SMOBA Documentation

SMOBA: A 3-Dimentional Daily Ozone Analysis Using SBUV/2 and TOVS Measurements

ABSTRACT

In this paper, we describe a 3-D daily global ozone analysis, Stratosphere Monitoring Ozone Blended Analysis (SMOBA), that is adopted for the input to CERES/EOS heating rate calculation. The analysis can also be used for other applications, such as the initial conditions for numerical ozone forecasts. The analysis utilizes the operational SBUV/2 measurements blended with the total ozone measurements from the 9.7 mm channel of HIRS/TOVS in the polar- night regions, where SBUV/2 data are not available. Currently, both measurements are from the polar-orbiting satellite, NOAA-14.

Monthly mean cross sections of ozone mixing ratio for the period of March through June, 1997 are presented. As compared to the SBUV-only climatology of Nagatani et al. (1988), SMOBA shows that ozone profiles in the polar regions are a smooth extension from the sun-lit latitudes. An analysis of temporal variations of mixing ratio is conducted for March and June. The results show that the largest temporal variation are located in the lower stratosphere around 175 hPa for most of the latitudes where the ozone concentration is the highest. In the upper stratosphere from 30 hPa up, polar regions exhibit large seasonality, while the tropics remain fairly calm. Spatial variations for Boreal equinox and solstice are also analyzed. Region-wise, the results are very similar to the temporal variations. The implication from the results of temporal and spatial variations is that without the dynamical effects, there could be errors of heating rates were the zonal averaged climatology used in the radiative transfer calculations in global numerical weather prediction models or general circulation models.

The vertical profiles of SMOBA are compared with Microwave Limb Sounder (MLS). The results show that there is very good agreement in shape and magnitude for the sun-lit northern hemisphere case. There is also reasonable agreement with some characteristic differences for the southern hemisphere case. From 50^o S to 70^o S, MLS profiles all peak at 5 hPa, while the SMOBA profiles peak from 8 hPa to 3 hPa. From 60^o S to 80^o S, SMOBA has higher magnitudes between 5 hPa and 50 hPa. The higher magnitudes is very possibly caused from TOVS overestimation of the total ozone.

1. Introduction :

Ozone has absorption features in both the solar (UV band) and infrared (9.6 mm and 14 mm ) portions of the spectrum, and affects tropospheric and stratospheric dynamics, radiation and chemistry. The vertical distribution of ozone concentration is important in structuring the earth-atmospheric temperature profile. In a sensitivity study, Ramanathan and Dickinson (1979) illustrate that the redistribution of ozone profile imposes larger perturbation to the tropospheric fluxes than a uniform changes of the profile. This finding implies that the recent ozone trend (Miller et al, 1995 and 1996), which occurs in the lower and upper stratosphere, can impose significant forcing on the climate changes. Evaluating the possible impact, Lacis et al. (1990) simulate a scenario of ozone change with a one-dimensional model. Their result found that a decrease of ozone above ~30 km, and increase in tropospheric and lower stratospheric ozone below ~30km, cause the global surface temperature to warm up. More recent experiment with a general circulation model (GCM) by Ramaswamy et al (1996) also reports that the ozone-induced cooling of the lower stratosphere leads to a negative radiative forcing of the surface-troposphere system. As ozone is highly variable both in time and space, these results indicate a need for a 3-dimensional (3-D) ozone analysis for studying the impacts of ozone on the heating/cooling profiles in the atmosphere, and thus the climate.

The Stratosphere Monitoring Group at Climate Prediction Center (CPC) / National Centers for Environmental Prediction (NCEP)/NWS/NOAA has recently developed an algorithm for 3-D daily ozone analysis. The algorithm uses the measurements from the Solar Backscatter Ultraviolet instrument (SBUV/2) (Bhartia, et al, 1996) which doesn't observe in the polar-night, blended with TOVS infrared ozone estimates (Neuendorffer, 1996) in the regions of polar night to achieve global coverage. It is thus called the Stratosphere Monitoring Ozone Blended Analysis (SMOBA). The current operational polar-orbiting satellite, NOAA-14, contains both instruments. For its operational stability and global coverage, SMOBA is adopted as the ancillary input data to the Cloud and Earth Radiant Energy System (CERES) of EOS interdisciplinary study (Wielicki et al, 1996) which is to estimate atmospheric radiative divergence/ heating rates by utilizing the meteorological, aerosol and ozone data from independent sources. CPC is also using SMOBA as the initial conditions for experimental numerical forecast of ozone.

