A Trajectory Study of the Seasonal Cycle of the Tropical Lower Stratospheric Water Vapor in Chemistry-Climate Models in Comparison with Observations

In this study, we evaluate how well the Goddard Earth Observing System Chemistry Climate Model (GEOSCCM) and the Whole Atmosphere Community Climate Model (WACCM) reproduce the seasonal cycle of the tropical lower stratospheric water vapor in the Microwave Limb Sounder (MLS) and the European Centre for Medium-Range Weather Forecasts (ECMWF) ERA-Interim (ERAi). We also evaluate how well the chemistry-climate models (CCMs) reproduce the key processes that regulate the seasonal cycle using a forward, domain filling, diabatic trajectory model. The seasonal cycle from the MLS, the ERAi, and the CCMs show general agreement. The troposphere-to-stratosphere transport of water vapor in the trajectory model driven by the GEOSCCM (traj-GEOSCCM) and the trajectory model driven by the WACCM (traj-WACCM) show differences when compared to the trajectory model driven by the ERAi (traj-ERAi). The traj-GEOSCCM underestimates the contribution to stratospheric water vapor from the Asian monsoon region during summer, while the traj-WACCM run overestimates the contribution to stratospheric water vapor from the Tropical West Pacific throughout the year. The final dehydration point (FDP) of parcels in the traj-ERAi run is well reproduced in the GEOSCCM run. The traj-WACCM run, however, overestimates the FDP density in the Tropical West Pacific and underestimates it in Tropical Africa and South America. The traj-WACCM run also shows a higher average FDP altitude. Both the traj-GEOSCCM and traj-WACCM do a good job reproducing the contribution to the seasonal oscillation of the tropical lower stratospheric water vapor in the traj-ERAi. Finally, both the CCMs predict that the seasonal cycle will be moister throughout the year by the end of the 21st century. The mean of the seasonal cycle increases by 0.93 ppmv (25%) in the GEOSCCM and 1.85 ppmv (41%) in the WACCM. The amplitude of the seasonal cycle in the GEOSCCM is predicted to increase by 0.64 ppmv (36%), while that in the WACCM will decrease by 0.32 ppmv (28%). The trajectory model driven by the two CCMs underpredict the changes in the moisture and amplitude of the seasonal cycle from the CCMs