4.2
Variability and Trends in Dynamical Forcing of Tropical Lower Stratospheric Temperatures

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner
Tuesday, 6 January 2015: 8:45 AM
212A West Building (Phoenix Convention Center - West and North Buildings)
Stephan Fueglistaler, Princeton University, Princeton, NJ; and M. Abalos, T. Flannaghan, P. Lin, and W. J. Randel

We analyse the relation between tropical lower stratospheric temperatures and dynamical forcing over the period 1980-2011 using NCEP, MERRA and ERA-Interim reanalyses. The tropical mean thermodynamic energy equation with Newtonian cooling for radiation is forced with two dynamical predictors: (i) the average eddy heat flux of both hemispheres; and (ii) tropical upwelling estimated from momentum balance following Randel et al. (2002). The correlation (1995-2011) for deseasonalised tropical average temperatures at 70~hPa with the eddy heat flux based predictor is 0.84 for ERA-Interim (0.77 for the momentum balance calculation), and 0.87 for MERRA. The eddy heat flux based predictor indicates a dynamically forced cooling of the tropics of ~ -0.1K/decade (~ -0.2K/decade excluding volcanic periods) for the period 1980-2011 in MERRA and ERA-Interim. ERA-Interim eddy heat fluxes drift slightly relative to MERRA in the 2000's, possibly due to onset of GPS temperature data assimilation. While NCEP gives a small warming trend, all 3 reanalyses show a similar seasonality, with strongest cooling in January/February (~ -0.4K/decade, from northern hemispheric forcing) and October (~ -0.3K/decade, from southern hemispheric forcing). Months preceding and following the peaks in cooling trends show pronounced smaller, or even warming, trends. Consequently, the seasonality in the trends arises in part due to a temporal shift in eddy activity. Over all months, the southern hemisphere contributes more to the tropical cooling in both MERRA and ERA-Interim. The residual time series (observed minus estimate of dynamically forced temperature) are well correlated between ERA-Interim and MERRA, with differences largely due to temperature differences. The residual time series is dominated by the modification of the radiative balance by volcanic aerosol following the eruption of El~Chichon (maximum warming of ~ +3K at 70hPa) and Pinatubo (maximum warming of ~ +4K at 70hPa), with a strong dynamical response during Pinatubo partially masking the aerosol heating.