Assessment of MIPAS ESA V7 Products And First Verification Of MIPAS ESA V8 Products

The Michelson Interferometer for Passive Atmospheric Sounding (MIPAS) on ESA's ENVISAT satellite performed almost continuous measurements of the global atmospheric composition for approximately 10 years, from July 2002 to April 2012. The operational ESA data retrieval processor, based on the algorithm ORM (Optimized Retrieval Model), originally designed for the Near Real Time analysis and developed by an European Consortium led by IFAC, is currently used for the reanalysis of the full MIPAS mission. The maintenance and the upgrade of the ESA processor are made in the frame of the Quality Working Group, where a fruitful collaboration among Level 1, Level 2 and validation teams can be exploited. This collaboration is essential to pursue improvements in the accuracy of the products and their characterization. This paper is meant to describe the most recent upgrades in the ESA processor performed to improve the quality of ESA products. In particular, the full mission was recently reprocessed with the L1 V7 and L2 V7 processors, and further improvements are in preparation, that will be collected in version 8 of the ESA processors. With respect to V6, V7 and V8 products take advantage of significant improvements in both L1 and L2 processors, as well as in the auxiliary data. Improvements in the L1 processor consist in the correction of the instrumental drift, improved spike detection algorithm and new Instrument Line Shape, as well as in the use of measured daily gain calibration instead of weekly gain. Improvements in the L2 processor include a different approach for retrieving atmospheric continuum, the use of an a posteriori regularization with altitude dependent constraint, a better approach for handling interfering species, a reduced bias in CFC-11, the handling of horizontal inhomogeneities and the use of ECMWF altitude/pressure relation for determining more accurate tangent altitudes. Improvements in the auxiliary data consist in the use of microwindows with larger information content, more accurate non-LTE errors, new spectroscopic database and diurnally varying climatological dataset. Furthermore, with each new version additional trace-gases are provided, leading to 20 retrieved species by the L2 V8 processor. Improvements in the V7 products will be revised in the light of the results of the validation with correlative measurements, and, by comparing the first new L2 V8 products with the L2 V7 ones, a preliminary assessment of the performance of the new V8 processors will be presented