Inter-comparisons of the V8 SBUV instruments (Nimbus 7, NOAA 9, 11, and 16): Time, Altitude and Latitude dependency

Weekly zonal mean data sets created in 5-degree latitude intervals are used to make time series of the ratios of the retrieved layer ozone to the a priori layer ozone. The figure below shows the ratio time series at four selected altitudes (21 km, 30 km, 37 km, and 50 km) in the Equator latitudinal zone. The different instruments are distinguished by color (Red: Nimbus-7, Black: NOAA-9, Green: NOAA-11, and Blue: NOAA-16). The midpoint pressures of the V8 atmospheric layers are converted to a logarithm-scale standard altitude in kilometers by using a scale height of 7 km. The V8 a priori has an annual cycle and annual mean that vary with altitude (in approximately 5-km layers) and latitude (10-degree bands), but has no inter-annual or long-term variability. Therefore, these ratio plots show both V8 algorithm retrieval performance as well as possible instrument problems.
[Click on any of the figures in this document to see a full size version.]

Representative plot Equator Ratio (Retrieved/Apriori) Time Series

At the Equator, a Quasi-Biennial Oscillation (QBO) cycles can be seen in the lower layers, and volcanic effects in 1982 and 1991 can be seen in the 30 km layer. In general, most of instruments compare well during the overlapped time periods over all latitudes -- within ± 8 % except for early years of NOAA-9 (pre 1993 not shown on these plots, but included on the DVD) due to the unresolved instrument performance issues. The agreement among instruments has latitude and altitude-dependencies, particularly in the southern hemisphere.

The APPENDIX contains ratio time series plots at other latitudes. Contour plots showing the average mixing ratio differences in pairs of instruments during their overlap periods are shown in the Appendix.

Profile Quality Checking : the Percent Occurrence of Good Profiles

The figure below shows the percentage of profiles with profile error codes of 0 (good retrieval) or 1 (solar zenith angle greater than 84 degree) for each day. The significant decrease in Nimbus-7 occurs during the non-synchronization period. The decreases in NOAA-9 and NOAA-11 are due to grating drive problems and all of the NOAA-11 1995 data are at high solar zenith angles.

V8 SBUV Profile Total Ozone Comparison with Dobson Stations in the Northern Hemisphere

The profile total ozone is more reliable with fewer wavelength-dependent errors than conventional TOMS-like total ozone products. The matched-up V8 SBUV profile total ozone data are averaged on a weekly basis and compared with ground-based Dobson total ozone measurements from 40 stations in the northern hemisphere. The data smoothed with 15-point running averages, are shown as solid lines for each instrument. The V8 SBUV profile total ozone is up to 2 % (~ 6 D.U) higher than Dobson total ozone, but has no significant long-term trends.

Seasonal Variations of Layer Ozone over Lauder, New Zealand

The figure below shows seasonal variations for the NOAA-11 SBUV/2 data over Lauder, New Zealand (45oS, 170oE) and corresponding measurements from a ground-based microwave instrument at four altitudes. All available matched-up samples (n=872) for the NOAA-11 time period are used to make these plots. Thick red lines are microwave data smoothed with 7-point running averages. Blue lines are V8 SBUV data, also smoothed with 7-point running averages. A dynamic transport from ozone rich air in the tropics builds up the maximum ozone in late winter and early spring in the lower stratosphere at 24 km. A typical annual cycle (maximum in summer and minimum in winter) is shown in the middle stratosphere at 34 km. In the upper stratosphere above 43 km, a winter (June, July, August) ozone maximum is dominant due to the reduced ozone destruction rate. The V8 SBUV data agree well with the microwave data in terms of both offset and seasonal cycle variation. This result is consistent with SAGE II, LIDAR, and ozonesonde comparison results presented in Brinksma et al. [2002].

