A Lagrangian view of the Stratospheric Mixing

(1)

A Lagrangian view of the Stratospheric Mixing

Simulations with CLaMS

CLaMS – Chemical Lagrangian Model of the Stratosphere

P. Konopka et al.

P.Konopka@fz-juelich.de

http://www.fz-juelich.de/icg/icg-i/www export/p.konopka

Research Centre J ¨ulich, ICG-I: Stratosphere, Germany

(2)

CLaMS-Model

CLaMS - Lagrangian Chemistry Transport Model Potential temperature as vertical coordinate

Horizontal and vertical velocities from meteor. winds (ECMWF) and/or a radiation scheme Lagrangian mixing

Full stratospheric chemistry

Lagrangian particle sedimentation scheme parallelized code (JUMP)

McKenna et al., JGR, 2002, Konopka et al., JGR, 2004, Grooß et al., 2005, ACP TRAJECTORIES

MIXING CHEMISTRY

SEDIMENTATION

(3)

Lagrange versus Euler

Lagrange (irregular grid)

!!""

##$$

% %

% %& &

& &

' ' ' '

( ( ( () )

) )

* *

* *+ + + +

+ + + +

, , , ,

- - - -

. . . .

. . . . / / / / / / /

/ / / / / / /

0 0 0 0 0 0 0

1 1

1 12 2

2 2

3 3 3 3

4 4 4 4

4 4 4 45 5 5 5 5

5 5 5 5 5

6 6 6 6 6

r(t)

mixing (exchange of mass) is under control

Euler

(regular grid)

777777777777777777777777777777777777777777777777777777 888888888888888888888888888888888888888888888888888888

9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9: : : : : : : : : : : : : : : : : : : : : : : : : : :

; ; ; ;

< < < <

< < < < = = = =

= = = =

> > > >

> > > > ? ? ? ?

? ? ? ?

@ @ @ @

@ @ @ @ A A A A A

A A A A A

B B B B B

C C C

C C CD D

D D

D D E E E E

E E E E

F F F

F F F G G G

G G G

H H H

H H H I I I I

I I I I

J J J J

K K

K KLLLLLLLL M M M M

M M M M

N N N

N N N O O O O

O O O O

P P P P

P P P P Q Q Q Q

Q Q Q Q

R R R R

S S S S S

T T T T

T T T T U U U U U

U U U U U

V V V V

V V V V W W W W

W W W W

X X X X

X X X X Y Y Y Y Y

Y Y Y Y Y

Z Z Z Z Z

mixing is determined

(4)

Mixing in CLaMS

Large-scale wind

Small-scale deformations

Filamentation

Mixing

(irreversibility)

(5)

Grid Adaptation ⇒ Mixing

A C B

quasiuniform distribution of air parcels

Delaunay triangulation ⇒ next neighbors

sheared flow

∆t = 6− 24 hours

A C B

D

grid adaptation =

regridding of the deformed grid

⇒ new air parcels

⇒ interpolations (num. diffusion)

⇒ mixing

(6)

Grid Adaptation ⇒ Mixing

A C B

sheared flow

∆t = 6 −24 hours

A C B

D

grid adaptation =

⇒ new air parcels

⇒ mixing

(7)

Grid Adaptation ⇒ Mixing

A C B

sheared flow

∆t = 6 − 24 hours

A C B

D

grid adaptation =

⇒ new air parcels

(8)

Lyapunov Exponent λ

r₀

t=t₀

Consider an air parcel sur- rounded by a small circle of ra- dius r0.

t=t + t r r

+

−

0

After a time ∆t and for sufficiently small values of r0, the circle is deformed into an ellipse with minor and major axes r⁻ and r+

Definition: (Lyapunov exponent)

λ^± = ± 1

∆t ln r^± r0

for sufficiently small ∆t and r0

Incompressible flows ⇒ (r₀² = r−r+) ⇒ λ− = λ+

(9)

Lyapunov Exponent λ

r₀

t=t₀

Consider an air parcel sur- rounded by a small circle of ra- dius r0.

t=t + t r r

+

−

0

After a time ∆t and for sufficiently small values of r0, the circle is deformed into an ellipse with minor and major axes r⁻ and r+

Definition: (Lyapunov exponent)

λ^± = ± 1

∆t ln r^± r0

for sufficiently small ∆t and r0 2

(10)

