Long-Term Evolution of the Tropical Cold Point Tropopause

Simulation Results and Attribution Analysis

Thomas Reichler, U. of Utah, Salt Lake City, USAJohn Austin, UCAR-NOAA-GFDL, Princeton, USA

Motivation• Tropical tropopause is closely related to climate change• Examples:

– Tropopause controls stratospheric water vapor– Tropopause is controlled by tropical upwelling, QBO, tropo-

spheric and stratospheric temperature changes

• Exact cause-and-effect relationship between climate change and tropopause change is unclear

• Outline– Simulations with coupled chemistry climate model– Past and future evolution of tropical tropopause– Identify factors for change:

1. regression analysis2. conceptual model

Methodology• GFDL-AMTRAC: stratosphere resolving coupled

chemistry climate model (2.5x2.5, L48)• Simulation: 1960-2100, 3 members

– PAST: 1960-1990, historical forcings (SST, GHG, ODS)– FUTURE: 1990-2100, IPCC-A1B and WMO (2003)

forcings, SSTs from GFDL-CM2.1

• Cold point tropopause– Interpolation after Reichler et al. (2003)– Lapse rate definition: 0 K/km– Zonal averages (22S-22N) and annual means

Tropical Tropopause Evolution• Heights

– PAST: Increase– FUTURE: Increase– 1960-2100: ca. 1 km or

ca. 10%

• Temperatures

– Climatologically important transition

– PAST: Cooling– FUTURE: warming

Decadal Trends

Tropics PAST FUTURE

Global1980-2004 OBS

Height [m/dec.] 70 64 123 64

Temperature [K/dec.] -0.13 0.25 -0.27 -0.41

1. Linear Regression ModelFit tropical tropopause parameters (temperature, pressure, height) to a linear regression model using the following four factors:

AER Aerosols (60 hPa at equator)SST Tropical SSTs (22S-22N)O3 Total ozone (globally averaged)UPW Tropical mass upwelling (77 hPa), BDC

These factors represent major processes known to influence the tropopause.

Regression Parameters Evolution

Regression Analysis: Heights

Plots are decadally smoothed

• Contribution of each term to tropopause height1. UPW - most important2. SST - comes second

Plots are decadally smoothed

1. O3 - dominates PAST2. SST - dominates FUTURE3. UPW - probably strongly related to SST4. AER - small impact

Regression Analysis: Temperatures

1. ΔTt > 0: height ↑ temperature ↑2. ΔTs < 0: height ↑ temperature ↓

• Staten and Reichler (2008) show:

• Shepherd (2002): Constant lapse rates above (γs) and below (γt) tropopause

• Explain tropopause change by temperature change below (ΔTt) and

above (ΔTs )

2. Conceptual Tropopause Model

and

Testing the Simple Model

PAST

FUTURE

• Test impact of simulated temperature trends above ΔTs

(0, LS) and below ΔTt (0, LT, UT) on tropopause itself

γs = -4 K/km, γt = 6 K/km

PAST LS (and UT)FUTURE (LS and) UT

Cause and Effect AnalysisChange per century

ΔZtrop ΔTtrop ΔTUT ΔTLS O3 UPW GHG

PAST 700 -1.3 2.5 -5 ↓↓ ↑↑ ↑FUTURE 640 2.5 5 0 ↑ ↑ ↑↑

• Tropopause increases mostly due to LS cooling (PAST) and UT warming (FUTURE)

ΔZtrop

• Similar• PAST: LS cooling dominates• FUTURE: UT warming dominates

ΔTtrop

GHGFUTURE ΔTUT

O3, UPW, GHGPAST ΔTLS

Conceptual vs. Regression Model

• In the PAST, ozone depletion was most important for cooling and lifting the tropopause

• In the FUTURE, greenhouse gas increase will be most responsible for warming and lifting the tropopause

Conceptual Regression

PASTΔT↓

O3 + UPW + GHGO3

ΔZ↑ UPW

FUTUREΔT↑ GHG (SSTs) SST + O3ΔZ↑ UPW

Thank You

More info: Austin and Reichler (2008, JGR)

Trend Analysis