Data‎ > ‎

Total column ozone

Last updated: 4 September 2018

Version 3.4 the National Institute of Water and Atmospheric Research - Bodeker Scientific (NIWA-BS) total column ozone database is available.

Version 3.4 of the NIWA-BS total column ozone (TCO) database is now available through a project funded by the New Zealand Deep South National Science Challenge. This version of the database improves on Version 3.3 by improving the screening of the raw satellite data and removing anomalous data prior to their use in the process that splices together the different data sets. As with previous versions, the database is constructed by combining measurements from a number of different satellite-based instruments and extends from 1 November 1978 to 31 December 2016. Offsets and drifts between the different data sets are resolved through comparisons with the Dobson and Brewer ground-based instruments and through inter-satellite instrument comparisons. 

Version 3.4 of the database is available here and a paper describing the new database is in preparation for submission to Earth System Science Data. You will note that the 'Patched' data set, where all missing values are replaced by estimates, is not available on the FTP server. We are still in the process of generating those filled fields. Previous versions of the database had problems with the data filling as a result of over fitting. We have now implemented a much more advanced, machine-learning-based algorithm to fill in the missing data but it takes a long time to run. We expect that processing to be finished by late October 2018. We will make the filled data sets available shortly thereafter.

To access the data, you will need a username and password. These can be obtained by emailing datasets@bodekerscientific.com. If you would like to be informed of updates to this database, please email me at greg@bodekerscientific.com and I will add you to my list. 

The daily data are available in NetCDF files with one file holding a year of data. The NetCDF files include daily fields of total column ozone, the number of measurements used to derive the daily total column ozone in each grid cell, and the 1 sigma uncertainties on each total column ozone measurement. NetCDF files of monthly mean and annual mean fields are also available.

A paper is in the process of being written to describe this new database. Descriptions of previous versions of the data set are available in a number of publications that have made use of the data:

  • Struthers, H., G. E. Bodeker, J. Austin, S. Bekki, I. Cionni, M. Dameris, M. A. Giorgetta, V. Grewe, F. Lefèvre, F. Lott, E. Manzini, T. Peter, E. Rozanov, and M. Schraner (2009), The simulation of the Antarctic ozone hole by chemistry-climate models, Atmospheric Chemistry and Physics, 9, 6363–6376.  [acp-9-6363-2009.pdf].
  • Müller, R., J.-U. Grooß, C. Lemmen, D. Heinze, M. Dameris, and G. E. Bodeker (2008), Simple measures of ozone depletion in the polar stratosphere, Atmospheric Chemistry and Physics, 8, 251-264. [acp-8-251-2008.pdf]
  • Bodeker, G. E., H. Shiona, and H. Eskes (2005), Indicators of Antarctic ozone depletion, Atmospheric Chemistry and Physics, 5, 2603-2615.  [acp-5-2603.pdf]
If you are going to be using this database in a publication, please let me know (greg@bodekerscientific.com). At the very least please include the following acknowledgement: We would like to thank Bodeker Scientific, funded by the New Zealand Deep South National Science Challenge, for providing the combined NIWA-BS total column ozone database.

Changelog

Version 3.3 (October 2017):
This update to previous versions of the database included:
  • New and updated sources of satellite-based TCO measurements were used viz. data from NPP-OMPS (National Polar-orbiting Partnership-Ozone Mapping and Profiler Suite), GOME-2 (Global Ozone Monitoring Experiment-2) and SCIAMACHY (Scanning Imaging Absorption Spectrometer for Atmospheric CHartographY) are now included in the combined data set.
  • Improved statistical methods are used to model the difference fields between data sets; zonal means of the difference fields are modelled using Legendre expansions which comprise the meridional component of a spherical harmonic expansion which is best suited for statistically describing a field on a sphere.
  • Measurement uncertainties on the source data sets, and the corrections applied to those data sets, have been collated and are propagated through to the combined ozone data set so that, for the first time, this data set is now provided with uncertainty estimates for each datum.
Version 3.2 (January 2017):
Updated the V3.0 methodology to more robustly model the difference fields between data sets at the poles. A patched data set was also provided

Version 3.0 (November 2015):

