ann-sofi smedman uppsala university uppsala, sweden · ann-sofi smedman uppsala university uppsala,...

Effects of transfer processes on marine atmosphericboundary layer

orEffects of boundary layer processes on air-sea

exchange

Ann-Sofi SmedmanUppsala University

Uppsala, Sweden

Swell

Long wavescreateform drag

Upward momentum flux

Quasi-frictional decouplingat the surface

Turbulence resembles free convection conditions

Ocean

MABL

Effect of transfer process on MABL

Effect of MABL process on air-sea exchange

MABL

Detached eddiesimpinging on the surface

Increased small scaleheat flux

Increased CHNand CEN

Ocean

UVCN regime

In models air-sea exchangeis described through Monin-Obukhov similarity

theorya theory which is well tested over land

but

is it valid over the ocean?

Large heat capacity

Small diurnal variation

Roughness (waves)increasing with

increasing wind speed

What are the differences ?

10002200

2 buoys (temp, waveheight , dir. and CO2 Footprint area

Tower 30 mTemp and wind profiles 5 levels

Turbulence 3 levelsHumidity and CO2 at 2 levels

Long term measurements in the Baltic Sea

Turbulence, wind speed, temperature and waveparameters

Short term experiments with RV Aranda and ASIS buoy

In the mean, excellent agreement between tower and ASIS

Data set from Agile experiment in lake Ontario 1994-95

A 15- m research vessel Agile was equipped to measure waves and turbulence with

• Sonic R2A• Thermocouple• Licor• Wave staff array

Measurements were performed at 7.8 m during two autumn periods

Growing sea (slow waves)Unstable stratification

• Monin-Obukhov similarity theory is validbut

roughness length, zo, is a function of wave age

0.04 0.05 0.06 0.0710

−6

10−4

10−2

Growing sea

u*/c

p

z 0/σ

Swell (long waves traveling faster than the wind)

andunstable stratification

• A quite different type of boundary layer develops

The swell driven BL is characterized by

• Small or positive momentum flux• Low level jet• M-O scaling does not hold• The vertical extent can be global• Turbulence spectra resemble free convection conditions• Swell represents reverse atmospheric-ocean coupling

−0.04 −0.035 −0.03 −0.025 −0.02 −0.015 −0.01 −0.005 0 0.0050

5

10

15

20

25

30

u’w’ (m2s−2)

heig

ht (

m)

C2 C

p/U

8=1.79C3

CP/U

8=1.61Shearing stress

C1C

p/U

8=4.72

F1C

p/U

8=4.66

F2C

p/U

8=1.73

Small momentum flux

uw x 1000

positive uw

Aircraft measurements

Tower measurements

Positive momentum flux

LES simulations Peter Sullivan

0.5 0.7 0.9 1.1 1.3 1.5 1.7 1.9 20

5

10

15

20

25

30

35

U (ms−1)

heig

ht (

m)

Mean wind profiles

Cp/U

8 = 4.72 C

p/U

8 = 4.60

Case C1 Case F1

Low level jet

2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 70

5

10

15

20

25

30

35

U (ms−1)

heig

ht (

m)

Mean wind profiles

Cp/U

8 = 1.79 C

p/U

8 = 1.61 C

p/U

8 = 1.73

Case C2 Case C3 Case F2

0 2 4 6 8 1010

0

101

102

103

104

U (m/s)

z (m

)

Pibal tracking 03090610−23 LST

The vertical extent is global

−1 0 1 2 3 4 5 6 70

5

10

15

20

25

30

φm

heig

ht (

m)

φm

= (1−15z/L)−1/4for L= −30 m

C1Eq.C2F1F2C3

Φm calculated for five swell cases

Cp/U8=4.7

Cp/U8=1.6

Turbulent kinetic energy budget

ερ

θ +∂′∂

+∂

′∂⋅+′−

∂∂⋅′′=

zew

zwpw

Tg

zUwu

oo

10

P B Tp Tt D

−7 −6 −5 −4 −3 −2 −1 0 1 2 3 4 5

x 10−3

0

5

10

15

20

25

30

TKE (m2s−3)

heig

ht (

m)

