Improvements in stratospheric data assimilation

Data assimilation produced winds have been successfully used in driving off line transport chemistry models. For accurate results, the assimilation winds must have realistic transport characteristics. Both parts of the data assimilation system (DAS), the analysis scheme and the general circulation model (GCM), are important to obtain realistic transport. The analysis scheme must insert observations without unduly disturbing transport fluxes, and general circulation model (GCM) must represent data void regions as well as provide consistent unobserved fields, such as vertical velocity. Wind observations are sparse in the stratosphere and, while the analysis scheme can provide good approximations to the balanced extratropical winds based on satellite temperature observations, the GCM carries most of the burden of producing winds in the tropics and subtropics.

Results are presented from the latest assimilation products and tests done at Goddard Space Flight Center's Data Assimilation Office. Several changes in the DAS have been designed to improve stratospheric assimilations, including a higher vertical top, better vertical resolution, addition of a gravity wave drag parameterizaion, and removal of biases in satellite temperature retrievals. These have led to a much improved representation of the quasi-biennial oscillation, a more accurate tropical tropopause temperatures, and some improvement in potential vorticity analyses. In addition, results from the Goddard chemistry transport model show improved descent in the northern hemisphere winter middle atmosphere and sharper subtropical tracer gradients.