Version 3.0 of the NIWA-BS total column ozone (TCO) database has been constructed in a project funded by the New Zealand Deep South National Science Challenge. The database is constructed by combining measurements from a number of different satellite-based instruments. Offsets and drifts between the different data sets are resolved through comparisons with the Dobson and Brewer ground-based instruments and through inter-satellite instrument comparisons. This update to previous versions of the database includes:

  • New and updated sources of satellite-based TCO measurements are used viz. data from NPP-OMPS (National Polar-orbiting Partnership-Ozone Mapping and Profiler Suite), GOME-2 (Global Ozone Monitoring Experiment-2) and SCIAMACHY (Scanning Imaging Absorption Spectrometer for Atmospheric CHartographY) are now included in the combined data set.
  • Improved statistical methods are used to model the difference fields between data sets; zonal means of the difference fields are modelled using Legendre expansions which comprise the meridional component of a spherical harmonic expansion which is best suited for statistically describing a field on a sphere. A Bayesian information criterion is used to determine the number of terms used in the Legendre expansions.
  • Measurement uncertainties on the source data sets, and the corrections applied to those data sets, have been collated and are propagated through to the combined ozone data set so that, for the first time, this data set is now provided with uncertainty estimates for each datum.
This new version of the database is available here. The data are supplied as unpatched (no spatial or temporal interpolation) CF-Compliant NetCDF files. 

At this time only daily 'unpatched' data are available. We are working on generating monthly mean and patched data files as had been available in previous versions of the database. This is now a little more challenging as we intend to capitalize on the uncertainty estimates being available to calculate monthly means and patched data that incorporate realistic uncertainties. If you need the monthly mean or patched data, please continue to use version 2.8 of the database for now (see below).

Version 2.8:

Version 2.8 of the total column ozone database made the following changes to previous versions:

  • corrections applied to the OMI data for their offset and drift against the ground-based Brewer and Dobson network have been recalculated based on the longer overlap period now available.
On the FTP server  for version 2.8 of this database you will see three different directories, viz.:

Daily: These are the data at daily resolution. There are, in turn, three subdirectories, viz.:

LongPatched: In these files, longitudinal interpolation has been performed to fill data gaps. However, no interpolation is performed over more than 60o in longitude and 5o in latitude. In addition, temporal interpolation was performed when missing data were bracketed by available data at the same location on the preceding and following days.

Patched: In these files spatial linear interpolation has been performed to fill data gaps. However, no interpolation is performed over more than 60o in longitude and 10o in latitude. However, these criteria are relaxed poleward of 60o to permit interpolation over the polar caps. This interpolation method is described and validated in: Bodeker, G. E., B. J. Connor, J. B. Liley, and W. A. Matthews (2001), The global mass of ozone: 1978-1998, Geophys. Res. Lett., 28(14), 2819-2822. No temporal interpolation is performed.

Unpatched: In these files no spatial or temporal interpolation is performed.

All three of these data sets are available as NetCDF files. I have tried to make them CF compliant so that they can be used directly e.g. in the SPARC CCMVal activity.

Locations: These are daily data extracted from the daily resolution grid files using bilinear interpolation. Each file is a text file and contains a header describing the file contents. The values in the last column describe the extent of the interpolation required to derive a valid value. If a valid value was available from the 'Unpatched' daily grid files, this value will be 0. If it was necessary to go to the 'LongPatched' daily grid files, this value will be 1. If it was necessary to go to the 'Patched' daily grid files this value will be 2. If no value could be derived, this value will be -1000. If a location that you want data for is not currently available, please email me at greg@bodekerscientific.com with the name, latitude and longitude of the station you want data for, and I will see what I can do.

Monthly: These are data at monthly resolution. There have to be at least 25 valid values within a month to calculate a valid monthly mean. As with the daily data, there are three subdirectories containing data in which different degrees of spatial and/or temporal interpolation has been applied. Read the explanatory notes in the daily data section to see what these are. They are available both as NetCDF and as text files.

Text files: These are files in a raw text format. Within each annual file there is a block of data for each month. The first line in each monthly data block gives the month number of the year. Note that there may be missing months of data. The next 180 lines give the data. The first line is for the cells between 89oN and 90oN and the last line is for the cells between 89oS and 90oS. Each line of data holds 288 numbers. Each value is always three digits with -99 meaning no data available. There are no spaces between values to save on file size. The first value on each line is for the cell between 180oW and 178.75oW (each cell is 1.25 degrees in width). The last value on each line is for the cell between 178.75oE and 180oE.