Near-real time retrieval of tropospheric NO₂ from OMI

Abstract. We present a new algorithm for the near-real time retrieval – within 3 h of the actual satellite measurement – of tropospheric NO₂ columns from the Ozone Monitoring Instrument (OMI). The retrieval is based on the combined retrieval-assimilation-modelling approach developed at KNMI for off-line tropospheric NO₂ from the GOME and SCIAMACHY satellite instruments. We have adapted the off-line system such that the required a priori information – profile shapes and stratospheric background NO₂ – is now immediately available upon arrival (within 80 min of observation) of the OMI NO₂ slant columns and cloud data at KNMI. Slant columns for NO₂ are retrieved using differential optical absorption spectroscopy (DOAS) in the 405–465 nm range. Cloud fraction and cloud pressure are provided by a new cloud retrieval algorithm that uses the absorption of the O₂-O₂ collision complex near 477 nm. On-line availability of stratospheric slant columns and NO₂ profiles is achieved by running the TM4 chemistry transport model (CTM) forward in time based on forecast ECMWF meteo and assimilated NO₂ information from all previously observed orbits. OMI NO₂ slant columns, after correction for spurious across-track variability, show a random error for individual pixels of approximately 0.7×10¹⁵ molec cm⁻². Cloud parameters from OMI's O₂-O₂ algorithm have similar frequency distributions as retrieved from SCIAMACHY's Fast Retrieval Scheme for Cloud Observables (FRESCO) for August 2006. On average, OMI cloud fractions are higher by 0.011, and OMI cloud pressures exceed FRESCO cloud pressures by 60 hPa. A sequence of OMI observations over Europe in October 2005 shows OMI's capability to track changeable NO_x air pollution from day to day in cloud-free situations.