2d-CLaMS: Dynamically Adaptive Grid

Before the advection step

A C B

Determine nearest neighbors (e.g. for point A), r0 - mean distance between air parcels

Set the critical Lyapunov exponent λ_c and the time step ∆t (free parameter) Define r^c_±

r±^c = r0 exp±λc∆t After the advection step

A C B

D

If r > r₊^c , then a new grid point D is inserted midway between A and B (insertion)

If r < r^c₋, then grid points A and C are removed and a new grid point is intro- duced midway between the positions of A and C (merging)

(11)

2d-CLaMS: Dynamically Adaptive Grid

Before the advection step

A C B

Determine nearest neighbors (e.g. for point A), r0 - mean distance between air parcels

Set the critical Lyapunov exponent λ_c and the time step ∆t (free parameter) Define r^c_±

After the advection step

A C B

D

If r > r₊^c , then a new grid point D is inserted midway between A and B (insertion)

If r < r^c₋, then grid points A and C are removed and a new grid point is intro- duced midway between the positions of

(12)

Mixing in the vicinity of the subtropical jet

Subtropical jet

over Himalayas

(13)

Mixing in the vicinity of the subtropical jet

Subtropical jet over Himalayas

Strong

deformations ...

(14)

Mixing in the vicinity of the subtropical jet

Subtropical jet over Himalayas

... and mixing !

(15)

Critical deformation

Mixing in CLaMS is controlled by:

1. critical deformation γ ^c or critical eccentricity of the deformation ellipse r ₊ ^c /r ₋ ^c

γ ^c = λ ^c ∆t = 0.5 ln ^r _r

⁺^c^c

−

2. horizontal, r 0 (2d) and vertical resolution, ∆z (3d)

Comparison with experiment ^⇒ mixing parameters

(16)

Mixing versus CRISTA-1 observations

CRISTA observations

(November, 10, 1994)

(17)

Mixing versus CRISTA-1 observations

CRISTA observations (November, 10, 1994)

CLaMS-2d

no mixing

(18)

Mixing versus CRISTA-1 observations

CRISTA observations (November, 10, 1994)

CLaMS-2d

mixing too strong

(19)

Mixing versus CRISTA-1 observations

CRISTA observations (November, 10, 1994)

CLaMS-2d

optimal mixing

(20)

N ₂ O variability in terms of PDFs

Synoptic map of N2O from CRISTA-1, 4-6 Nov 1994 at θ = 700 K.

−200 −100 0 100 200

∆ N2O [ppbv]

0.0001 0.0010 0.0100 0.1000

PDF

CRISTA observations Gaussian fit Exponential tails, p=0.63

CRISTA observations Gaussian fit

Exponential tails, p=0.63

PDF – probability density function of N2O differences between neighboring observations with 150< r <250 km

⇒ “fat” tails inidicate filaments and 2d-turbulence

(21)

N ₂ O variability in terms of PDFs

−200 −100 0 100 200

∆ N2O [ppbv]

0.0001 0.0010 0.0100 0.1000

PDF

CRISTA observations Gaussian fit Exponential tails, p=0.63

CRISTA observations Gaussian fit

Exponential tails, p=0.63

PDF – probability density function of N2O differences between neighboring observations with 150< r <250 km

(22)

CLaMS versus CRISTA PDFs

−300 −200 −100 0 100 200 300

∆ N2O [ppbv]

0.0001 0.0010 0.0100 0.1000

PDF

CRISTA observations

reduced

satellite optimized in−situ optimized

enhanced

KASIMA

(Khosrawi et al., 2005, ACP)

(23)

Optimal mixing (in situ observation)

Spatial variability of the simulated time series γ:

γ ≈ ∆_theor

∆exp

, ∆ =

n

X

i=1

|f(t_i+1) −f(t_i)|

09:00 10:00 11:00 12:00 13:00 14:00 15:00 time [UTC]

0.6 0.8 1.0 1.2 1.4 1.6

CH4 [ppm]

0.6 0.8 1.0 1.2 1.4 1.6

CH4 [ppm]

γ > 1

no mixing

09:00 10:00 11:00 12:00 13:00 14:00 15:00 time [UTC]

