Advances in remote sensing are generating a huge range of new measurements of chemical species in the middle atmosphere. These measurements compliment existing measurements from ozone sondes and lidar. Measurements from satellites generally give better spatial coverage, but the precision and calibration may be less stable than ground based measurements. Intercomparison of a wide range of independent measurements provides a rigorous check on the systematic differences between different instruments.

Gridded MIPAS ozone data

Gridded stratospheric ozone fields (netcdf format), with approximately 2 degree spatial resolution, 4 hourly and on 10 isentropic levels are available here.

Manuscript in open discussion

A description of the method has been published in the Atmospheric Chemistry and Physics Discussion online jorunal, and can be downloaded and commented on here


Different measurments have often be compared by nearest neighbour comparisons. More recently, data assimilation is being used to generate global fields from a observation set which can then be compared with all available independent measurements. Both approaches are used in this project, and it is shown that the data assimilation gives more precise answers because of its ability to represent the spatial and temporal variations in the fields being observed and the contribution these variations make to differences between observations which are near but not coincident.


Acronym Full name Type Time
GOME Global Ozone Monitoring Experiment global May 1995, ongoing
GOMOS Global Ozone Monitoring by Occultation of Stars global Mid 2002, ongoing
HALOE Halogen Occultation Experiment partial November 1991 to April 2003
HIRDLS High Resolution Dynamics Limb Sounder global Expected 2004
ILAS Improved Limb Atmospheric Spectrometer partial August 1996 to June 1997
ILAS II Improved Limb Atmospheric Spectrometer II partial December 2002, ongoing
MIPAS Michelson Interferometer for Passive Atmospheric Sounding global Mid 2002, ongoing
MLS, UARS Microwave Limb Sounder global October 1991 to June 1997
MLS, EOS Microwave Limb Sounder global Expected 2004
POAM II Polar Ozone and Aerosol Measurement partial 1993 to November 1996
POAM III Polar Ozone and Aerosol Measurement partial March 1998, ongoing
SAGE II Stratospheric Aerosol and Gas Experiment II partial October 1984 to June 2001
SAGE III Stratospheric Aerosol and Gas Experiment III partial 2002, ongoing
SBUV II Solar Backscatter Ultraviolet global 1984, ongoing
TOMS Total Ozone Mapping Spectrometer column ongoing
TOVS TIROS Operational Vertical Sounder column ongoing
Ozonesondes groundbased partial ongoing
Umkehr partial ongoing
Lidar partial ongoing
Intercomparisons have so far been carried out with MIPAS, HALOE, POAM III, SAGE III and GOME.


Comparisons between assimilated MIPAS ozone data and observations from HALOE, POAM III and SAGE III are summarised in figure 1. The plot shows 30 bars charts, corresponding to 5 levels and 6 latitudinal bands. Each bar chart show the mean difference between available observations and the assimilation. A black bar indicates twice the variance about this mean normalised by the square root of the number of observations. From 850K to 1600K POAM II and SAGE III are consistent with MIPAS to within 3%. HALOE, on the other hand, is measures significantly lower values at upper levels in this period. Generally, the plot is consistent with MIPAS having zero bias. If a similar plot is prepared using nearest neighbour comparisons, with a cut off of 12 hours and 5 degrees, the variances and mean differences are much larger (figure 2).