Three Recent Data Anomalies

Three recent data anomalies have been observed in the OMI measurements in the VIS channel. These anomalies manifest themselves as suppressions in the radiance spectra, and are currently under investigation.

• Between 25 and 27 June 2007, a change occurred in the Ozone Monitoring Instrument such that in rows 54-55 (1-based) a decrease in the radiance signal of 20-30% under illuminated conditions is observed over the complete illuminated part of the orbit. All wavelengths in the VIS (and UV2) channel are affected by this effect.
• Since 11 May 2008, another signal suppression has been observed in rows 38-41 of the VIS (and UV2) channel at solar zenith angles larger than 44.0 degrees (though data at lower SZAs is also suspect). It is recommended not to use the affected scenes at all latitudes and SZAs for now.
• On 24 and 25 January 2009, it was observed that a signal suppression now affects rows 27-46 over the complete illuminated part of the orbit. It is recommended not to use the affected data for now.

The anomalies are currently under investigation to examine whether correction for the signal suppressions can be implemented in the level 1b data. No correction has been implemented at the moment. It is recommended that the rows in question are not used for scientific data analysis from the start of the anomaly onwards until a correction in the level 1b data has been implemented.

Because current (Collection 3) level 1b and level 2 data, including the DOMINO data product, have not been flagged or corrected for these anomalies, data users should be cautious in using the OMI data. One effective way to ensure meaningful data analysis is by excluding all the data from the affected rows from the reported start date of the anomaly onwards. As an alternative, data users may set a flag based on the reported uncertainty in the NO₂ slant column density. Preliminary results suggest that slant column errors (SlantColumnAmountNO2Std) > 1.7·10¹⁵ molecules cm^-2 are strongly correlated with the affected rows.

For more information on the status of the OMI instrument and the level 1b data, please check this page