OMI Product Quality Assurance Checks

How do we measure quality?

Data quality is affected in several ways. They are:

• Instrument Errors
• Algorithm Errors
• Recording Errors

OMI scientists make sure that the quality is checked at every step. Several error flags are thus designed as indicators of the quality of the measured data. Examples: GroundPixelQualityFlags, AlgorithmFlags, MeasurementQualityFlags, QualityFlags, RadianceScienceQualityFlag, are all designed to indicate the type of data record for the particular collection. Recently a new flag XTrackQualityFlags is added to keep track of the pixels that are affected by row anomaly. Recording errors are minimized in OMI data since once the errors are found, they are immediately corrected and all the data are reprocessed at that instant.

How do we ensure data integrity?

Once a new set of data is processed, several validity checks are made to ensure the data quality. The hypothesis regarding the physical phenomena based under which the algorithm of data retrieval is designed, is also checked for accuracy.

Original data is always protected since only authorized personnel have access to the original data.

How can one use OMI data ?

OMI have several products of significant importance for climate studies. OMI products can be analyzed in local, regional and global scale to understand the trend of greenhouse gasses. OMI data can be used with meaurements by other satellites for validity checks. Atmospheric chemistry models and global circulation models can use OMI data for comparative analysis and understanding of modeling formulations.

How can we ensure the accuracy and reliability of the data?

OMI data is accurate under the conditions of the assumptions used for data retrieval. As the scientific knowledge is maturing everyday, OMI scientists are always alert to implement a better methodology that is available for data retrieval. Algorithms are timely updated. Attempts are made to minimize calibration errors by checking instrument conditions from time to time.