Data & Products
Missions & Sensors
DFD Data Products
GOME Level 3+ Data Products
All level 1 and level 2 data products are produced on behalf of ESA and can be obtained from that source. Level 3+ data products are generated by DLR under national contract and are freely available via DFD's ISIS-WWW interface as well as over the internet.
Level 2 data form a cloud of heterogenously distributed data points in time and space (see Figure 4). This kind of information is hard to handle and therefore usable mainly to the experts, only. The philosophy behind higher data products is to make theses expensive remotely sensed data, - or to be more precise -, to make the information behind the data visible and available not only to the relatively small group of experts but also to scientists in other disciplines, to thousands of individuals, companies and government groups who can use it. Therefore, definition and development of GOME level 3+ data products strongly depends on the "customer's" requirements.
The present constellation of GOME level 3 data products can be grouped into five categories (status as of March1998):
I. Daily global and regional (Europe) maps of total column ozone
The most basic level 3 data products are global and regional maps of the distribution of total column ozone in different geographical projections. As mentioned above, GOME can be operated at various horizontal resolutions. Even if GOME is to use the relatively coarse resolution of approximately 40 x 320 km, daily complete coverage of the Earth's surface is achieved only for latitudes above about 65 degrees north and south. It takes three days to entirely cover the globe. It is quite obvious that the information content from such kind of three day averaged maps is rather limited. This is because the distribution of total column ozone is primarily controlled through dynamic processes in the atmosphere (About 90-95% of ozone total column density stems from altitude regions up to about 35 km. In this region, ozone can be regarded as being photochemically stable for time scales of a few days and weeks. Therefore, the distribution of total column ozone is mainly due to transport processes, that is dynamics.). However, it is well-known that there are considerable dynamic processes in the atmosphere with time scales well below three days. Examples are gravity waves (typical time scales are within several hours), shorter-period planetary waves (e.g. the two-day wave), Kelvin and mixed Rossby-gravity waves near the equator, large-scale circulation, and surges above thunderstorms, cyclones or over sea surface temperature anomalies such as El Niņo, to mention only a few.
Daily mean total column ozone
To overcome these problems and to allow to monitor structures on time scales of at least one day, an atmospheric planetary wave model in conjunction with sophisticated spectral analysis methods is used to generate daily maps of the total column ozone distribution using the GOME level 2 data of only one day at maximum. The daily mean maps are available operationally in different projections globally (Figure 6) and specifically for Europe (Figure 7).
An improved Harmonic Analysis technique is used to estimate the planetary wave model parameters. The accuracy of the daily composites typically is within only 3%. Refer also to our validation page.
Near real time service
Many applications need information on the ozone distribution in near real time. For example, TV-media, environmental and public health authorities need this up-to-date information to early warn people against health risks since low total column ozone implies increased ground-level UV-B radiation (UV-B is biologically harmful and can create cancers, cataracts and impacts the human immune system). To serve this request, DLR-DFD operates a near real time service: based on GOME data as they are downlinked at ESA's ground receiving station at Kiruna, Sweden, (up to 10out of 14 orbits per day are received at Kiruna) data products are available within only a few hours after acquisition. A one day forecast service is being worked at.
Such maps have proven to help to early detect and to monitor peculiar situations such as the considerable ozone decrease in the Northern Hemisphere in spring 1997. Such maps also reflect the extremely high variability of the ozone content in time and space. For instance, Figure 7 (left) impressively demonstrates that, at least occasionally, there are strong gradients in total ozone on horizontal scales of only a few hundred kilometers. Such features are believed to be related, at least in part, to weather phenomena in the troposphere (see below and Figures 8, 9). Therefore, this kind of level 3 data product is of considerable interest also to weather services since this information can help to improve the quality of longer range weather forecasting.
Moreover, these maps help to observe the development of so-called 'streamer-events'. Such a typical finger-like feature is clearly visible in Figure 7 (center) reaching out from lower to higher latitudes off-coast of Europe and North Africa. Streamer induce high temporal variability of the ozone content over a fixed site and can catch people unaware when a sudden ozone decrease can allow a brief burts of intense radiation to reach the ground. This is dangerous especially during summer time when the sun is high. These aspects are subject to the EU-project 'STREAMER' which is to set up an early warning system against streamers and ozone mini holes. For example, Figure 8 shows a considerable area of ozone decreased air approaching
Europe on February 8, 1998. To judge whether this could imply environmental risks or not one has to take into account the current cloud cover since clouds effectively block solar UV-B radiation from penetrating to the ground. Therefore, DLR-DFD will soon provide simultanous information from METEOSAT (Figure 8, right) on an operational scheme.
Instantanous distribution of total column ozone - 'snapshot'
For some applications, the temporal uncertainty of half a day that is associated with the Harmonic Analysis approach is not acceptable. For example, scientists performing measurements on rockets, balloons or aircrafts would like to compare their measurements with GOME-data near in time and space. Therefore, to meet the request to provide instantanous 'snapshots' of the total coulmn ozone distribution for a specific time, DLR-DFD offers the special service to generate global or regional maps of total column ozone for a requested time. This is done using the Kalman-Filter technique.
