Vertical Structure of the Tropical Troposphere (including the TTL) Ian Folkins Department of Physics and Atmospheric Science Dalhousie University

Deep Convection ubiquitous shallow convection (cumulus congestus), ~ 28% of rainfall during TOGA/COARE (Johnson et al., JAS, 1999)

Trimodal cloud top distribution (lidar obs): Shallow Boundary Layer Deep Shallow Boundary Layer LandOcean

Mapes and Houze, JAS, 1995 Rawinsonde wind measurements from the TOGA/COARE IFA when deep convection present Johnson and Cieslinski, 2000 Deep Outflow Layer Inflow to feed downdrafts (mainly)

Clear Column: Radiative Descent Cloudy Column Convective Outflow can be estimated from clear sky mass fluxes (radiative + evaporative). evaporative moistening (downdrafts) Deep Outflow Shallow Outflow

Folkins and Martin, JAS, 2005 Mass Flux Mass Flux Divergence Deep shallow

 ~ 1000 km ----  coolingheating Two Distinct Circulations? 1)Tropical-scale Hadley/Walker circulation: deep condensational heating balances radiative cooling. 2) Regional scale downdraft/shallow convection circulations: shallow convective heating balances Evaporative cooling. radiative cooling

Shallow Outflow Layer Deep Outflow Layer Outflow Layers related to changes in stability

Ozone is rapidly destroyed in the tropical marine boundary layer. Deep convection pumps this low ozone air to higher altitudes. Low O3 Ozone is chemically produced at a rate of 1-2 ppbv/day above 6 km in the background atmosphere Low O3

Convective Tracers: O3 and CO At deep convective marine locations, there is an ozone minimum near 12 km, probably associated with deep convective outflow Deep convective outflow maintains high CO mixing ratios till 15 km, presumably the height at which the convective replacement time is similar To the chemical lifetime.

Tropical mean cloud mass flux and divergence profiles from 3 convective schemes [Emanuel, Zhang and McFarlane (GEOS-4), Relaxed Arakawa Schubert (GEOS-3)]

Dry Mixing dT = 2 C M = 1 kg dT = 0 C M = 1 kg dT = 1 C M = 2 kg Entrainment of dry air reduces the buoyancy B of a rising air parcel, but has no effect on the buoyancy flux MB. (MB = mass flux *buoyancy) Mixing Emanuel and Bister JAS, 1996

Moist Mixing Dry Air Evaporation of cloud droplets: Moist Mixing More rapid decrease in buoyancy, and a decrease in MB Buoyancy Reversal higher condensate loading

Wei et al, JAS, 1998 PDF of Updraft Buoyancy (850 mb – 600 mb) Moist mixing is very effective at reducing updraft buoyancies. (at least in the lower troposphere.)

PDF of Downdraft Buoyancy (850 mb – 600 mb) Wei et al, JAS, 1998

Physics of Moist Mixing (strongly damped buoyancy flux) Shallow Convection Physics of Dry Mixing (constant buoyancy flux) Deep Convection colder temperatures (higher altitudes) Higher background RH (water vapor feedback) reduced condensate loading (rapid rainout)

Brewer Dobson Circulation TTL

What is the Tropical Tropopause?

What is the TTL? Convective Outflow Hadley Circulation Brewer Dobson Circulation Level of Mean Ascent Top Level of Convective outflow TTL 17.5 km 15.5 km

RH ~ 60% RH ~ 80% LZH: level of zero radiative heating TTL: uplift moistening; need dehydration mechanism 17 km 15.5 km Subsidence Drying RH > 100% T~198 K T~192 K RH should increase as you approach the TTL from below Detrainment Moistening 10 km

TTL Aircraft measurements show high RH in the TTL

Harvard Group

Positive heating rates at the cold point tropopause are due to LW heating from Ozone

Ozone has a seasonal cycle At the tropical tropopause (probably cause by a seasonal variation in convective outflow)

Ozone and Water Vapor Budgets Coupled Ozone high Ozone affects seasonal cycle in radiative heating rates

good BD mass flux good convective mass fluxes + good convective outflow profile good ozone profile chemistry STE good temperature profile cloud radiative effects good strat H2O entry mixing ratio good dehydration mechanism better climate/ozone depletion forecasts TTL “Virtuous Circle” Start Here

Summary 1. Moist convection has a rich vertical structure. 2. Accurate modelling of the cold point temperature, and of chemical species profiles in the TTL, requires convective schemes which can accurately simulate the shape of the deep outflow layer. 3. There are significant variations in convective outflow between convective parameterizations (at least when run in assimilated modes). 4. The water vapor budget of the TTL is unique – it appears to require an in situ irreversible dehydration mechanism to prevent large scale supersaturation. 5. Ozone-Temperature coupling in the TTL