The selection of operational measurement of SBUV/2 and TOVS ozone as the basis for developing the SMOBA algorithm is prompted by the long term stable global data stream required for supporting the CERES/EOS mission in studying climate variation and climate changes. Although there have been a number of experimental measurements that provide ozone profiles, such as and Cryogenic Limb Array Etalon Spectrometer (CLAES) (Bailey et al, 1996), the Microwave Limb Sounder (MLS) (Froidevaux et al, 1996), Improved Stratospheric and Mesospheric Sounder (ISAMS) (Connor et al, 1996) and Halogen Occultation Experiment (HALOE) (Bruhl et al, 1996) onboard NASA Upper Atmosphere Research Satellite (UARS), as well as Stratospheric Aerosol and Gas Experiment II (SAGE II) (McCormick et al, 1989) on board NASA Earth Radiation Budget Satellite, none of the above provides the global continuity of SBUV/2. TOVS total ozone is derived from the HIRS 9.7 mm channel, which can provide information for the regions which is not sun-lit. In fact, the effective coverage of SBUV/2 is limited by the solar zenith angle of less than.80^o, which is a function of the satellites' equatorial crossing time. Moreover, this blacked-out region tends to expands as the operational satellite precesses to later equatorial crossing times. Hence we have formulated an algorithm, blending SBUV/2 and TOVS to add additional information for the polar-nights. Continuous daily SMOBA covers the most comprehensive domain from surface to stratopause, from the pole to the pole.

This paper presents the procedure of SMOBA in Section 2. Data analysis, which includes the monthly means and temporal and spatial variations from the March to June 1997 is discussed in Section 3. Selected comparison with the measurements from MLS/UARS is in section 4.

2. Data Analysis Algorithm:

The process of SMOBA starts from the orbital information of the daily measurements of ozone mixing ratio and total column ozone that are obtained from the NOAA polar orbiting satellite, currently NOAA-14. The retrieved SBUV/2 ozone data include 19 levels of mixing ratio from 0.3 hPa to 100 hPa and total column ozone. Layer ozone amounts are also included for 12 discrete layers from the surface to ~60 km, but the vertical resolution is very coarse in the troposphere (only 2 layers).The input data are arranged along satellite orbit tracks, over the sun-lit portion of the earth. We have performed objective analysis from the orbital data to a global, 2.5 x 2.5 degree latitude-longitude grid by a commonly used successive correction method (SCM) (Daley, 1991). The output includes ozone mixing ratio at 24 pressure levels (0.2, 0.3, 0.4, 0.5, 0.7, 1, 1.5, 2, 3, 4, 5, 7, 10, 15, 20, 30, 40, 50, 70, 100, 150, 200, 250, 300 hPa) where we have added the 5 levels (300, 250, 200, 150, and 0.2 hPa) to the 19 level input mixing ratio, which are derived from ozone amount data in layers 2 (256 - 128 hPa), 3 (128 - 64 hPa) and 12 (0.25 - 0.12 hPa) respectively. We also assume a uniform ozone mixing ratio from 300 hPa to 1000 hPa, derived from the ozone amount in the lowest layer (1000 - 256 hPa). The total ozone is directly from the retrieved data, which is the sum of 12 layer ozone amounts. Note that due to the different vertical resolution of the raw data, the reliability of ozone mixing ratio is best above 30 hPa, good between 100 hPa and 30 hPa, and marginal below 100 hPa. We note, however, that the lower stratospheric ozone retrievals are guided by climatology used in the ozone retrieval algorithm, and corrected by the measured layer ozone amounts. The uniform tropospheric mixing ratio also captures the main feature shown in ozone sonde measurements.

