C. Michael Volk, J. Baehr, C. Homan, A.C. Kuhn, A. Werner, S. Viciani, A. Ulanovsky, F. Ravegnani, P. Konopka, D. Brunner

4th SPARC General AssemblyBologna, 4 September 2008

Transport across the tropical tropopause

by convection, mixing, and slow upwelling:

Insights from recent in situ observations with

the Geophysica aircraft

Transport Processes in the Tropical Tropopause Region

Stratospheric

Subtropical barrierIsolated tropical stratosphere

Goal: Quantitative understanding of the balance between the dominant transport

processes as function of time and space

(from Konopka et al. ACP 2007)

M55 In situ Tracer Measurements

HAGAR (Univ. Frankfurt)

N2O, F11, F12, H1211, SF6, CO2

(CH4, CO, H2)

FOZAN (CAO, Russia): O3

COLD (INOA, Italy): CO

Campaign locations 1999-2006 (incl. transfer flights)

Adapted from S. Borrmann

MarrakeshMarocco

OuagadougouBurina Faso

AccraGhana

VeronaItaly

Seychelles

TroCCiNOx 2/2005Continental convection

APE-THESEO 2-3/1999Non-convective

SCOUT-O3 11-12/2005Maritime Convection

AMMA 8/2006Cont. Convection

Total # of tropical flights: 44

300

250

200

150

100

FO

ZA

N O

3 (

pp

b)

320318316314312

HAGAR N2O (ppb)

AMMA Flights 060804, 060808, 060811, 060813

Correlation Coefficients R:355-365 K: R = -0.15

365-375 K: R = -0.025

375-385 K: R = -0.39

385-395 K: R = -0.51395-405 K: R = -0.39

Confidence level > 95%

320310300290

HAGAR N2O (ppb)

420

400

380

360

340

Th

eta

(K

)

40 30 20 10

Latitude (°N)

Flight Dates 060804

060808 060811 060813 060729 060731

060801 060816 060817

no low

N2O in

TTL

TTL: Stratospheric (horizontal) inmixing above W. Africa 2006?

No significant negative

O3-N2O correlation

below CPT

fraction of aged air small (

Significant anti-correlations between O3 and F12 found everywhere

between 365 K and 380 K indicate significant stratospheric influence.

(see poster by A.C. Kuhn et al.)

The CLaMS model calculates a fraction of 30-40% of TTL air over subtropical Brazil originating from the

stratosphere

TTL: Stratospheric inmixing above Brazil 2005?

Brazil 2005

Convective uplift of boundary layer air into the TTL: CO2

051130b

355 K

Max. outflow level ~ 355 K

Darwin 2005

378376374372

CO2 [µmol/mol]

20

15

10

5

0

Altitude [km

]

Storm chase February 4th(ascent and descent)

median profile(non-convectiveflights)

convective influence up to 17 km

377376375374CO2 [nmol/mol]

Flight Dates 050201 050204 050205 050212

050215 050217 050218

050204 Theta: 355-378 K

300

250

200

150

100

50

0

FO

ZA

N O

3 [

nm

ol/m

ol]

120100806040

TDL CO [nmol/mol]

Flight Dates

050204 050205 050212 050215 050218

Troposphere

TTL

Stratosphere

Mixing of overshooting air

(see poster by J. Baehr et al.)

vertical transport of convectively detraining air to the upperTTL by mixing

Mixing of overshooting air in the TTL: Brazil 2005 correlations

10090807060504030

COLD CO (ppb)

150100500

FOZAN O3 (ppb)

400

380

360

340

320

Th

eta

(K

)

"Background" Flights 051119

051123

051125

051129

Background Average

051130 Flights M55 afternoon

M55 evening

O3 sonde afternoon O3 sonde evening

Vertical mixing in the TTL over Darwin 11/2005

Enhanced O3 and CO levels on 051130 consistent with vertical mixing(not consistent with horizontal stratospheric inmixing into TTL)

Enhanced CO

Enhanced O3

D. Brunner

Most likely explanation for high O3 and CO levels:

Vertical mixing in the vicinity of the subtropical jet

Vertical cross section at 110°E of the wind speed (color contours) and PV (black contours).

=> air travels alongthe edge of the

subtropical jet

=> vertical wind

shear facilitatessmall-scale mixing

White squares: crossing of backward trajectories from Geophysica flight track

The subtropical jet as a major agent for vertical mixing

CLaMS: Chemical Langrangian Model of the Stratosphere

(Konopka et al., 2007)

Vertical diffusivity at 380 K (February)

382380378376374CO2 (ppmv)

440

420

400

380

360

340

The

ta (

K)

Flight Dates 060804 060808 060811 060813

Convection versus slow ascent: CO2 profiles West Africa

~15 km

Max. convective outflow ~ 14-15 km

Coherent tape

recorder signal due to slow ascent

Satellite picture flight 11th Aug.

Small features at 08-08-2006 due to overshooting convection ?

Define mean CO2surface = average Mauna Loa +Samoa

385

380

375

370

CO

2 (

pp

m)

200720062005200420032002Year

CO2 tracer clock to quantify slow ascent

CO2 tape recorder

Use seasonal change of CO2 to measure mean age since entering the TTL

NOAA ESRL (Conway et al.)

Assumptions:

• At TTL bottom (~350 K): CO2(t) = CO2surface(t)

• Uniform ascent rate in the TTL w = dθ/dt

• => match “model” profile in region dominated by slow ascent

(360 < θ < 390 K): CO2(θ, t) = CO2surface (t – [θ-350K]/w)

AMMA

SCOUT-O3TROCCINOX

382381380379378377376CO2 (ppmv)

420

400

380

360

340

Th

eta

(K

)

Flight Dates060804060811060813

grey: all datared: data with F12>532 ppm 0.5K/day

0.6 K/day

0.4 K/day

AMMA – August 2006

Mean ascentrate in the TTL:

0.5 +/-0.1 K/day

=> Air ascended since mid June

390 K

360 K

SCOUT-O3- November/December 2005

Mean ascentrate in the TTL:

0.5 +/-0.1 K/day

380379378377CO

2 (ppmv)

400

390

380

370

360

350

340

330

Th

eta

(K

)

Flight Dates051123051125051128051129051206

grey: all flights red: f12> 532 ppm

0.6 K/day 0.5 K/day

0.4 K/day

390 K

360 K=> Air ascended since early October

380379378377376375374373CO2 (ppmv)

420

400

380

360

340

Th

eta

(K

)

Flight Dates050212050215050217050218050224050227a050227b

grey: all datared: data with F12>532 ppm

1.7 K/day

1.4 K/day 2.0 K/day

TROCCINOX- February 2005

390 K

370 K

Mean ascentrate in the TTL:

1.7 +/-0.6 K/day

=> Air ascended since late January

Conclusions

• Horizontal inmixing of aged stratospheric air into TTL: appears to be common close to the subtropical jets, but not in inner tropics

• Convection: Max. outflow levels @ ~365 K or ~15 km, max. in Darwin

• Vertical mixing following convective overshooting:

can effect much higher levels (~ 390K or 17 km observed over Brazil 2005) => causes strat.-trop. exchange

W. Africa: possibly influence of overshooting up to 420 K

• Vertical mixing along subtropical jets:

likely a major agent for tropical stratosphere-troposphere exchange !

• Mean diabatic ascent (360-390K) derived from CO2 clock:

June/July 2006: 0.5 +/- 0.1 K/day

Oct./Nov. 2005: 0.5 +/- 0.1 K/day

Jan./Feb. 2005: 1.7 +/- 0.6 K/day