Convective Influence on the Heat Balance of the Tropical Tropopause Layer

1. Convective Influence on the Heat Balance of the Tropical Tropopause Layer Chris Bretherton and Peter Blossey Department of Atmospheric Sciences University of Washington Zhiming Kuang Harvard University

2. The TTL is the transitional layer between the troposphere and the stratosphere cold point tropopause TTL mixing barrier or a secondary tropopause Mean soundings over Samoa

3. Many processes may combine to affect TTL T, q, trace species Holton

4. Is cold point T, q significantly affectedby convection?Maybe not… Argument 1 Convective clouds are rare in the TTL (Gettelman et al., JGR, 2002). … but rare might be enough, for a slow TTL radiative response time.

5. Argument 2 A cold point tropopause solely in radiative equilibrium can exist well above the convectively adjusted region in a 1D model (Thuburn and Craig, JGR, 2002) …but is a 1D model good enough for this problem? Convection top

6. Our approach: Use a cloud resolving model to examine TTL heat balance in radiative-convective equilibrium. Kuang and Bretherton 2004 JAS Similar approach, different result: Kuepper et al. 2004 JGR • Colorado State University CRM (Khairoutdinov and Randall, JAS, 2003), anelastic, bulk microphysics, interactive radiation, 100 day run. • Doubly periodic, 64km×64km domain, 1km horizontal grid. • Uniform SST, zero wind shear, no diurnal cycle. • Fixed ozone. • Init lower-strat water vapor = 4 ppmv. • Control simulation has <w> = 0. • Results robust for larger domain and finer grid sizes. Sponge layer

8. Cold point tropopause level radiative convective Heat Balance in the TTL Convective cooling

9. Entrainment Layer LNB Entrainment of warm TTL air by occasional overshooting tops Cold point level Convective Mass flux Δθ Level of zero rad. heating LNB Radiation Convection 0

10. How much effect does convective cooling have on the temperature of the TTL? With convective cooling Convective cooling removed The radiative spring is weak in the TTL: trad = DT/Qrad ~ 30 d.

11. A convective event Domain averaged values 500m above the cold point BL tracer (12 h decay) cloud ice

12. Long term behavior at the cold point tropopause

13. In perturbed simulations, convective cooling maximum is also collocated with the cold point in this model. radiative dynamical Cold point Cold point SST+2K W+0.3mm/s Filled: convective heating rate Open: radiative heating rate

14. Convective impact negligible (like Thuburn+Craig) Why different? - Radiation? - Humidity? - Transport? …an open problem.

15. Mock-Walker 2D CRM simulation (Blossey et al. poster) 100-200 d means • Radiatively-forced subsidence above convection. • Occasional convective penetration to cold point but little conv. cooling

16. Summary • In some, but not other, CRM simulations of radiative-convective equilibrium, eddy cooling due to overshooting convection affects cold point height and temperature. • High CRM vertical resolution (100 m) is necessary to avoid spurious gravity-wave damping in the TTL. • Walker-circulation results suggest that radiatively-driven subsidence above convective anvils may be at least as important. • We still need to develop a TTL cloud-resolving modeling framework that includes appropriate feedbacks between convection and large-scale circulations.