Because the SBUV/2 algorithm is based on the sunlight absorbed and reflected by the atmosphere (iincluding ozone), there are no data in polar night. To fill in the polar night gap, we use NOAA TOVS total ozone derived from infrared measurements as the auxiliary information. However, TOVS does not measure vertical ozone profiles. From other satellite data such as UARS MLS, CLAES, ISAMS and HALOE, we have determined that the shapes of ozone vertical profiles are similar in polar latitudes to those at high mid-latitudes. Therefore, we assume polar night ozone mixing ratio to be such that its shape is the same as that of SBUV/2 ozone at the boundary of the polar night, and the integrated polar total column ozone is equal to TOVS measurement. As will be shown, this assumption is a reasonable approximation when the polar night gap is not too large. Since TOVS total ozone is slightly different from SBUV/2 total ozone (~5% at the boundary of polar night), a linear adjustment is applied to the polar ozone so that the filled data joint smoothly at the boundary . This procedure is the so-called blended data analysis, which combines the two data sources to obtain more complete information, on the basis of physical and instrumental understanding. In operation, in case TOVS ozone is not available for any reason, we simply use extrapolation to fill in the data gap. Therefore, there are no gaps in the data analysis files.

To demonstrate the effect of the blended analysis, we show in Figure 1 example of total ozone maps in the southern hemisphere. Figure 1a is SBUV/2 only for June 17, 1997. The polar night gap is near its winter solstice maximum, and the data gap is even larger than the physical polar night region due to the time of equatorial crossing, which is 1415 pm local time. In Figure 1b, TOVS data have been used to fill the polar night gap, providing us with detailed polar ozone characteristics such as wave patterns and a minimum at the east shore of Antarctica. The TOVS data has been slightly adjusted according to the SBUV/2 data at the polar night edge, so that there are no discontinuities or sharp gradients at the edge due to instrumental difference. As a comparison Figure 1c shows SBUV/2 data with extrapolations into polar night gap. Apparently, those real polar ozone features could not be reproduced due to relatively large expanse of the polar night region. Fortunately, we have both SBUV/2 and TOVS in NOAA operational satellites. Thus we do not have to extrapolate data in order to generate a globally covered, 3-D near real time daily ozone field.

3. Results from the first four months operation:

3.1 Monthly Means

SMOBA began daily operation on March 1, 1997. Figure 2 illustrates the monthly mean pressure-latitude cross sections for the first four months. The general features of ozone cross section have a mixing ratio maximum located near 8 hPa, about 33 km altitude. The maximum moves from 8^o S in March northward to about 5^o N in June. The maxima of the vertical profiles raise gradually from 8 hPa at equator to about 5 hPa at the poles. For May and June, there exists a second maximum larger than 8 ppmv located at about 8 hPa and 60 S, which also moves northward slightly from May to June. This result agrees very well with other published SBUV atlases, such as Nagatani et al. (1988), which contains 8-year average SBUV measurements for the sun-lit area and uses Cressman objective analysis (1959). SMOBA expands the domain to include the regions of polar night, and this analysis demonstrates that the cross section of ozone mixing ratio in the polar region is a smooth extension from the sun-lit regions.

3.2 Temporal Variations :

Stratospheric ozone constitutes an important greenhouse gas with substantial temporal and spatial variation that is not revealed in the seasonal variation shown in Fig. 2. Significant temporal variations of ozone usually occur in the day-to-day scale, which could be significant departures from the monthly climatologies or monthly means. Fig. 3 depicts the standard deviations of the ozone mixing ratio sampled from the daily analysis and normalized to the monthly means for March and June 1997. This result is in contrast to the interannual variance shown in Nagatani et al. (1988) which samples the monthly means from 8 years data. The normalization enhances the significance of the variability in the lower stratosphere, between 150 hPa and 75 hPa, where the ozone concentration is the highest.