P B Tp Tt D

Growing sea

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 10

0.005

0.01

0.015

0.02

0.025

0.03

0.035

0.04

0.045

0.05

Frequency − n (Hz)

Ene

rgy

− S

(n)

(m2 s−

1 )

Growing sea case

npeak

= 0.42 Hz

Turbulent energy budgetGrowing sea

−3 −2.5 −2 −1.5 −1 −0.5 0 0.5 1 1.5 2 2.5 3

x 10−4

0

5

10

15

20

25

30

TKE (m2s−3)

heig

ht (

m)

Case F1

BTp P Tt ε

Turbulent energy budgetcp/U=4.7

0.5 0.7 0.9 1.1 1.3 1.5 1.7 1.9 20

5

10

15

20

25

30

35

U (ms−1)

heig

ht (

m)

Mean wind profiles

Cp/U

8 = 4.72 C

p/U

8 = 4.60

Case C1 Case F1

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 10

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

Frequency − n (Hz)

Energ

y −

S(n

) (m

2s

−1)

Case F1

nswell

= 0.22 Hz

−5 −4 −3 −2 −1 0 1 2 3 4

x 10−3

0

5

10

15

20

25

30

TKE (m2s−3)

heig

ht (

m)

Case C3

P B Tt Tpε

2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 70

5

10

15

20

25

30

35

U (ms−1)

he

igh

t (m

)Mean wind profiles

Cp/U

8 = 1.79 C

p/U

8 = 1.61 C

p/U

8 = 1.73

Case C2 Case C3 Case F2

Turbulent energy budgetcp/U=1.6

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 10

0.005

0.01

0.015

0.02

0.025

0.03

Frequency − n (Hz)

Energ

y −

S(n

)(m

2s−

1)

Case C3

nswell

= 0.22 Hz

nmin

=0.35 Hz

)2

''''(''

1 2

εθ ++∂∂

+−−=∂∂

pv Tqwz

wTg

wuzU

'')(

wuT

zU p

Tp −=∂∂

dzuT

mUzUz

m

p∫=−5.2

2*

)5.2()(

From the turbulence energy eq,

The effect of Tp can be isolated

4.5 4.7 4.9 5.1 5.3 5.5 5.7 5.9 60

5

10

15

20

25

30

35

U (ms−1)

heig

ht

F2 (based on numerical expressions for Tp)

blue: measured U−curve

red: measured U−curve minus Tp/u*2

Heat fluxes

All turbulent fluxes in models must be expressed as mean variables

)( soH TTU

wC−′′

=θ

Wind speed Air-sea temperaturedifference

Stanton numberHeat flux

In the same way for latent heat, CE

)}/)}{ln(/{ln( 00

2

THN zzzz

C κ=

Reduced to neutral stability

Von Karman constant=0.4

Roughness length Roughness lengthfor temperature

In the same way for latent heat, CEN

Heat fluxes, unstable stratificationΔT>3o

• Traditional parametrization (COARE algorithm) is valid

• No variation with wave age

Heat fluxes, stable stratification

• Low-level jets cause shear suppression leading to decreased fluxes

Heat fluxes, unstable stratification, very close to neutral

Wind speed > 10 m/s and ΔT< 3o

Fluxes are underestimated with COARE algorithm

10−4

10−3

10−2

10−1

100

101

102

10−4

10−3

10−2

10−1

100

101

nSu,

w(n

)/u*2 Φ ε2/

3 nS T(n

)/T*2 Φ ε−1

/3Φ N

T

f=nz/U

Tuw

Unstable stratification -- horizontal rolls

U T

w

10−4

10−3

10−2

10−1

100

101

10−3

10−2

10−1

100

101

f=nz/U

nSu,

w(n

)/u *2 Φ

ε2/3 n

ST(n

)/T

*2 Φε−1

/3Φ

NT

Tuw

Unstable near neutral -- detached eddies

UT

w

10−4

10−3

10−2

10−1

100

101

102

0

0.05

0.1

0.15

0.2

0.25

f=nz/U

nCO

wt(n

)/w

t

Norm. heat flux Östergarnsholm

L=−10 m

L=−42 mL=−140 m

L=−250 m

Normalized heat fluxes

2 4 6 8 10 12 140

0.5

1

1.5

2

2.5

3x 10

−3

U10

CH

NMIUU, L > −150m

data2

R2, 1 m/s bins

MIUU, L <−150m; 0.5<(Tw − T10) <1.0K

MIUU, L < −150m; 1.0 < (Tw − T10) < 1.5K

MIUU, L < −150m; 1.5 < (Tw − T10) < 2.0 K

ΔT<3o

Sonic temp.