0.6 0.8 1.0 1.2 1.4 1.6

γ ∼ 1

Exp, ARGUS CLaMS

optimal mixing

09:00 10:00 11:00 12:00 13:00 14:00 15:00 time [UTC]

0.6 0.8 1.0 1.2 1.4 1.6

γ < 1

mixing too strong

(24)

Optimal mixing (in situ observation)

Spatial variability of the simulated time series γ:

γ ≈ ∆_theor

∆exp

, ∆ =

n

X

i=1

|f(t_i+1) −f(t_i)|

09:00 10:00 11:00 12:00 13:00 14:00 15:00 time [UTC]

0.6 0.8 1.0 1.2 1.4 1.6

CH4 [ppm]

0.6 0.8 1.0 1.2 1.4 1.6

CH4 [ppm]

γ > 1

no mixing

09:00 10:00 11:00 12:00 13:00 14:00 15:00 time [UTC]

0.6 0.8 1.0 1.2 1.4 1.6

γ ∼ 1

Exp, ARGUS CLaMS

optimal mixing

09:00 10:00 11:00 12:00 13:00 14:00 15:00 time [UTC]

0.6 0.8 1.0 1.2 1.4 1.6

γ < 1

mixing too strong

(25)

Optimal mixing (tracer correlations)

Mean deviation from the CH4/Halon-1211 correlation

² = 1 m

m

X

i=1

²_i ²_i = 1 − f_i^theor f_i^exp

!2

+ 1 − g_i^theor g_i^exp

!2

no mixing

optimal mixing

mixing too strong

(26)

Optimal mixing (tracer correlations)

Mean deviation from the CH4/Halon-1211 correlation

² = 1 m

m

X

i=1

²_i ²_i = 1 − f_i^theor f_i^exp

!2

+ 1 − g_i^theor g_i^exp

!2

no mixing optimal mixing mixing too strong

= 0 means a perfect agreement between CLaMS and observations.

(27)

Optimal mixing

1

Spatial variability γ 10

Deviation from H1211/CH4 correlation ε [%] 0.5 0.75 1.0 1.5 2.0 ∞λc∆t=

300 km 200 km

150 km 100 km excess

mixing no mixing

Optimal mixing: γ = 1, = 0

Conclusions:

Mixing improves tracer transport only for r0 < 300 km

Best agreement with satellite and

in situ observations for:

γ_c = 0.8 (satellite). . .1.5 (in situ) γc=λ_c∆t - critical deformation

Optimized aspect ratio: α ≈ 250 Konopka et al. 2004, JGR

Per day ≈ 10% of air parcels are mixed. If mixinig occurs:

(28)

How homogeneous and isotropic is strat. mixing?

CLaMS v. Exp. Res. r 0 [km] Crit. def. γ ^c Ecc. r ₊ ^c /r ₋ ^c

2d v. CRISTA 200 0.8 5

2d v. in situ 40 1.2 11

3d v. in situ 80 1.5 20

anisotropy ^∼ (r ₊ ^c /r ^c

− ) ²

(29)

How homogeneous and isotropic is strat. mixing?

mixing inhomogeneous in space

(30)

How homogeneous and isotropic is strat. mixing?

mixing inhomogeneous in time

0.0 0.2 0.4 0.6 0.8 1.0 1.2

Deformation param. γ

γc=0.8

0.5 1.0 1.5 2.0 2.5

D−× 104 [m2 /s]

Deform. param., γ=λ∆

t − Diffusion coef., D

Deform. param., γ=λ∆t Diffusion coef., D₋

80 70 60 50 40 30

Equivalent Latitude [deg S]

80 70 60 50 40 30

22.10.94 01.11.94 10.11.94

7.8

1.5 2.2 2.9 3.6 4.3 5.0 5.7 6.4 7.1 7.8 9.5

<D> [10³ m²/s]

Formation of fragments A, B

Strong mixing

Moderate and weak mixing

Vortex edge

(31)

Permeability of the vortex edge

400 600 800 1000 1200

Pot. Temperature [K]

400 600 800 1000 1200

01.01.03 01.03.03 01.05.03

time

100.00

0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 80.00 90.00 100.00 Vortex Air

mW1 MW mW2 mW3 FW

1.5

1.5 1.5

1.5

Vortex Air=50% EQLAT>70 EQLAT>65 EQLAT>75

Vortex Air=50%

EQLAT>70 EQLAT>65 EQLAT>75

Despite a strong activity of planetary waves in winter 2002-2003, the vortex between θ =450