The maximum of the normalized standard deviation generally occurs around 125 hPa for most of the latitudes, and ranges from 9% to more than 24% for March (Fig. 3a). At this altitude, the maximum of the normalized standard deviation occurs at 33 ⁰ N with the intensity of 27%. Not mirroring the northern hemisphere, southern hemisphere has a maximum of more than 15% at 45^o S. Although the features in the northern hemisphere is not symmetric to the southern hemisphere from 30 hPa upward, there is a gradual poleward increase of normalized standard deviation from 3% at 30^o N and S to 9% at poles, while the tropics is quite homogeneous.

Fig. 3b shows the corresponding normalized standard deviation for June. Albeit there exists similar synoptic features as March, the global pattern is severely skewed toward the southern hemisphere, especially from 30 hPa upward. The whole field from 30^o S northward to the Arctic is very homogeneous. In the lower stratosphere, the maximums are still located around 125 hPa for most of the latitudes, and ranges from 9% to 21%.

3.3 Spatial variations:

The boreal spring equinox and summer solstice are selected for illustrating the deviation from the zonal means around the globe. Fig. 4 is constructed by calculating the standard deviations about the zonal means normalized by the zonal means. Region-wise the general global pattern of this figure somewhat resembles the temporal variations of Fig. 3, with very homogeneous field in the tropics from 30 hPa upward, i.e. the ozone distribution in this region is very zonal. At the lower stratosphere, around 175 hPa, the normalized standard deviation has much larger magnitudes ranges from 10% to 20% for the spring equinox, Fig. 4a. A maximum of 20% locates at 175 hPa, 27^o N, is the largest in the lower stratosphere, while there are less clearly defined maximums in the southern hemisphere at 15^o S and around 55^o S with a magnitudes of 10%. Very distinctive from Fig. 4a is from 30 hPa upward at southern polar region, where a well defined maximum of 20% is located at 20 hPa 70^oS with substantial gradients from the center. While there appears to be a similar maximum in the northern hemisphere around 20 hPa 70^o N with the intensity of 10%, it is very mild.

On the boreal summer solstice, Fig. 4b, it appears that the ozone distribution in both hemispheres are far from zonal in the lower stratosphere from the tropics to the mid latitudes, particularly from the equator to 25^o N, where a maximum of 30% exists at 175 hPa. In the low latitudes, the mean zonal mixing ratio is small, about 0.1 to 0.2 ppmv as compared to 0.6 ppmv for the polar regions. Thus, any departure from the mean could be enhanced. From 30 hPa upward, most of the deviations are in the southern hemisphere from 30^o S poleward, while the northern hemisphere is fairly zonal. The implication from the results of temporal and spatial variations is that there could be errors of heating rates were the zonal averaged climatology used in the radiative transfer calculations in global NWP models or general circulation models.

4. Profile Comparison with MLS

Limited ozone profiles from Microwave Limb Sounder (MLS) (Froidevaux, 1996) are available for comparison from UARS, which is in a 58^o inclined orbit. For this particular period under study, the UARS is operating under a power deficiency, so that all instruments can not be on all the time. SAGEII, also, is limited in coverage of the polar regions. Therefore, we have focused attention on the MLS (Version 4) , which provides 2 days of comparision from April 1 (South facing) to April 10 (North facinging). The facing of UARS enable MLS to measure further poleward then the terminator of orbital inclination. the MLS provides profiles from 100 hPa up with coarser vertical resolution than SMOBA.

On April 10, 1997, the SMOBA profiles at three locations, 90^oE and 50^o N, 60^o N, and 80^o N, respectively, are on Fig. 5a. The corresponding MLS profiles are on Fig. 5b. This is a case that SBUV/2 has coverage up to 80^o N, thus most of the profiles are from measurements instead of blending with TOVS ozone. It is clear that the higher resolution of SMOBA makes the profiles smoother. Nevertheless, there is very good agreement as both MLS and SMOBA have profiles peaks around 5 hPa for all the three locations. The magnitudes of the peaks at 50^oN (60^oN) for SMOBA agree with that of MLS, within 0.2 (0.1) ppmv out of 8.4 (7.7) ppmv, while the one at 80^oN differ slightly larger, 0.5 ppmv out of 6.2 ppmv. Although there is precise agreement for the peak at 60^oN, profile-wise, SMOBA overestimates MLS from 20 hPa to 5 hPa with largest difference of 0.7 ppmv at 10 hPa. The profiles at the other two locations are in good agreement within 0.4 ppmv for all the altitudes. Although not shown, we have also compared similar profiles for the same latitudes at 0^o E, 180^o, and 90^o W. All of them have the similar characteristics and good agreement.