An example, demonstrating the capability of the Kalman-Filter technique to monitor dynamically highly variable structures, Figure 9 shows a sequence of 'snapshots' of the total column ozone distribution over Europe from February 7, 1998, 12:00 UT until February 9, 1998, 00:00 UT. The interval between the different 'snapshots' is 6 hours.
Note that the temporal variability of total ozone over a specific site on Earth can easily vary up to about 20% within only a few hours. As a rule of thumb, ground-level UV-B intensity variability is a factor of 1.5 higher (assuming clear skies and local noon). In other words: a 20% decrease in total column ozone means an increase in ground-level UV-B intensity of up to about 30%. However, one item that is being worked at is to operationally provide simultanously maps of UV-B gorund-level intensity.
To derive vertically resolved ozone distributions the 3-dimensional ROSE chemical-transport model is used. Total ozone columns are assimilated sequentially into the ROSE model and modify the models ozone distribution characteristically. Daily updated ozone profiles, a ROSE model description and validation results can be found here.
Analyses of data from satellites and ground stations clearly reveal noticeably reduced ozone column densities everywhere except at tropical latitudes. The trend is, however, not uniform; it varies according to region and season. Ozone losses are most noticeable in the Northern Hemisphere during winter and springtime (up to eight percent per decade). At intermediate latitudes in the Southern Hemisphere the trends are almost independent of season and amount to about five percent per decade. Ozone decrease is highest during the months of September to November at high southern latitudes (over ten percent per decade). Because of the long atmospheric life of CFC's, which are regarded as the prime cause of ozone reduction, a continuation of the ozone depletion trend can be expected in both hemispheres. In addition to its function as a shield from biologically harmful amounts of the sun's UV irradiance, ozone also has an effect on the climate (absorption of high energy solar UV determines to a large extent the atmosphere's vertical temperature structure). For such reasons, careful long-term measurements of atmospheric ozone are needed to monitor ozone trends. DLR-DFD contributes to this task by providing regularly updated information.
Contourplots of zonal mean total column ozone
Monthly updated contourplots of GOME's zonal mean total column ozone are provided (Figure 10, left). To better judge on the measurement it is compared to the CIRA-climatology (Figure 10, right). These data products are especially used to help to validate atmospheric global circulation models.
Zonal mean data do not allow to study longitudinal structures, of course. However, one focus in current scientific activities worldwide is to investigate regional structures in ozone trends. For example, analyses of different ground-based measurements have shown that especially in Europe and during January there are negative trends in total ozone that can be twice as high as the zonal mean trend for these latitudes. It is suspected that planetary waves play a major role within this respect. DLR-DFD contributes to these research actvities by providing maps and data of monthly mean total ozone. Figure 11, for example, gives the monthly mean situation for March, 1997. Please note that in this case there was a strong stationary planetary wave number one, superimposed by a somewhat weaker wave number two present in the Northern Hemisphere.
Maps, showing regional trends in total ozone for each month is being worked at.
In recent years considerable progress has been made in understanding the complexities of ozone generation and breakdown in the atmosphere. Nevertheless, todays computer models still underestimate ozone destruction at polar latitudes by factors ranging from 1.5 to 4. Ozone trends at middle latitudes are hardly understood at all. As mentioned above, the primary reason for this is that dynamic processes have so far been inadequately taken into account (limited computer power is one of the show stoppers, of course). Dynamic atmospheric processes take place over a wide temporal and spatial ranges. Currents, eddies, waves and turbulence are part of the picture and are likewise responsible for the transport and mixing of trace gases. It was already mentioned, that especially planetary waves play a major role within this respect. GOME measurements of total column ozone can make significant contributions in monitoring these large scale waves.
Planetary wave climatology
Planetary scale waves are frequently observed in the middle atmosphere. In their simplest form they occur because of the variation of the Coriolis parameter with latitude(see Figure 12). They show up as traveling waves with periods of several days to a few weeks while excitation and decay occurs on time scales of about 1-2 months. Some of these large scale waves are forced by features at the surface, by orography and thermal contrast, e.g., between land and sea, in which case they would be stationary with respect to the surface to produce, for example, the Aleutian antcyclone. There are several reasons that constitute the substantial interest in these phenomena. For example, planetary waves drive the circulation out of radiative equilibrium, they are known to be important mechanisms for transport processes in the stratosphere, are responsible for the intermittent breakdowns of the polar vortex structure called sudden stratospheric warmings and are involved in vortex erosion processes. Especially, the latter point is of substantial interest since the dramatic springtime depletions of ozone in polar regions rquire that polar stratospheric air has a high degree of dynamical isolation and extremely cold temperatures necessary for the formation of polar stratospheric clouds. Both of these conditions are produced within the undisturbed stratospheric winter polar vortex.
There are many indicators of wave activity such as temperature or winds. Each of these indicators has been studied in the past for insight into a variety of wave mechanisms. Besides the atmospheric parameters mentioned above the distribution of quasi-conservative constituents such as ozone provide valuable information on wave structures since their chemical lifetime is so long that their distribution is specified by transport processes.
By far the largest part of atmospheric variability is explained by the stationary planetary waves number one and two. As an outcome of the generation of daily global ozone maps (see chapter I.), amplitudes for the waves one and two are gathered day by day. Contourplots of these daily amplitudes are provided by DLR-DFD and updated on a monthly time line (Figure 13).