The quality of ozone data produced from the Electrochemical Concentration Cell (ECC) ozonesonde depends on the care exercised in their pre-flight preparation. The ozonesonde flown from Goddard Space Flight Center's, Wallops Flight Facility undergoes a number of preparation mechanisms designed to find and remove anomalies that could compromise valid ozone data. Complete electronic testing of the instrument, individually, and when coupled to its radiosonde, reduces instrument failures during flight. A number of other factors considered are, specific pump efficiency calibration, volumetric flow rate checks, temperature of the air being sampled, and background current. These are thoroughly checked and results recorded. A UV photometer measurement of the same ozone concentrations used to calibrate the ECC is compared and an adjustment factor is added to the ozone equation. Similar procedures are employed in Brazil, and Ascension Island, where NASA has cooperative agreements in place to obtain ozonesonde data for monitoring stratospheric changes and for the SHADOZ project. ECC instruments are typically prepared 3-4 weeks prior to the observation day. We briefly describe instrument comparison tests of simultaneous observations to reconcile KI concentration differences, to compare sensors of different manufacturers, and between surface- and remote instrumentation. Comparisons between ECC and HALOE measurements, ECC integrated total ozone overburden with EP-TOMS and Dobson Spectrophotometers show reasonable agreement. We postulate reasons for the differences, or biases, when these occur.

In spite of the work conducted in situ, results from laboratory testing during JOSIE2000, carried out at the Institute for the Study of Chemistry of the Polluted Atmosphere located in Jülich, Germany, are particularly interesting. ECC measurements from seven independent laboratories conducted in the high-altitude pressure chamber yielded variable results. This suggests that a considerably larger number of ECC ozonesondes, whether compared in a chamber or in situ, would improve the statistical record and provides greater confidence in ECC products.