• Lagrangian transport;
  • tropical tropopause

[1] Winter-summer differences in the transport of air from the boundary layer to the lower stratosphere at low latitudes are investigated with ensembles of back trajectory calculations that track parcels from the 380 K isentropic surface to their convective detrainment in the tropical tropopause layer (TTL) during the winter of 2006–2007 and summer of 2007. Horizontal displacements for the trajectories are calculated from reanalysis data; potential temperature displacements are calculated from radiative heating rates derived from observed cloud, water vapor, ozone, and temperature variations; and the locations' convective detrainments are determined by satellite observations of convective clouds. Weaker upwelling in the TTL during boreal summer compared with that of winter both slows the ascent through the TTL and raises the height threshold that convective detrainment must surpass in order for ascent to occur, restricting the injection of new air into the stratosphere during summer. In addition, anticyclonic circulations associated with convective activity contribute to vertical transport in the TTL by guiding detrained air parcels through regions with the strongest upwelling. These features combine to make monsoon-related convection over the Indian subcontinent the dominant source of new air during summer. In contrast, winter sources are spread over the southern continents and the western Pacific Ocean. These seasonal differences imply that air entering the tropical stratosphere during summer is older but might nevertheless be more polluted than air entering during winter. While poor data sampling in the TTL makes it difficult to validate our results, they are bolstered by favorable comparisons with previous studies of the TTL, by sensitivity tests that reveal important dynamical influences on surface-to-stratospheric transport, and by the robustness of dynamical interactions that systematically associate deep convection with anticyclonic circulations and strong radiative heating in the TTL. Sensitivity experiments suggest that the aforementioned seasonal differences are sensitive to strong “large-scale” (on global space scales and seasonal time scales) perturbations. In particular, uncertainties in the vertical motion fields constrain our ability to draw definitive conclusions. However, trajectory statistics are not sensitive to small-scale perturbations, with the encouraging implication that our results are primarily associated with those features of the circulation that are the most likely to be robust.