Validation of ozone profiles measured by SMR and OSIRIS instruments
One of the chief goals of the Odin satellite is to observe the stratospheric ozone layer and its temporal and spatial evolution. Satellite measurements must always be validated to be certain that results obtained are valid. As a general rule, this is achieved by comparing satellite measurements with readings obtained at the same time from the ground. For Odin, as the satellite has two instruments observing the same quantities simultaneously and independently, we compared the results of SMR and OSIRIS.
SMR measurements are also analysed by a team at the Radio and Space Science Department of the Chalmers University of Technology in Gothenburg, Sweden, and a team at the Astrodynamics, Astrophysics and Aeronomy Laboratory at the Bordeaux Observatory. We therefore have three sets of different results for comparison. The figure below shows the mean result of the comparison of some 35,000 profiles of vertical ozone concentration, obtained at altitudes between 20 and 65 km in September 2002 and March 2004. We note that the SMR profiles for Bordeaux and Chalmers are in close agreement, varying 2 to 7% according to altitude, whereas the variations between the OSIRIS profiles and Chalmers are 7 to 18%.
These comparisons are complemented by other comparisons with ground-based measurements (using lidar or microwave techniques and ozone probes) or measurements from balloons or other satellites like POAM.
Validation of nitrogen dioxide profiles measured by OSIRIS
As with ozone, measurements of nitrogen dioxide (NO2) obtained by the OSIRIS instrument need to be validated by comparison, in this case only with ground-based measurements. The figure below shows a comparison of OSIRIS measurements (yellow) with those obtained by a SAOZ instrument on a stratospheric balloon launched by CNES on 16 March 2003. The red measurements are those acquired during the balloon’s ascent, with the instrument pointed at the Sun, while the black measurements are those acquired when the balloon had reached its ceiling altitude and the Sun was setting. We can see that the two SAOZ profiles do not match perfectly, as the balloon was drifting with the winds and the geographic distribution is not uniform. Conversely, the two profiles do match the measurements acquired by OSIRIS, including in the laminar flow layer observed near 22 km in altitude.
Validation of water vapour profiles measured by SMR
The SMR instrument observes stratospheric water vapour at two different frequencies: 489 GHz and 557 GHz. Taking advantage of the launch of a Swedish rocket from Kiruna carrying a hygrometer, the two microwave receivers acquired a battery of measurements. This figure shows the comparison of SMR profiles at 489 GHz (orange) and 557 GHz (blue) with the rocket-borne hygrometer profile, which offers better vertical resolution. The 489-GHz profile gives results very close to those of the hygrometer up to 45 km, then deviates from it, while the 557-GHz profile exhibits large oscillations that nevertheless remain close to those of the hygrometer up to about 80 km.
Validation of nitrogen protoxide observations by SMR
The first figure shows a comparison of two profiles of nitrogen protoxide (N2O) obtained by the Bordeaux group and the Chalmers group (Sweden) from the same SMR instrument data with a profile measured on a stratospheric balloon flight from Kiruna by a Fourier transform spectrometer from the LPMA (Laboratoire de Physique Atmosphériques et Applications). Note how the three profiles closely match.
The second figure also shows a comparison of different SMR profiles measured over Bauru, Brazil. The mean value is represented by the thick red curve plotted, with measurements acquired by a balloon-borne British DIRAC instrument (black dots) and a profile generated by the REPROBUS model. Despite the dispersion of the measurements, there is a close enough agreement to be able to use the N2O profiles obtained by Odin.
Comparison of geographic distributions of H216O and H218O water vapour isotopes
In December 2001, the SMR instrument observed two water vapour isotopes, H216O and H218O, simultaneously. The figure here shows the distributions observed as a function of latitude altitude(x-axis) and longitude latitude (y-axis). The clearly reveal the maximum level of water vapour, generally around 50 km in altitude. An interesting result appears at high latitudes in the northern hemisphere, where this maximum is rather at altitudes of 30 to 35 km for H216O and 30 to 40 km for H218O. This is a direct consequence of the winter polar vortex, where subsidence phenomena occur: the maximum water vapour concentration is therefore at lower altitudes.
Geographic distribution of carbon monoxide
Synoptic picture of levels of carbon monoxide (CO) at the South Pole (left) and North Pole (right) and from 100 to 0.005 hPa (approximately 20 to 100 km in altitude) on 18 November 2001. The high levels of CO in the upper atmosphere (thermosphere and mesosphere) are explained by the photodissociation of carbon gas. This CO descends in winter inside the vortex at the North Pole down to the lowest layers of the stratosphere, at around 1 hPa (approximately 30 km). Blue-mauve colours indicate low values, red-orange colours high values. [From Dupuy et al., Geophys. Res. Lett., 2004.].
Temporal evolution of the splitting of the polar vortex over the Antarctic
Evolution of stratospheric chemical constituents at an altitude of 20 km during the splitting of the polar vortex over the Antarctic in September-October 2002. The curves represent constituents measured by Odin—ozone (O3), chlorine monoxide (ClO), nitrogen protoxide (N2O), nitric acid (HNO3) and nitrogen dioxide (NO2)—at sunrise (SR) and sunset (SS), at different dates (from top down): 19-20 September, 25-26 September, 1-2 October and 4-5 October 2002. An increase in ozone-depleting ClO and a decrease in nitrogen constituents (HNO3 and NO2) in the two lobes of the vortex (materialized by low N2O values) are associated with a drop in ozone levels (ozone hole). The red line represents the edge of the vortex. Blue-mauve colours indicate low values, red-orange colours high values. [From Ricaud et al., J. Geophys. Res., 2005.].
Comparison of levels of ozone, chlorine monoxide and nitrogen protoxide measured by Odin with REPROBUS model results
The REPROBUS model has been developed by F. Lefèvre at the Service d'Aéronomie of the Institut Pierre Simon Laplace (IPSL). Note that the model faithfully reproduces satellite observations, except for chlorine monoxide, which could call for additional research in particular for low concentrations, although the values remain within the uncertainty limits of the observing instruments and the model. These results are important, as they confirm the dynamic evolution and molecular concentration hypotheses on which the model is based.
Comparison of Odin observations and REPROBUS model results
Evolution of stratospheric chemical constituents at an altitude of 20 km during the splitting of the polar vortex over the Antarctic in September 2002. The top curve represents constituents measured by Odin—ozone (O3), chlorine monoxide (ClO), nitrogen protoxide (N2O), nitric acid (HNO3) and nitrogen dioxide (NO2)— at sunrise (SR) and sunset (SS). The bottom curve represents the same constituents computed by the REPROBUS transport model. An increase in ozone-depleting ClO and a decrease in nitrogen constituents (HNO3 and NO2) in the two lobes of the vortex (materialized by low N2O values) are associated with a drop in ozone levels (ozone hole). The red line represents the edge of the vortex. Blue-mauve colours indicate low values, red-orange colours high values. [From de Ricaud et al., J. Geophys. Res., 2005.].
Comparison of geographic distributions of HDO isotope of water vapour observed by SMR and computed by a two-dimensional model
The figure left shows the relative variation in the distribution of the HDO isotope of water vapour observed by the SMR instrument from pole to pole at altitudes between 20 and 60 km on 11 September 2002, whgile the figure right shows the same values computed by a two-dimensional model developed by Dr. Martin Ridal at the Department of Meteorology of the University of Stockholm. This comparison shows that the model nevertheless needs to be further refined with respect to the quality of observations obtained by Odin.