For the case in the southern hemisphere on April 1, 1997 at 90^o W, SBUV/2 has coverage down to around 68⁰ S. The SMOBA profiles higher than 68⁰ S, Fig. 5c, are from blending TOVS ozone, namely the profiles at the boundary of the polar nights are carried into the higher latitudes.

For this reason, the shape of the profiles at 70^o S and 80^o S are very similar. Comparison with MLS, Fig. 5d, shows that there is quite reasonable agreement between the two data sets especially below 50 hPa and above 2 hPa, and with some characteristic differences in between. From 50^o S to 70^o S, MLS profiles all peak at 5 hPa, while the SMOBA profiles peak from 8 hPa to 3 hPa. At 70^o S and 80^oS from 5 hPa to 50 hPa, SMOBA mixing ratio decreases substaintially slower than that of MLS. Conversely is the case at 50^o S from 15 hPa to 50 hPa. From 60^o S to 80^o S, SMOBA has higher magnitudes between 5 hPa and 50 hPa. Integrating the column ozone from 100 hPa to 1 hPa found that SMOBA is about 10 % higher than MLS from 60^oS poleward. Although there has been few studies on the polar night ozone profiles, the most reasonable conjecture for the cause of this difference is possibly from the overestimates of the total ozone by TOVS. Table 2 of Neuendorffer (1996) documents "Auxiliary TOVS Total Ozone Adjustments" derived from comparison with total ozone-mapping spectrometer (TOMS) between 75^o S and 75^o N. For mean April difference averaged from 1979 through 1992, it states that TOVS is 36 Dobson Unit higher than TOMS. To comply with TOMS, it is possible in the future to develop a similar daily table for adjustment once SMOBA has collected a full year data. The adjustment can narrow the gap between SMOBA and MLS. However, it can not change the shape of SMOBA profiles in polar-nights, which calls for further research.

Summary:

For computing global heating rate, CERES/EOS has adopted a 3-D daily global ozone analysis, SMOBA, as an ancillary input. The analysis developed at CPC/NCEP employes a successive correction method and uses operational SBUV/2 measurements in the sun-lit regions, blended with the total ozone measurements from the 9.7 mm channel of HIRS/TOVS in the polar- night regions. Currently, both measurements are from the polar-orbiting satellite, NOAA-14. This near real-time analysis can also be used for other applications, such as the initial conditions for numerical ozone forecasts that CPC/NCEP is experimenting.

Since began daily operation on March 1, 1997, SMOBA has proven itself to be a very stable and reliable algorithm required by the coming CERES operations. Monthly mean cross sections of ozone mixing ratio for the period of March through June, 1997 are presented. As compared to the SBUV-only climatology of Nagatani et al. (1988), which uses Cressman objective analysis, SMOBA shows that ozone profiles in the polar regions are a smooth extension from the sun-lit latitudes. The maximum of the ozone mixing ratio located near 8 hPa and moves from 8^o S in March northward to about 5^o N in June. An analysis of temporal variations of mixing ratio is conducted for March and June. The results show that the largest temporal variation are located in the lower stratosphere around 175 hPa for most of the latitudes where the ozone concentration is the highest. In the upper stratosphere from 30 hPa up, polar regions exhibit large seasonality, while the tropics remain fairly calm. Spatial variations for Boreal equinox and solstice are also analyzed. Region-wise, the results are very similar to the temporal variations. The implication from the results of temporal and spatial variations is that without the dynamical effects, there could be errors of heating rates were the zonal averaged climatology used in the radiative transfer calculations in global numerical weather prediction models or general circulation models.

