1. Page 1 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 ACVT-GBMCD subgroup GOMOS ozone profiles, analysis of comparison with GMBCD datasets (bright/dark, star magnitude, star temperature) Yasjka Meijer, RIVM yasjka.meijer@rivm.nl

2. Page 2 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 Validation Team O 3 profiles Validation teamPI-nameInstituteInstrument  AO 153S. PalSAAI/MSCLidar  AO 158J.-C. LambertBIRA Microwave/lidar/sondes  AO 179A. MatthewsNIWAMicrowave/sondes  AO 191T. BlumenstockINTAFTIR  AO 300 D. De MuerRMISondes  AO 360P. KeckhutCNRSLidar  AO 429E. KyroFMISondes  AO 1103A. PetritoliISACSondes  AO 9003D. SwartRIVMLidar

3. Page 3 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 Intercomparison of Ozone Profiles Geolocation criteria:  lidar (< 24 hrs, 1000 km)  sonde (< 24 hrs, 1000 km)  microwave radiometer (< 4 hrs, 1000 km) GOMOS data:  from ACRI prototype processor  added solar zenith angle at tangent point GBMCD data:  collocations provided by AO-teams  all files available from NILU database  all data converted to ozone number density vs altitude

4. Page 4 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 Geolocation of GOMOS & GBMCD data Total collocated GOMOS-GBMCD paired profiles:226 no altitude overlap: - 13 missing files GOMOS:- 82 available for analysis 131

5. Page 5 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 Example 1 Lauder lidar

6. Page 6 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 Example 2 Toronto lidar

7. Page 7 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 Example 3 Uccle sonde

8. Page 8 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 Important GOMOS parameters Sun position (SZA)  dark (110 o -180 o )  twilight (90 o -110 o )  bright (0 o -90 o ) Star temperature (K)  hot (7,000- 100,000)  cold (1,000-7,000) Star magnitude (MV)  strong (-2 to 1)  weak (1 to 5) Less signal from weaker stars More strayligh t Less UV in colder stars

9. Page 9 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 LIDAR measurements vs GOMOS

10. Page 10 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 All data Lidar N = 57

11. Page 11 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 BRIGHT Lidar N = 4

12. Page 12 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 TWILIGHT Lidar N = 13

13. Page 13 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 DARK Lidar N = 40

14. Page 14 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 DARK STRONG Lidar N = 5

15. Page 15 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 weak DARK Lidar N = 35

16. Page 16 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 COLD DARK Lidar N = 19

17. Page 17 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 HOT DARK Lidar N = 21

18. Page 18 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 Conclusions vs LIDAR :  bright limb ozone profiles: poor results  twilight limb ozone profiles: better than bright limb, but still large deviations (cause to be determined)  dark limb ozone profiles: good results  no systematic biases between 18-45 km  no clear influence of star magnitude or temperature

19. Page 19 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 SONDE measurements vs GOMOS

20. Page 20 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 All data Sonde N = 39

21. Page 21 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 BRIGHT Sonde N = 26

22. Page 22 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 DARK Sonde N = 13

23. Page 23 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 Only 1 strong STAR (of 13), with MV<1 Sonde

24. Page 24 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 COLD DARK Sonde N = 9

25. Page 25 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 HOT DARK Sonde N = 4

26. Page 26 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 Conclusions vs SONDE :  bright limb ozone profiles: poor results  twilight limb ozone profiles: no cases  dark limb ozone profiles: good results  small systematic bias of 5-10% between 18-35 km (GOMOS lower)  star magnitude: no info  star temperature: below 22 km cold better than hot and above vice versa(?, more statistics needed)

27. Page 27 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 MICROWAVE measurements vs GOMOS

28. Page 28 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 All data Microwave N = 35 Note: lower mesosphere included

29. Page 29 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 BRIGHT Microwave N = 23

30. Page 30 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 DARK Microwave N = 12

31. Page 31 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 NO strong STARS, with MV<1 Microwave

32. Page 32 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 COLD DARK Microwave N = 4

33. Page 33 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 HOT DARK Microwave N = 8

34. Page 34 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 Conclusions vs MICROWAVE :  bright limb ozone profiles: poor results  twilight limb ozone profiles: no cases  dark limb ozone profiles:  (all stars) 20-45 km bias within 20%  (cold stars) 45-65 km : poor results  (hot stars) 45-65 km bias within 20%  (hot stars) 45-65 km significant non-random bias suggests possibility for improvement  star magnitude: no info

35. Page 35 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 All instruments compared to GOMOS

36. Page 36 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 All data All instruments N = 131

37. Page 37 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 BRIGHT All instruments N = 53

38. Page 38 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 BRIGHT STRONG All instruments N = 9

39. Page 39 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 DARK All instruments N = 65

40. Page 40 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 DARK STRONG All instruments N = 6

41. Page 41 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 weak DARK All instruments N = 59

42. Page 42 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 COLD DARK All instruments N = 32

43. Page 43 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 HOT DARK All instruments N = 33

44. Page 44 Envisat Validation Workshop, ACVT-GBMCD, GOMOS O 3 (r) 12/12/2002 Conclusions vs all GBMCD instruments :  bright limb ozone profiles:  only for bright (MV<1) stars and only above 30 km  GOMOS lower by 10 to 15% (30-50 km)  twilight limb ozone profiles:  needs further research  dark limb ozone profiles:  star magnitude: no clear influence  below 18 km: poor results  18-45 km: bias 5 to 10% (all stars)  45-65 km : cold stars: poor results  hot stars: bias within 20%, significant non-random bias suggests possibility for improvement