θ 1+0.61q−ql( ) θl ≈θ−Lvcp
ql
liquid water potential temperature [K]
284 288 292 2960
100
200
300
400
500
600
700
800
temperature [K]
cloud top
cloud base
stratocumulus: vertical structure
Key questions
• How well is stratocumulus represented in models?
• Entrainment
- what is it?
- why important?
- how parameterized?
• Boundary-layer mixing and cloud liquid water path
- perfect boundary-conditions, perfect cloud structure?
• FIRE I observations revisited
- a different view on entrainment
ISCCP stratocumulus cloud climatology
ECMWF RE-ANALYSIS shortwave radiation errors
GCSS intercomparison cases
• Stratocumulus case based on observations (FIRE I)
• Prescribe
- initial state
- large-scale horizontal advection
- large-scale subsidence rate
• Simulation of diurnal cycle
- 1D versions of General Circulation Models
- Large-Eddy Simulation Models (LES)
GCSS intercomparison cases
0 5 10 15-10
-5
0
Δθl [ ]K
Δq
t
[ / ]g kg
03 ( )ASTEX RF EUCREM
( )FIRE I EUROCS
01 ( )DYCOMS II RF GCSS
initial jumps for three
GCSS stratocumulus cases• Stratocumulus case based on observations (FIRE I)
• Prescribe
- initial state
- large-scale horizontal advection
- large-scale subsidence rate
• Simulation of diurnal cycle
- 1D versions of General Circulation Models
- Large-Eddy Simulation Models (LES)
GCSS FIRE I intercomparison participants
Fine-scale turbulence models [Large-Eddy Simulation Models (LES)] : Δx=Δy=50m, Δz=10m1. IMAU - Peter G. Duynkerke, Stephan de Roode, M. C. van Zanten and P. Jonker2. MPI - Andreas Chlond, Frank Müller, and Igor Sednev3. WVU - David Lewellen 4. INM - Javier Calvo, Joan Cuxart, Dolores Olmeda, Enrique Sanchez 5. UKMO - Adrian P. Lock 6. NCAR - Chin-Hoh Moeng (NCAR)
1D versions of General Circulation Models [Single-Column Models (SCM)]1. LMD - Sylvain Cheinet 2. MPI - Andreas Chlond, Frank Müller, and Igor Sednev3. Meteo France I - Hervé Grenier4. Meteo France II - Jean-Marcel Piriou5. ECMWF - Martin Köhler6. CSU - Cara-Lyn Lappen7. KNMI - Geert Lenderink8. UKMO - Adrian P. Lock9. INM - Javier Calvo, Joan Cuxart, Dolores Olmeda, Enrique Sanchez
3D results from Large-Eddy Simulation results -The cloud liquid water path
Local time [h] LWP [g/m2] SWnet,sfc [W/m2]
night-time 0100 ≤ t ≤ 0400 156 ± 11
daytime 1100 ≤ t ≤ 1400 69 ± 20 551 ± 104
0
50
100
150
200
250
0 8 16 24 32 40 48
MMobs
obs
IMAU
MPI
UKMO
INM
NCAR
WVU
LWP [ g m
-2 ]
local time [hours]
What is entrainment?Why is entrainment important?
Entrainment- mixing of relatively warm and dry air from above the inversion into the cloud layer- important for cloud evolution
3D results from Large-Eddy Simulation results -Entrainment rates
Local time [h] LWP [g/m2] SWnet,sfc [W/m2] entrainment rate [cm/s]
night-time 0100 ≤ t ≤ 0400 156 ± 11 0.58 ± 0.08
daytime 1100 ≤ t ≤ 1400 69 ± 20 551 ± 104 0.36 ± 0.03
0.2
0.4
0.6
0.8
0 8 16 24 32 40 48
IMAUMPIUKMOINMNCARWVUmean
local time [hours]
Boundary-layer representation
w'ψ' =−Kψ∂ψ∂z
w'ψ' =Mc ψc −ψ( )
1D results from General Circulation Models -The cloud liquid water path (LWP)
0
50
100
150
200
250
0 8 16 24 32 40 48
MMobsobsKNMI RACMOINM MESO-NHINM HIRLAMCSU MassfluxLMD GCMMPI ECHAMARPEGE Clim.UKMOARPEGE NWPECMWF
LWP [ g m
-2 ]
local time [hours]
Single Column Model liquid water path results very sensitive to
• entrainment rate
• drizzle parameterization
• convection scheme (erroneous triggering of cumulus clouds)
Key questions
• How well is stratocumulus represented in models?