Time Series of Microwave to V8 SBUV Layer Ozone over Mauna Loa

The figure below shows the time series analysis of the ratio of microwave to V8 SBUV layer ozone over Mauna Loa (20oN, 156oW). The three solid lines are the linear regression fits for the time periods of NOAA-9 (Jul. 1995 to Jul. 1997), NOAA-11 (Jul. 1997 to March 2001), and NOAA-16 (Oct. 2000 to Nov. 2002), respectively, at four altitudes. Mauna Loa is a good site to do intercomparisons because it is not only the primary Network for the Detection of Stratospheric Change (NDSC) station for the tropics and subtropics, but also has low ozone variability that minimizes the uncertainty caused by differences in the same air volumes measured by different instrument [McPeters et al., 1999]. This plot does not show any significant trend for this short-term period (7 and 1/2 years), and also demonstrates a good agreement between SBUV instruments in the overlaps except in NOAA-9 time period coincidences at 37 km. Nevertheless, the linear regression fits have no significant slope larger than the standard deviation (numbers not shown here) as Tsou et al. [2000] described.

Mean Profile Differences (%) between V8 SBUV and Ground Measurements (Microwave, Lidar and Sonde)

The V8 SBUV profile ozone data are validated using ground-based data (microwave, LIDAR, and ozonesonde) with independent calibrations and derivation methods at various locations around globe. Microwave and LIDAR data are available from NDSC . Ozonesonde data are archived in World Ozone and Ultraviolet Radiation Data Center(WOUDC). The first two initials of the names of stations are used as legends on the plots as follows;

Microwave: LA(Lauder), MA(Mauna Loa), TA(Table Mt).
Lidar: OH(OHP), MA(Mauna Loa), TA(Table Mt).
Sonde: HO(Hohenpeissenberg), PA(Payerne), WA(Wallops Island), HI (Hilo).

  • Microwave comparisons with N9, N11, and N16
  • Lidar comparisons with N9, N11, and N16
  • Sonde comparisons with N7, N9, and N11
The mean percentage differences are calculated by averaging all matched-up samples for those time periods: (Nimbus-7: Dec. 1978 to Dec. 1990), (NOAA-9: Jan. 1993 to Feb. 1998), (NOAA-11: Dec. 1988 to Dec. 2001), (NOAA-16: Oct. 2000 to Sept. 2002). The ±1 standard error bars shown on the plots are calculated by dividing the standard deviation of percent difference by the square root of the number of coincidences at that altitude minus one. The numbers in parentheses indicate the number of coincidences for a station. For most of the upper and lower layers, these numbers are slightly reduced because of the sparse sampling of coincidences. The comparison results indicate that V8 SBUV ozone profiles generally agree with external data sources in the altitude range 24 to 50 km (30 hPa and 1 hPa) to within ± 10% (~5% on average). These results are consistent with Microwave, SAGE II, LIDAR, and ozonesonde comparison results from Brinksma et al., [2002], Leblanc et al., [2000], and Tsou et al., [2000].

References

Brinksma, E. J., J. Ajtic, J. B. Bergwerff, G.E. Bodeker, I. S. Boyd, J. F. de Haan, W. Hogervorst, J. W. Hovenier, and D.P. J. Swart, Five years of observations of ozone profiles over Lauder, New Zealand, J. Geophys. Res., 107(D14), 10.1029/2001 JD000737, 2002.

Leblanc, T., and I. S. McDermid, Stratospheric ozone climatology from lidar measurements at Table Mountain (34.4N, 117.7W) and Mauna Loa (19.5N, 155.6W), J. Geophys. Res., 105, 14,613-14,623, 2000.

McPeters, R. D., D. J. Hoffman, M. Clark, L. Flynn, L. Froidevaux, M. Gross, B. Johnson, G. Koenig, X. Liu, S. McDermid, T. McGee, F. Mucray, M. J. Newchurch, S. Oltmans, A. Parrish, R.Schnell, U. Singh, . J. Tsou, T. Walsh, and J. M. Zawodny, Results from the 1995 stratospheric ozone profile intercomparison at Mauna Loa, J. Geophys. Res., 104, 30,505-30,514, 1999.

Tsou, J. J., B. J. Conner, A. Parrish, R. B. Pierce, I. S. Boyd, G. E. Bodecker, W. P. Chu, J. M. Russell III, D. P. J. Swart, and T. J. McGee, NDSC millimeter wave ozone observations at Lauder, New Zealand, 1992-1998: Improved methodology, validation, and variation study, J. Geophys. Res., 105, 24,263-24,281, 2000.