(32)

Permeability of the vortex edge

85 80 75 70 65 60

400 500 600 700 800 900 1000

85 80 75 70 65 60

400 500 600 700 800 900 1000

1.50

0.25 0.38 0.50 0.62 0.75 0.88 1.00 1.12 1.25 1.38 1.50 CH4 [ppmv]

75

98 98

Well-isolated vortex remnant 5 weeks after “vortex split” in September 2002, “frozen-in”

ozone loss, (Konopka et al. 2005, JAS)

(33)

Long-lived vortex remnants

θ = 450K (≈ 18 km)

⇐ Vortex remnants on April 29, 2000 about 40 days after the vortex breakup.

These remnants show a strong vertical coherence and correlation with PV.

(34)

Long-lived vortex remnants

θ = 450K (≈ 18 km) θ = 585K (≈ 24 km)

Vortex remnants on May 22, 1997, about 12 days after the vortex breakup. The remnants are trapped in the “solid body” summer circulation (Konopka et al., ACP, 2003)

(35)

Mixing ⇒ chlorine deactivation

Chlorine deactivation = formation of ClONO2

ClO +NO2 → ClONO2

North pole mid−latitude air

vortex air ClO

NOx weak mixing

Question:

What is the mechanism of the chlorine deactivation ?:

1. mixing of the vortex ClO with NO_x-rich mid-latitude air

2. in situ NO_x production due to the photolyt- ical decomposition of HNO3

(36)

Mixing ⇒ chlorine deactivation

ClONO2 collar at θ = 450 K on March 11, 2000 (CLaMS-2d)

Optimal mixing

(37)

Mixing ⇒ chlorine deactivation

ClONO2 collar at θ = 450 K on March 11, 2000 (CLaMS-2d)

Optimal mixing Pure advection

in situ chemistry in air parcels ⇒

(38)

Impact of mixing on O ₃ chemistry

Ozone depletion at 450 K on 29.04.2000

no mixing adjusted mixing mixing too strong

(39)

Ozone loss: Impact on mid-latitudes

01.03.2000 01.04.2000 01.05.2000 time

−10

−5 0 5 10 15 20

Mean O3 loss [%]

30<lat<90 N

adjusted mixing − D

=1.1 10 −

m 3

/s /s 22 2 m m 5 4 =1.1 10 =2.6 10 − − mixing too large − D − D

/s 2 m 5 =4.2 10 − − D

/s =0 y , Br y strong denitrification Cl

adjusted mixing − D₋=1.1 10³ m²/s mixing too large − D₋=2.6 10⁴ m²/s − D−=1.1 10⁵ m²/s − D₋=4.2 10⁵ m²/s strong denitrification

(40)

Ozone loss: Impact on mid-latitudes

01.03.2000 01.04.2000 01.05.2000 time

−10

−5 0 5 10 15 20

Mean O3 loss [%]

30<lat<90 N

=1.1 10 −

m 3

/s 2 m 5 =4.2 10 − − D

Cl_y, Br_y=0

(41)

Ozone loss: Impact on mid-latitudes

01.03.2000 01.04.2000 01.05.2000 time

−10

−5 0 5 10 15 20

Mean O3 loss [%]

30<lat<90 N

=1.1 10 −

m 3

/s 2 m 5 =4.2 10 − − D

(42)

Ozone loss: Impact on mid-latitudes

01.03.2000 01.04.2000 01.05.2000 time

−10

−5 0 5 10 15 20

Mean O3 loss [%]

30<lat<90 N

=1.1 10 −

m 3

/s 2 m 5 =4.2 10 − − D

Cl_y, Br_y=0

(43)

Ozone loss: Impact on mid-latitudes

01.03.2000 01.04.2000 01.05.2000 time

−10

−5 0 5 10 15 20

Mean O3 loss [%]

O₃ destruction within the vortex

O₃ destr. in vortex remnants (ClO_x)

Dilution of depleted vortex remnants

30<lat<90 N

=1.1 10 −

m 3

/s 2 m 5 =4.2 10 − − D

(44)

Conclusions

CLaMS - Lagrangian 3d-CTM:

trajectories (reversibel) + mixing (irreversibel) + chemistry 1. mixing driven by large-scale deformations