• Entrainment
- what is it?
- why important?
- how parameterized?
• Boundary-layer mixing and cloud liquid water path
- perfect boundary-conditions, perfect cloud structure?
• FIRE I observations revisited
- a different view on entrainment
The clear convective boundary layer (CBL) -Entrainment scaling from observations
Entrainment rate we scales as
• A ≈ 0.2
• H boundary-layer height
• (g/θ0) Δθv buoyancy jump across the inversion
• w* convective velocity scale: vertically integrated buoyancy flux
we=A w*
3
gθ0
H Δθv
Buoyancy flux in stratocumulus
convective velocity scale w* depends on entrainment rate we
w'θv'T =−weΔθv,sat
-0.04 -0.03 -0.02 -0.01 0 0.01 0.020
200
400
600
800
virtual potential temperature flux <w' θv> [ / ] ' Km s
entrainment
longwave radiative cooling
condensation
Solve entrainment rate
we=A w*
3
gθ0
H Δθv we =
2.5AWNE
Δθv +2.5A T2Δθv,dry+T4Δθv,sat( )
solve for entrainment rate
we __________forcing WNE
"jumps"
Solve entrainment rate
we=A w*
3
gθ0
H Δθv we =
2.5AWNE
Δθv +2.5A T2Δθv,dry+T4Δθv,sat( )
we __________forcing WNE
"jumps"
-0.04 -0.02 0 0.02 0.040
0.2
0.4
0.6
0.8
1
<w'θv>'
WE
WNE
<w'θv>'
solve for entrainment rate
Solve entrainment rate
we=A w*
3
gθ0
H Δθv we =
2.5AWNE
Δθv +2.5A T2Δθv,dry+T4Δθv,sat( )
we __________forcing WNE
"jumps"
-0.04 -0.02 0 0.02 0.040
0.2
0.4
0.6
0.8
1
<w'θv>'
WE
WNE
<w'θv>'
solve for entrainment rate
Solve entrainment rate
we=A w*
3
gθ0
H Δθv we =
2.5AWNE
Δθv +2.5A T2Δθv,dry+T4Δθv,sat( )
we __________forcing WNE
"jumps"
-0.04 -0.02 0 0.02 0.040
0.2
0.4
0.6
0.8
1
<w'θv>'
WE
WNE
<w'θv>'
solve for entrainment rate
Stability jumps
Stability jumps
Stability jumps
Entrainment parameterizations for stratocumulus -Results based on LES results
• Nicholls and Turton (1986)
• Stage and Businger (1981) Lewellen and Lewellen (1998) VanZanten et al. (1999)
• Lock (1998)
• Lilly (2002)
we = 2.5AWNE
Δθv +2.5A T2Δθv,dry +T4Δθv,sat( )
we = 2.5AWNE
Δθv,NT +2.5A T2Δθv,dry+T4Δθv,sat( )
we = AWNE
T2Δθv,dry+T4Δθv,sat
we = 2AAL WNE +αtAwΔFL / ρcp
Δθv
we = ADLWNE,DL
Δθv,DL +ADL L 2Δθv,dry+L4Δθv,sat( )
• Based on observations of clear CBL
Sensitivity of entrainment parameterizations to inversion jumps
observations from ASTEX Flight A209__________________________________cloud base height = 240 mcloud top height = 755 msensible heat flux = 10 W/m2
latent heat flux = 30 W/m2
longwave flux jump = 70 W/m2
max liquid. water content = 0.5 g/kgLWP = 100 g/m2
Compute entrainment rate from parameterizations as a function of inversion jumps
Entrainment rate [cm/s] sensitivity to inversion jumps
Entrainment rate [cm/s] parameterizationsof observed cases
Parameterization Case Observed
Moeng Lock Lilly Nicholls-Turton
Lewellen
North Sea NT620 0.55 0.50 0.13 0.30 0.30 0.33
North Sea NT624 0.56 0.76 0.28 0.55 0.66 0.61
ASTEX A209 0.9 ± 0.3 1.23 0.42 0.86 1.06 0.97
ASTEX RF06 1.0 ± 0.6 1.24 0.48 1.04 1.31 1.33
DYCOMSII RF01 0.38 ± 0.10 0.72 0.69 0.62 0.60 0.64
FIRE I 0.58 ± 0.08
(mean LES)
0.57 0.16 0.37 0.35 0.50
high low
Entrainment results mirror the LES results where they are based on
• Turbulent flux at the top of the boundary layer due to entrainment:
("flux-jump" relation)
• Top-flux with K-diffusion:
Entrainment parameterizations -
Implementation in K-diffusion schemes
w'ψ'T =−weΔψ
w'ψ'T =−KψΔψΔz
⇒ Kψ =weΔz
Key questions
• How well is stratocumulus represented in models?
