lmd_Bonazzola2004_abstracts.html
2004 .
(1 publication)M. Bonazzola and P. H. Haynes. A trajectory-based study of the tropical tropopause region. Journal of Geophysical Research (Atmospheres), 109:20112, October 2004. [ bib | DOI | ADS link ]
Large ensembles of 90-day backward trajectory calculations from the tropical lower stratosphere are performed for Northern Hemisphere (NH) winters 1997-1998 and 1998-1999 and summer 1999 on the basis of European Center for Medium-Range Weather Forecasts operational analysis data. The calculated trajectories are analyzed to determine patterns of transport and encountered temperatures and implications for lower stratospheric water vapor. For each set of back-trajectories, a troposphere-to-stratosphere (TS) ensemble, originating below 355 K, is identified. Trajectories in the TS ensemble sample the coldest regions of the tropical tropopause region very efficiently. Corresponding water vapor concentrations are calculated using two simple dehydration models, one (model 1) assuming instantaneous dehydration and the other (model 2) taking some account of time delays associated with microphysical processes. Model 1 predicts average concentrations for the TS ensembles of 1.5 and 2.0 ppmv in the two NH winters and 3.8 ppmv in NH summer. Model 2 predicts concentrations that are about 0.5 ppmv larger. The effect of temperature variability along the trajectories is considered and is shown to arise primarily through horizontal advection through strong gradients rather than through temporal variability. A quantitative method is described to assess the efficiency of sampling of cold regions, the roles played by different transport processes, and differences between seasons or years. Both vertical transport (the ”stratospheric fountain” effect) and horizontal transport are shown to play important roles in dehydration, with the former more important in NH winter and the latter more important in NH summer. Differences in predicted water vapor between NH winter 1997-1998 (El Niño) and 1998-1999 (La Niña) are due to the warmer region of coldest temperatures in 1997-1998 than in 1998-1999 and to the less efficient sampling of cold temperatures by both horizontal and vertical circulations during the former.