Platinum sensor ( o ***)

Heat flux from Östergarnsholm

Latent heat flux, Östergarnsholm

COARE

ΔΤ<2ο

2 4 6 8 10 12 14 16 180.5

1

1.5

2

2.5

3

3.5

4x 10

−3

CH

N

U10

SonicTC

ΔT>3

ΔT>3

0.8

0.9

0.8

0.6 0.7

2.18

0.752.2

2.7

2.0 3.0

2.3 2.0

..

Agile heat flux

thermocouple

sonic

ΔT<3o

2 4 6 8 10 12 140

0.5

1

1.5

2

2.5

3x 10

−3

U (ms−1)

CE

N

2.98 3.04

3.1

3.13.17

3.13

3.2

2.06

2.05

2.32

6.19

6.14

6.22

5.562.26

Agile Licor, ΔT<3

Agile latent heat flux

Solid curves data from Östergarnsholm

2 4 6 8 10 12 14 16 18−0.01

−0.005

0

0.005

0.01

0.015

0.02

0.025

0.03

0.035

0.04Comparison of heat fluxes, differences, Agile−94,95

w´t

´ TC −

w´t

´ S (

m s

−1 K

)

U10

(m s−1)

Agile heat flux: (wt)TC- ( wt) sonic

The unstable very close to neutral regime

• The UVCN regime is defined for ΔT<3o and U>10m/s

• It is found both over land and sea in the lowest region (z<15 m) of the surface layer

• Small scale heat flux is a result of detached eddies originating from the upper part of the surface. This is in contrast to a convective surface layer ruled by horizontal rolls.

Frequency (%) of occurrence of UVCN conditions

ERA-40 reanalysis

Increase (W/m2) in latent heat flux with UVCN regime included

ERA-40 reanalysis

Global mean values of the increase of heat fluxes with UVCN regime included

• Sensible heat flux= 0.8 W/m2

• Latent heat flux (evaporation)= 2.4 W/m 2

• The effect of increased anthropogenic greenhouse gases (2006) on net radiation = 2.63±0.26 W/m 2

Conclusions

Unstable stratification

• M-O similarity theory is only valid for growing waves

During swell the MABL is characterized by

• Low-level jet (LLJ) at 5-10 m height• Constant wind speed above LLJ• Small or positive momentum flux varying with height• Energy is transported from waves to atmosphere by

pressure transport term Tp

Mixed seas

As soon as a small portion of the waves are travelling faster than the wind (mixed seas) MABL is slowly changing from an ordinary boundary layer to a ‘swell dominated ‘ boundary layer and

Monin-Obukhov similarity theory is not valid

During stable stratification wave effects are small

Heat fluxes

• The dynamics of the MABL influence the air-sea exchange of sensible and latent heat.

• When the convective boundary layer is ruled by horizontal rolls the well known COARE algorithm is valid.

• When detached eddies originating from the upper part of the surface layer is present (UVCN regime) the heat fluxes are increased and the Φh-function goes to zero

Why has not the UVCN regime been observed earlier?

• sonic temperature fails to measure temperature fluctuations correctly when σT is small (high wind speed, small ΔT)

• data with small ΔT are often removed from the data set

• measurements are taken at heights above the UVCN layer.

Thanks to• Ulf Högström, Erik Sahlee, Anna Rutgersson,

Hans Bergström, Cecilia Johansson, MIUU

• Will Drennan, RSMAS, Miami, Kimmo Kahmaand Heidi Pettersson, FMRI, Helsinki

• Phd-students: Andreas Karelid , Björn Carlsson, Maria Norman, Alvaro Semedo

• Former Phd-students: Cecilia Johansson Birgitta Källstrand, Anna Rutgersson, Anna Sjöblom, Xiaoli Guo Larsen