• Entrainment
- what is it?
- why important?
- how parameterized?
• Boundary-layer mixing and cloud liquid water path
- perfect boundary-conditions, perfect cloud structure?
• FIRE I observations revisited
- a different view on entrainment
Compute eddy- diffusivity
coefficients from FIRE I
LES
Kψ =−w'ψ'
∂ψ / ∂z
288 292 296 300 3040
200
400
600
800
1000
Liquid water potential temperature θl [ ]K
0.005 0.008 0.010
200
400
600
800
1000
total water content [g/kg]
-0.04 00
200
400
600
800
1000
<w'θl> [ / ]' mK s
0 100
1.5 10-5
0
200
400
600
800
1000
<w'qt'> [(g/kg) m/s]
K-coefficients from FIRE I LES
Kψ =−w'ψ'
∂ψ / ∂z
0 100 200 300 400 500 6000
100
200
300
400
500
600
K_ θl
_K qt
[Eddy diffusivity coefficient m2 / ]s
Importance of eddy-diffusivity coefficients on internal boundary-layer structure
• Change magnitude K profiles
• Compute vertical profiles θl and qt from integration
0 200 400 600 800 10000
100
200
300
400
500
600
Kref
x 0.2
Kref
x 0.5
Kref
Kref
x 2
Kref
x 5
Eddy diffusivity K [m2/s]
∂ψ∂z
=−w'ψ'Kψ
same
change
Total water content profiles for different K-profiles but identical vertical flux
8 8.5 9 9.5 100
100
200
300
400
500
600
Kref
x 0.2
Kref
x 0.5
Kref
Kref
x 2
Kref
x 5
Kref
x inf
total water content [g/kg]
Liquid water content profiles for different K-profiles
K factor LWP [g/m2]
0.2 2
0.5 52
1.0 79
2.0 94
5.0 103
109
Magnitude K-coefficient in interior BL important for liquid water content!
0 0.1 0.2 0.3 0.4 0.5 0.6 0.70
100
200
300
400
500
600
Kref
x 0.2
Kref
x 0.5
Kref
Kref
x 2
Kref
x 5
Kref
x inf
liquid water content [g/kg]
Key questions
• How well is stratocumulus represented in models?
• Entrainment
- what is it?
- why important?
- how parameterized?
• Boundary-layer mixing and cloud liquid water path
- perfect boundary-conditions, perfect cloud structure?
• FIRE I observations revisited
- a different view on entrainment
FIRE I stratocumulus over the Pacific Ocean -
Aircraft lidar observations of cloud-top height
0
200
400
600
800
1000
0 10 20 30 40 50 60 70
horizontal distance [km]
Thermodynamic structure of clear air above cloud top depressions
clear air value
mean in-cloud value
Evaporation of cloud top by turbulent mixing horizontal winds
vertical velocity
liquid water content
liquid water potential temperature
total water content
turbulence turbulence
evaporation
12 km
Observations of moist and cold layers on top of stratocumulus
Entrainment mixing scenario
Conclusions
• Entrainment parameterizations
- extrapolation of Large-Eddy Simulation results
- considerable differences
different heat and moisture budgets
• Cloud liquid water path and K-diffusion turbulence schemes
- different solutions for identical surface and cloud-top fluxes
different albedo
• Entrainment observations
- may induce the formation of moist layers above cloud top
opposes general view on the entrainment process
stability jumps