A New Release of Data from the Total Ozone Mapping Spectrometer (TOMS)
Charles G. Wellemeyer, BUV Algorithm Development, Lanham (MD), USA, (email@example.com)
P. K. Bhartia, R. D. McPeters, S. L. Taylor, Ch. Ahn, and the Ozone Processing Team
The Total Ozone Mapping Spectrometer (TOMS) data series extends from November 1978 through the present as shown in Figure 1. TOMS provides daily global maps of total column ozone over the sunlit portions of the globe by measuring backscattered ultraviolet radiances at discrete wavelengths in the 310380 nm region. The TOMS algorithm has evolved over the years, with the previous Version 7 data processing being released in 1996 [McPeters et al., 1996]. A new Version 8 of the algorithm has been developed by NASA Goddards Ozone Processing Team to address a number of issues that have accumulated since then. Broadly speaking, the enhancements in the new algorithm target errors that occur under extreme conditions. Definite improvements are apparent for conditions like high tropospheric aerosol loading, sun-glint, persistent snow/ice, and very high solar zenith angles. The bulk properties of the new Version 8 TOMS dataset do not differ significantly from the previous Version 7, though a calibration shift for the Earth Probe instrument changes the long term ozone trend somewhat. These improvements are most important for applications like derivation of tropospheric ozone or aerosol loading calculations that push the limits for TOMS accuracy.
The TOMS production system will be converted to Version 8 in early 2004, and the reprocessed Version 8 data for the Nimbus 7 and Earth Probe TOMS instruments will be placed in the Distributed Active Archive Center (DAAC) at NASA/GSFC.
Figure 1. Global average total ozone from the Version 8 Algorithm applied to Nimbus 7, Meteor 3, and Earth Probe TOMS calculated from 60° S to 60° N latitude. This new dataset will be released in early 2004.
The Version 8 TOMS Algorithm has a number of enhancements that are designed to reduce errors under extreme conditions. We have used corrections calculated using regression studies based on modelling of errors due to tropospheric aerosols, sun-glint, and ozone profile shape dependence at very high solar zenith angles. A season and latitude dependent ozone climatology has been used to reduce errors due to limited sensitivity of the backscattered ultraviolet radiance to ozone in the lower troposphere. Climatological temperature variability is taken into account explicitly using season and latitude dependent temperature profiles, and the forward model has been improved to account for the presence of persistent snow cover and/or high terrain. For a more detailed description of the V8 TOMS Algorithm, see the Algorithm Theoretical Basis Document (ATBD) and Data Users Guides available at http://toms.gsfc.nasa.gov.
Impact and Validation of Algorithm Changes
Some of the primary features of changes in the V8 TOMS dataset are illustrated by the monthly zonal mean comparisons shown in Figure 2. The changes at high latitude are the result a combination of effects including: hemisphere and month dependent tropospheric ozone and temperature climatology, improved cloud model for snowy scenes, and the new correction for profile shape errors. In the summer hemispheres, the changes are validated by comparison with measurements from the network of Dobson and Brewer instruments, which are fairly accurate out to ozone path lengths (total column ozone multiplied by the geometric optical path) of up to 1000 D.U. In the winter hemispheres, however, the impact of V8 modifications is primarily occurring at ozone optical paths considerably larger than 1000 D.U. We have addressed the need for validation of TOMS at higher path lengths by comparing V8 TOMS with integrated profiles from Solar Backscatter Ultraviolet (SBUV). These comparisons show good agreement, but a more complete validation is needed for the high path length data from V8 TOMS.
Figure 2. Version 8 Version 7 differences of Nimbus 7 TOMS monthly zonal mean total ozone for June and December of 1980 illustrating the impact of the Version 8 Algorithm.
The major change in the tropical regions is the correction for the effects of tropospheric aerosol and sun-glint. These effects are not apparent in Figure 2 because the impact of these corrections on the zonal mean is small. During extreme events of dust, smoke from biomass burning, or sun-glint, however, this correction may be as large as 20 D.U. or more, but these events are sporadic and spatially limited. We should point out that in the event of a very large volcanic eruption that injects aerosol into the stratosphere, the V8 TOMS is still subject to errors similar to those for the previous Version 7.
EP/TOMS Performance Issues
Beginning in mid-2000 it became clear that the EP/TOMS was exhibiting unanticipated changes in performance that manifested as a cross-track dependence in instrument sensitivity. Ozone measured when looking to the right of the orbit track was significantly lower than that measured when looking to the left. The physical mechanism causing this dependence is not understood at this time, though we suspect it results from some asymmetric degradation of the scanning mirror. We have developed a relative correction to remove the cross-track bias, but in addition we have found it necessary to normalize the overall EP/TOMS ozone trend to the SBUV2 after 2000. EP/TOMS data from launch through June 2003 with the corrections applied after June 2000 will be placed in the GSFC/DAAC in early 2004.
These corrections are difficult to make in near real-time, but we will update them from time to time to keep the ozone errors less than about 5% in the NRT data available through the TOMS Web Site. However, as we are able to do so we will also extend the corrected dataset beyond June 2003.
Data Products and Availability
In order to make the V8 TOMS Data Products compatible with Earth Observing System (EOS), we are archiving them in Hierarchical Data Format Version EOS-5 (or HE-5). We will place the TOMS Level-2 orbital data product for Nimbus 7, Meteor 3, and Earth Probe TOMS (Figure 1) into the GSFC/DAAC in early 2004.
The V8 TOMS Web Site (http://toms.gsfc.nasa.gov) will have a new look, but the format of the data provided there will not change. All of the same products available on the Version 7 web site will be available on the V8 web site in identical format. This will include the native Level-3 products and images, monthly average grids and images, zonal means, and ground station overpasses. The ATBD, Data Users Guides, and the new ozone climatology used in the V8 Algorithm will also be available at the web site.
We also plan to provide the V8 TOMS native Level-3 and images for the entire N7 TOMS and EP/TOMS through June 2003 on DVD, which will be made available through the GSFC/DAAC and will be distributed at scientific meetings as appropriate.
We gratefully acknowledge the support of NASA/GSFCs Ozone Processing Team in providing the TOMS data presented here. The SBUV/2 data were obtained from NOAA/NESDIS with support from the NOAA Climate and Global Change Atmospheric Chemistry Element.
McPeters, R.D., et al., 1996, Nimbus_7 Total Ozone Mapping Spectrometer (TOMS) Data Products Users Guide, NASA Ref-erence Publication 1384, National Aeronautics and Space Administration, Washington, DC.