Next

NextThe evening boundary layer: turbulence or no turbulence?

Bas van de Wiel, Ivo van Hooijdonk & Judith Donda in collaboration with:

Fred Bosveld, Peter Baas, Arnold Moene, Harm Jonker, Jielun Sun, Herman Clercx, e.a.

A fruitful symbiosis

~ 1 paper / year

Scope

How much wind needed to keep turbulence going?

The very stable boundary layer: ‘cold’

The weakly stable boundary layer: ‘warm’

Need for reliable process parameterizations

The collapse of turbulence:Practical track -----Formal track

Concept: collapse driven by SEB

Idealistic case: Fresh snow (low heat capacity & conductance)

H=H0 H<H0

GHQdt

TdC n

sv

Here: Qn denoted as H0

Regime 1 Regime 2

Surface Energy Balance:

Non-linear diffusion

dz

dTKH H K decreases with increasing dT/dz

Surface flux defines velocity scale

“Weather lab”: ensembles

• Clear nights: similar radiative forcing

• “Wind strength decides on regime”

• Classes 40m wind:-0.5-1.0 m/s-1.0-1.5 m/s-2.0-... etc.

Temperature inversion:T(z)-T(1.0 m)

Can we predict critical wind ?

Turbulence ‘hockey stick’

Can we predict critical wind ?

10m 20m 40m 80m

Predicting minimum wind speed

-for given wind a flux maximum is found

vice versa

-for given demand (Qn-G) characteristic speed is found: Umin

After some calculations……

3/1

202

0min ))(ln(

4

27)(

zzz

c

GQgzU obsobs

p

nobs

The minimum wind speed for sustainable turbulence

Correction for soil heat

parameters Radiative Loss

Height-independent regime-classification

Height-independent regime-classification

Comparison with classical parameters

Stability indicators

3*0

/

u

cHgz

L

z p

Scaling based on fluxes:

Scaling based on gradients:

Combined scaling:

Also combines knowledge flow AND boundary condition

LzzHu /];;[ *

RizTzU ]/;/[ 20 )/(

/

zU

zTgRi

min0 /];;/[ UUzHzU min

/

/

zU

zU

Or formally:

“Shear Capacity”];;/[ zGQzU n

Problem with classical stability parameters

Comparison with classical parameters

Formal track

Step 1: Direct Numerical Simulation of regime

transition

Formal track

Step 2: Analogy with local similarity closure

Formal track

Step 3: analysis, prediction

-local scaling, gradient form

-model independent scaling:

-derivation from TKE-equation

min/

/

zU

zUSC

3/1

220

0min

1/

zc

HgzU

p

prediction ‘normalized hockey stick’

min

z

U

z

U

2m

laminarturbulent

Step 4: validation

Conclusion

• Shear Capacity (U/Umin) compares transport

capacity flow to flux demand at surface

• Prediction idealized configurations & observed

reality

Details:Van Hooijdonk et al. (2014; J.A.S. Submitted)Donda et al. (submission June 2014)

Outlook

• Parameterisation Forecast models

• DNS/LES/RANS simulations

• Other climatologies: Fluxnet – data

Thank you for your attention

What’s the use??

In practice (Louis 1979)

Physically preferable

-Non-physical curve aimed to enhance mixing in the very stable regime only

-But.....

-it causes too much mixing in the well-behaved, weakly stable case as well....!

regime based enhanced mixing

)()1()()( RiFRiFRiF LOUISMO

Continuous turbulentVery stable

Extra: relation to tke budget

Note: SC=shear capacity ~ U/Umin

Strong wind: steady balance

In the following: blue/green points

Weak wind: no balance initially

In the following: red points

Maximum found in observations

20 )( U

Tz

gRb

Here: z=40 m

Tendency for regeneration

Sensibele warmteflux

Instorten turbulentie tgv sterke koeling

Concept