Product Specification Document for SCIAMACHY Stratozone Limb ozone profiles

Date of origin: February 6, 2006

Author: Christian von Savigny

Institute of Environmental Physics

University of Bremen

Otto-Hahn-Allee 1

28334 Bremen

e-mail:

Product Specification Document for SCIAMACHY Stratozone

Limb ozone profiles

Date of origin: February 6, 2006

Author: Christian von Savigny

Institute of Environmental Physics

University of Bremen

Otto-Hahn-Allee 1

28334 Bremen

e-mail:

1. Product description

The data product essentially consists of stratospheric profiles of ozone concentration. The retrieval is performed on an altitude grid ranging from 2 km to 80 km, BUT it is important to realize, that the method applied is sensitive to ozone only between about 15 and 40 km. Above and below, the retrieved concentrations approach the a priori concentrations. Information on the sensitive altitude range can be obtained from the averaging kernels, which are available for every single retrieved profile. Also provided are the a priori ozone profiles, as well as geolocation information. The ozone concentrations are given in units of molecules cm-2 km-1. To convert to molecules cm-3 the listed values have to be divided by 105. The values given at a certain altitude level correspond to the retrieved concentration exactly at that altitude level. Within the 15 – 40 km altitude range the retrieval altitude grid is a 1 km altitude grid. The retrievals are performed on the dayside of the Earth for solar zenith angles smaller than 88º.

The solar viewing angles listed in the output files are given at the tangent point, not at the top of atmosphere (at the point where the line of sight intersects the 100 km altitude layer). The solar azimuth angle is given with respect to the line of sight. The geographical coordinates listed in the output files correspond to the coordinates of the tangent point averaged over the detector integration time for each state or subpixel (if available) at a tangent height of 30 km. This tangent height was chosen, since it is roughly in the middle of the retrieval range for stratospheric ozone profiles with the method used here.

The algorithm is described in the Algorithm Document for SCIAMACHY Stratozone limb ozone profile retrievals, February 2006. [von Savigny, 2006]

2. Product format specification

The data product consists of the following files:

1.1. One control file per Level 1 file containing all the parameters the radiative transfer model is run with :

control_YYYYMMDD_OOOOO_NNNN.dat

YYYY: Year

MM: Month

DD: Day

OOOOO: 5-digit orbit number

NNNN: 4-digit reference number required for the identification of the Level 1 files

1.2 One result overview file per Level 1 file containing the geolocation values for all ozone profiles retrieved:

YYYYMMDD_OOOOO_NNNN_ls0.inf

File format:

# Date Orb#1 Orb#2 Start (UTC) Meas# SubPix# SubPix Lat Lon SZA Azimuth

20050921 18614 3443 11:20:45 1 1 1 51.78 -10.58 53.62 37.66

20050921 18614 3443 11:22:50 2 1 1 44.60 -13.62 47.90 41.94

20050921 18614 3443 11:24:54 3 1 1 37.36 -16.10 42.57 47.46

20050921 18614 3443 11:26:59 4 1 1 30.09 -18.23 37.77 54.54

20050921 18614 3443 11:29:03 5 1 1 22.80 -20.14 33.75 63.58

…

The data lines contain list the date, orbit number, reference orbit number, UTC at start of state, number of state in file, number of azimuthal subpixels, present number of azimuthal subpixel, latitude, longitude, solar zenith angle, and solar azimuth angle.

1.3 Each individual retrieved ozone profile is stored in a separate file:

YYYYMMDD_OOOOO_NNNN_XX_ZZ_O3_ls0.dat

XX: 2-digit state number

ZZ: 2-digit number of subpixels per state

File format:

# Retrieval code by C. von Savigny, IUP Bremen, Version 1.62

20050921 18607 2620 01:06:57 6 1 1 81.15 -144.30 84.17 27.35

2.0000 0.1504E+18 0.1521E+18 0.0000E+00

4.0000 0.1172E+18 0.1198E+18 0.0000E+00

6.0000 0.1103E+18 0.1150E+18 0.0000E+00

…

The 2nd header line contains the following pieces of information: Date (YYYYMMDD), Orbit number, orbit reference number, UTC at start of state, state ID, number of azimuthal subpixels (up to 4), present number of azimuthal subpixel, latitude, longitude, solar zenith angle, solar azimuth angle.

