mesosphere from the Sounding of the Atmosphere using
Broadband Emission Radiometry (SABER) instrument launched in
2001 and the Aura Microwave Limb Sounder (MLS) launched in
2004 are the first to allow comprehensive evaluation of the
performance of models and data assimilation systems in the
upper stratosphere through the lower mesosphere. Recent
work with these data sets has shown that operational
analyses from European Center for Medium-Range Weather
Forecasting (ECMWF) and NASA's Global Modelling and
Assimilation Office (GMAO) show serious deficiencies in
reproducing extreme events such as the prolonged
stratospheric sudden warming in the 2005-2006 Arctic winter,
and show substantial biases in global stratopause structure
and evolution under more typical conditions. New research
assimilation products show promise in improving the
representation of the stratopause with higher model tops,
more sophisticated gravity-wave drag parameterizations,
and/or assimilation of MLS and SABER data. We show
improvements in stratopause structure in analyses that
assimilate MLS data into the GMAO assimilation system. The
interannual and interhemispheric variability of stratopause
evolution and its representation in data assimilation
systems is explored, focusing on recent assimilation
products from the Canadian Middle Atmosphere Model Data
Assimilation System (CMAM-DAS) and from Naval Research
Laboratory (NRL) NOGAPS-ALPHA runs assimilating MLS and
SABER temperatures.
(Work at JPL/CalTech is done under contract with the National Aeronautics and Space Administration.)