Two cases of the SMOBA vertical profiles are compared with the measurements from MLS (Version 4) onboard UARS, one case for each satellite facing, respectively. The results show that there is very good agreement in shapes and magnitudes for the sun-lit northern hemisphere case (South facing), April 10, 1997. There is also reasonable agreement with some characteristic differences for the southern hemisphere case (North facing), April 1, 1997. From 50^o S to 70^o S, MLS profiles all peak at 5 hPa, while the SMOBA profiles peak from 8 hPa to 3 hPa. From 60^o S to 80^o S, SMOBA has higher magnitudes between 5 hPa and 50 hPa. The higher magnitudes is very possible caused from TOVS overestimation. Adjustment at high latitudes can be developed once SMOBA collects one full year data. Since there are limited ozone profiles available for polar night, the result calls for more research before adjustment to the shape of profile can be made.

For further studying the profiles of polar nights, it is our intent to analyze July and December '97 data in the future when they are available. A comparison with MLS or other ozone sonde measurements can be conducted.

SMOBA started the routine operation on March 1, 1997. A ten-day rotating file has been set up for anonymous FTP at: ftp.cpc.ncep.noaa.gov/SMOBA/.

Acknowledgment:

This work is supported by a NASA Langley Research Center Grant L90988C. Discussions with Drs. Ron Nagatani, Hai-Tien Lee, Craig Long and Mel Gelman have been very helpful.

References:

Bailey, P. L., D. P. Edwards, J. C. Gille, L. V. Lyjak, S. T. Massie, A. E. Roche, J. B. Kumar, J. L. Mergenthaler, B. J. Connor, M. R. Gunson, J. J. Margitan, I. S. McDermid, and T. J. McGee, Comparison of cryogenic limb array etalon spectrometer (CLAES) ozone observations with correlative measurements, J. Geophy. Res., 9737-9756, 1996.

Bhartia, P. K., R. D. McPeters, C. L. Mateer, L. E. Flynn, and C. Wellemeyer, Algorithm for the estimation of vertical ozone profiles from the backscattered ultraviolet technique, J. Geophy. Res., 101, No. D13, 18793-18806, 1996.

Bruhl, C., S. R. Drayson, J. M. Russell III, P. J. Crutzen, J. M. McInerney, P. N. Purcell, H. Clande,, H. Gernandt, T. J. McGee, I. S. McDermid, and M. R. Gunson, Halogen Occulatation Experiment ozone channel validation, J. Geophys. Res., 10217-10240, 1996.

Connor, B. J., C. J. Scheuer, D. A. Chu, J. J. Remedios, R. G. Grainger, C. D. Rogers, and F. W. Taylor, Ozone in the middle atmosphere as measured by the improved stratospheric and mesospheric sounder, J. Geophys. Res., 9831-9842, 1996.

Cressman, G. P., An Operational Objective Analysis System, Mon. Wea. Rev., 87 No. 10, 367-374, 1959.

Daley, R., Atmospheric Data Analysis, Cambridge University Press, 457pp, 1991

Froidevaux, L., W. G. Read, T. A. Lungu, R. E. Cofield, E. F. Fishbein, D. A. Flower, R. F. Jarnot, B. P. Ridenoure, Z. Shippony, J. W. Waters, J. J. Margitan, I. S. McDermid, R. A. Stachnik, G. E. Peckham, G. Braathen, T. Deshler, J. Fishmanm, D. J. Hofmann, and S. J. Oltmans, Validation of UARS Microwave Limb Sounder ozone measurements, J. Geophy. Res., 10017-10060, 1996

Lacis, A. A., D.J. Wuebbles, J. A. Logan, Radiative forcing of climate by changes in the vertical distribution of ozone, J. Geophy. Res., 95, No. D7, 9971-9981, 1990.