The following data lines contain: altitude in km, retrieved ozone concentration in molecules cm-2 km-1, a priori ozone concentration in molecules cm-2 km-1, error estimate (presently not available).

1.4 The averaging kernel matrix for every retrieved profile is stored in a separate file:

A_kernel_YYYYMMDD_OOOOO_NNNN_XX_ZZ.dat

3. Software release history

Listed are the changes w.r.t the previous version for the most recent versions.

Version 1.63:

- Retrieval of azimuthally separated profiles implemented (i.e., up to 4 ozone profiles per limb state)

- Tangent height retrieval based on TRUE (Vs. 1.4) performed before ozone profile retrieval

Version 1.62:

- Solely based on Level 1 data, previous versions were entirely or partly based on Level 0 data

- Retrieval performed on a 1 km altitude grid. In previous versions the altitude grid was chosen to be identical to the tangent height grid, i.e. every profile had a different altitude grid.

4. Implementation details

The SCIAMACHY Level 1 data are calibrated using the SciaL1c flags 0,1,2,3,4,5, corresponding to: memory effect correction, leakage current correction, straylight correction, pixel-to-pixel-gain correction, etalon correction, and wavelength calibration. Flags 6 and 7 corresponding to polarization correction and absolute calibration are omitted.

5. List of known issues

5.1 Limb pointing:

5.1.1 There are several different indications that the TRUE tangent height retrievals show a systematic or slowly varying offset of about 500 m. The actual tangent height offsets appear to be larger than the values produced by TRUE. This is most likely due to differences between the true upper stratospheric/lower mesospheric ozone profile and the climatological profiles used for the TRUE retrievals.

5.1.2 TRUE only allows for a tangent height retrieval at tropical latitudes. The pointing correction performed for the ozone profile retrievals consists of applying the tropical pointing offset to all states of the same orbit. This implies that a possible pole-to-pole variation of the pointing errors cannot be corrected for.

5.2 Unrealistic features and other issues

5.2.1 The Brinskma et al. [2005] study showed that the Stratozone version 1.61 profiles exhibit an unrealistic zigzag shape from one tangent height step to the next. The amplitude of this oscillation was only a few percent, but clearly visible in the comparisons with Lidars and SAGE II. The oscillations are most likely related to the fact that in version 1.61 the retrieval altitude grid was identical to the tangent height grid. To mitigate this effect, versions 1.62 and higher use a 1 km altitude grid.

5.2.2 Most comparison studies showed a small systematical low bias of the Stratozone profiles (between 3 and 6 % between 20 and 35 km when compared to Lidars and SAGE II as shown in Brinskma et al. [2005]).

6. Data quality assessment

Unfortunately, the most recent version (1.63) has not been as comprehensively validated as some of the previous verions. All upgrades to version 1.63 were made in order to improve the data quality, but unexpected consequences on the ozone profiles cannot be excluded. With a pointing correction being performed on an orbit-by-orbit basis for version 1.63, the accuracy of the ozone concentrations should be better than about 10 – 15 % between 20 and 40 km. Any obvious errors or unrealistic features of the profiles should be reported to .

7. References

E. J. Brinksma, A. Bracher, D. E. Lolkema, A. J. Segers, I. S. Boyd, K. Bramstedt, H. Claude, S. Godin-Beekmann, G. Hansen, G. Kopp, T. Leblanc, I. S. McDermid, Y. J. Meijer, H. Nakane, A. Parrish, C. von Savigny, D. P. J. Swart, G. Taha, and A. J. M. Piters, Geophysical Validation of SCIAMACHY Limb Ozone Profiles, Atmos. Chem. Phys., 6, 197 – 209, 2006.

von Savigny, C., Algorithm document for SCIAMACHY Stratozone limb ozone profile retrievals, 2006.