ConcordIASI is a multi-year international project whose objectives include improving analyses and forecasts over polar regions. The field phase of ConcordIASI during Austral spring 2010 provided unprecedented meteorological data coverage over the southern polar region, with dropsondes launched from 19 super-pressure balloons providing over 30,000 measurements of wind, temperature and humidity at levels up to 60 hPa over a three-month period. The impact of assimilating these data in the context of NWP is a topic of ongoing investigation at several forecast centers.

In this study we use the adjoint data assimilation systems of NASA/GMAO and ECMWF to examine the impact of ConcordIASI dropsonde data on 24-h forecasts over the southern polar region during Sep-Dec 2010. The dropsonde impacts are compared with those of other in-situ and satellite observing systems, and the results aggregated by latitude, vertical pressure level and observed variable. Results show that the dropsondes, which are relatively few in number compared with most other data types in the assimilation system, have a small positive impact in an overall sense, but a relatively large positive impact per observation. Dropsonde temperature observations poleward of 80°S and below 400 hPa have the largest impacts, although both temperature and wind observations have significant impact in most locations. The spatial and temporal variability of the dropsonde impacts are compared with those of other observation types, and model biases that hinder data assimilation in the polar region are identified.