2. model validation:

polar stratosphere und extratropical UT/LS region Examples:

Permeability of the vortex edge

layerwise mixing into the vortex triggered by breaking planetary waves

long-living vortex remnants with frozen-in ozone loss (impact on the mid-latitudes) Summer circulation ⇒ “solid body rotation” ⇒ weak mixing

chlorine deactivation in the well-isolated vortex driven by in situ chemistry

(45)

Acknowledgments:

CLaMS group:

Danny McKenna (now at NCAR, Boulder),

Rolf Müller, Jens-Uwe Grooß, Gebhard Günther, Paul Konopka, Hildegard-Maria Steinhorst, Stephan Bausch, Carsten Lemmen, Verena Cals, Nicole Thomas, Jürgen Ankenbrand, Reimar Bauer Modeling support:

Yvan Orsolini (NILU, Oslo), Glenn Carver (University of Cambridge), Ken Carslaw (University of Leeds), Yasuhiro Sasano (NIES, Tsukuba), Richard Swinbank (UKMO), Joanna Haigh (Imperial College, London)

Experimental data:

CRISTA-Team:

Martin Riese, Dirk Offermann, Reinhold Spang, Volker Küll (University of Wuppertal) ...

SOLVE-THESEO-Team:

James Elkins (University of Colorado), Hans-Jürg Jost (NASA, Moffett Field), Geoffrey Toon (JPL, Pasadena), Andreas Engel, Ulrich Schmidt (University of

(46)

Transport of ozone and CO during SPURT

Alt [km]

18.0

15.5

13.0

10.5

8.0

Equator North Pole

TROPICS

MID−LATITUDES

POLAR REGION STRATOSPHERE

500 K

290 K

radiation scheme Vertical velocities from

?

OK

Configuration of CLaMS

for 2 years simulation

(47)

Transport of ozone and CO during SPURT

Alt [km]

18.0

15.5

13.0

10.5

8.0

TROPICS

MID−LATITUDES

STRATOSPHERE 500 K

290 K

radiation scheme Vertical velocities from

?

OK

Configuration of CLaMS

for 2 years simulation

(48)

Seasonal cycle of ozone

20 40 60 80

Equivalent Latitude [deg N]

320 340 360 380 400 420

20 40 60 80

320 340 360 380 400 420

1000.

0.

100.

200.

300.

400.

500.

600.

700.

800.

900.

1000.

O₃ [ppmv]

Stratosphere

Troposphere

Fall

Zonally averaged ozone

20 40 60 80

320 340 360 380 400 420

20 40 60 80

320 340 360 380 400 420

Winter

20 40 60 80

320 340 360 380 400 420

20 40 60 80

320 340 360 380 400 420

Spring

20 40 60 80

320 340 360 380 400 420

20 40 60 80

320 340 360 380 400 420

Summer

STT-

Transport

(49)

Seasonal cycle of ozone

20 40 60 80

320 340 360 380 400 420

20 40 60 80

320 340 360 380 400 420

1000.

0.

100.

200.

300.

400.

500.

600.

700.

800.

900.

1000.

O₃ [ppmv]

Stratosphere

Troposphere

Fall

Zonally averaged ozone

Mean tropopause

Subtropical jet

(transport barrier)

20 40 60 80

320 340 360 380 400 420

20 40 60 80

320 340 360 380 400 420

Winter

20 40 60 80

320 340 360 380 400 420

20 40 60 80

320 340 360 380 400 420

Spring

20 40 60 80

320 340 360 380 400 420

20 40 60 80

320 340 360 380 400 420

Summer

STT-

Transport

(50)

Seasonal cycle of ozone

20 40 60 80

320 340 360 380 400 420

20 40 60 80

320 340 360 380 400 420

1000.

0.

100.

200.

300.

400.

500.

600.

700.

800.

900.

1000.

O₃ [ppmv]

Stratosphere

Troposphere

Fall

320 340 360 380 400 420

Winter

20 40 60 80

320 340 360 380 400 420

20 40 60 80

320 340 360 380 400 420

Spring

20 40 60 80

320 340 360 380 400 420

20 40 60 80

320 340 360 380 400 420

Summer

A Lagrangian view of the Stratospheric Mixing