GFDL - Geophysical Fluid Dynamics Laboratory

Tropical Climate Change Control of the Lower Stratospheric Circulation

February 5th, 2015

Pu Lin, Yi Ming, V. Ramaswamy. Geophysical Research Letters. DOI: 10.1002/2014GL062823.

Summary

Tropical air has less ozone than polar air, even though the tropical stratosphere is where most atmospheric ozone is produced. The Brewer-Dobson circulation is considered key to understanding this apparent contrast. It also brings water vapor, aerosols and other species from the troposphere up into the stratosphere. The strength of the Brewer-Dobson circulation directly affects the thermal structure of the stratosphere and upper troposphere, and impacts the transport and distribution of important climate-influencing constituents including stratospheric water vapor, ozone, and volcanic aerosols.

GFDL global climate models were used to investigate how the Brewer-Dobson circulation would vary in response to different natural and anthropogenic climate forcings. The authors calculate the strengths of the Brewer-Dobson circulation simulated by GFDL global climate models CM3 and CM2.1, and find that the strengths correlate with the tropical mean surface temperature. This correlation is also supported by observational-based analysis.

The variations of the Brewer-Dobson circulation’s shallow branch can be explained largely by those of the tropical mean surface temperature. Interestingly, this robust mechanism holds for all forcing types, and operates both at interannual and at multi-decadal timescales. This correlation between the stratospheric circulation strength and tropical-mean surface temperature across timescales and forcings indicates that one can constrain the model-simulated long-term changes of the circulation with observations of much shorter durations, and gain more confidence in the climate model projection of stratospheric circulation, composition and its downward impacts on the troposphere and surface.

Scatter plot of the strength of the Brewer-Dobson circulation shallow branch versus the tropical-mean surface temperature in CM3. (a) For the interannual timescale in the control simulation. ENSO-neutral years are marked by red dots. (b) For the decadal to multi-decadal timescale in the control and historical forced simulations. Climatological means are removed. The gray line marks the result from the an idealized experiment in which SST is globally uniformly increased by 4K.
Scatter plot of the strength of the Brewer-Dobson circulation shallow branch versus the tropical-mean surface temperature in CM3. (a) For the interannual timescale in the control simulation. ENSO-neutral years are marked by red dots. (b) For the decadal to multi-decadal timescale in the control and historical forced simulations. Climatological means are removed. The gray line marks the result from the an idealized experiment in which SST is globally uniformly increased by 4K.