McCormick, M. P., J. M. Zawodny, R. E. Veiga, J. C. Larsen, and P. H. Wang, An overview of SAGE I and II ozone measurements, Planet. Space Sci., 37 No.12, 1567-1586. 1986

Miller, A. J., G. C. Tiao, G.C. Reinsel, D. Wuebbles, L. Bishop, J. Kerr, R. M. Nagatani, J. J. DeLuisi, and C. L. Mateer, Comparisons of observed ozone trends in the stratosphere through examination of Umkehr and ballon ozonesonde data, J. Geophy. Res., 100, N. D6, 11209-11217, 1995.

Miller, A. J., S. M. Hollandsworth, L.E. Flynn, G.C. Tiao, G.C. Reinsel, L. Bishop, R. D. McPeters, W. G. Planet, J. J. Deluisi, C. L. Mateer, D. Wuebbles, J. Kerr, and R. M. Nagatani, Comparisons of observed ozone trends and solar effects in the stratosphere through examinatio of ground-based Umkehr and combined solar backscattered ultraviolet (SBUV) and SBUV2 satellite data, J. Geophy. Res., 101, No C4, 9017-9021, 1996.

Nagatani, R.M., A. J. Miller, K. W. Johnson, and M.E. Gelman, An eight-year climatology of meteorological and SBUV ozone data, NOAA Tech. Report, NWS40. U.S. Dept of Commerce, 125pp, 1988.

Neuendorffer, A.C., Ozone monitoring with TIROS-N operation vertical sounders, J. Geophy. Res., 101, No. D13, 18807-18828, 1996.

Ramanathan, V. and R. E. Dickinson, The Role of Stratospheric Ozone in the Zonal and Seasonal Radiative Energy Balance of the Earth-Troposhere System, J. Atmos. Sci., 36, 1084-1104, 1979.

Ramaswamy,V., M. D. Schwarzkoph and W. J. Randel, Fingerpirnt of ozone depletion in the spatial and temporal pattern of recent lower-stratospheric cooling. Nature, 382, 616-618, 1996.

Wielicki, B. A., B. R. Barkstrom, E. F. Harrison, R. B. Lee, G. L. Smith and J. E. Copper, Clouds and the Earth's Radiant Energy System (CERES): An Earth Observing System Experiment, Bull. Amer. Meteo. Soc., 77, 853-868, 1996.

Figure Caption:

Fig. 1: Total column ozone for June 17, 1997, in Dobson Unit. (a) Analysis of SBUV/2 without fill-in for polar night; (b) SMOBA; (c) Fill in polar night of (a) by simple extrapolation.

Fig. 2: Monthly mean cross sections of ozone mixing ratio for March , upper left panel; April, upper right panel; May, lower left panel, and June 1997, lower right panel, in ppmv. Contour interval is 1 ppmv.

Fig. 3a: SMOBA temporal variation for March, 1997. The thick lines are the standard deviations from daily ozone mixing ratio divided by the monthly mean, in %. contour interval is 3%. The thin lines are the contours of monthly mean ozone mixing ratio as in Fig. 2, in ppmv. Contour interval is 1 ppmb. The ordinate is height in hPa.

Fig. 3b: same as Fig. 3a, but for June, 1997.

Fig. 4a: SMOBA spatial variation for March 21, 1997. The thick lines are the standard deviations of ozone mixing ratio divided by the zonal mean, in %. contour interval is 3%. The thin lines are the contours of zonal mean ozone mixing ratio, in ppmv. Contour interval is 1 ppmb. The ordinate is height in hPa.

Fig. 4b: Same as Fig. 4a, but for June 21, 1997.

Fig. 5a: SMOBA Ozone profiles at (50^o N, 90^oE), (60^o N, 90^oE) and (80^o N, 90^oE) for April 10, 1997, in ppmv

Fig. 5b: Same as Fig. 5a, but from MLS.

Fig. 5c: SMOBA Ozone profiles at (50^o S, 90^oW), (60^o S, 90^oW) and (80^o S, 90^oW) for May 19, 1997, in ppmv

Fig. 5d: Same as Fig. 5c, but from MLS.