TROPOSAT

The Use and Usability of Satellite Data for tropospheric research

Ulrich Platt1, John P. Burrows2, Peter Borrell3, Martin Dameris4, Albert P.H. Goede5, Hennie Kelder6,Paul S. Monks7, Andreas Richter2, Herman G.J. Smit8, Thomas Wagner1

1Institut für Umweltphysik, D-69120 Heidelberg, Germanye-mail: ulrich.platt@iup.uni-heidelberg.de

e-mail: thomas.wagner@iup.uni-heidelberg.de2Institut für Fernerkundung, D-28395 Bremen, Germany

e-mail: burrows@iup.physik.uni-bremen.dee-mail: andreas.richter@iup.physik.uni-bremen.de

3P&PMB Consultants, D-82467 Garmisch-Partenkirchen, Germanye-mail: p.borrell@gaponline.de

4Institut für Physik der Atmosphäre, DLR, Oberpfaffenhofen, D-82230 Wessling, Germanye-mail: martin.dameris@dlr.de

5Space Research Organization (SRON), Sorbonnelaan 2, 3583 CA Utrecht, Netherlands, e-mail. A.Goede@sron.nl

6KNMI, P.O. Box 201, 3730 AE De Bilt, Netherlandse-mail: kelder@knmi.nl

7University of Leicester, Department of Chemistry, University Road, Leicester LE1 7RH, UKe-mail: psm7@le.ac.uk

8Research Centre Jülich, P.O. Box 1913, D- 52428 Jülich, Germanye-mail: h.smit@fz-juelich.de

INTRODUCTION

The newly observed global distributions of tropospheric species such as NO2, SO2, HCHO and BrO, obtained byobservations from space platforms, mark the opening of a new era in tropospheric chemistry. Both the treatment of thedata presently available, and of data from the missions planned for the near future, will add a new dimension to thefield. The new data offer the possibility of studying concentration distributions over large areas, and in locations and atlevels in the troposphere, which were only accessible to occasional individual observations in the past. They will lead toa significant enhancement of our capability for investigating the chemistry and physics of the troposphere. An exampleof a global distribution for NO2 is given in Fig. 1 [1–9].

Furthermore the reliability and frequency of space observations will be of great use in determining the longer termchanges which are taking place in the troposphere. In addition the observations, appropriately combined withmodelling, will assist those responsible for environmental policy development to monitor the outcomes of legislativeinitiatives to control the quantities of pollutants in the troposphere.

The purpose of the present paper is to outline a new project, TROPOSAT, which has been formed to exploit the firstresults and, by demonstrating that a selected set of scientific objectives can be achieved with satellite data, bring to theattention of a wide scientific community the exciting possibilities offered by using space observations. In this sense,TROPOSAT is an essential capacity building exercise. In addition the project will acquaint the agencies responsible forpolicy development for the European troposphere with the application of satellite data for monitoring pollutants andtracing emissions on regional and global scales.

Fig. 1: The average global tropospheric NO2 density in 1997, determined from GOME. The high concentrations associated withindustrialised regions can readily be seen [10].

TROPOSAT: AIMS AND OBJECTIVES

The formal aim is to determine two- and three- dimensional distributions and time series of trace gases and otherparameters in the troposphere and so facilitate future research and environmental monitoring on regional and globalscales [11].

The purpose of the project is to explore and encourage the use of satellite data for tropospheric research. TROPOSAT isa new subproject of EUROTRAC-2, the EUREKA project studying the transport and transformation of pollutants in thetroposphere [12]. Studies will be carried out of pollutants over continental and regional scales, and the possibility ofusing remote sensing for Kyoto Monitoring will be investigated.

Four activities are envisaged: the development of algorithms for the retrieval of tropospheric species and parameters; the use of satellite data for understanding atmospheric processes; the synergistic use of different instrumentation and platforms for tropospheric measurements; the development of validation strategies for tropospheric satellite data products.

In addition TROPOSAT will undertake other underpinning and derived activities such as the development ofappropriate data assimilation techniques combining satellite measurements with modelling, and the specification of therequirements for future satellite instruments for tropospheric work.

CO-ORDINATION AND FINANCE

As already mentioned, TROPOSAT is part of EUROTRAC-2 which comprises twelve subprojects studying all aspectsof the chemistry of the troposphere [12]. TROPOSAT, which presently consists of some 40 principal investigators, isorganised into four task groups each addressing a principal aim of the project. They are described in the subsequentsections.

The project is guided by a steering committee consisting of the leaders of the task groups together with the subprojectcoordinator, Prof. Ulrich Platt from Heidelberg, and two deputy coordinators, Prof. John Burrows from Bremen and Dr.Peter Borrell from Garmisch-Partenkirchen. There also a number of affiliated investigators, mainly from NASA.

As with all EUROTRAC-2 subprojects, TROPOSAT principal investigators obtain their own funds for the workthrough normal research grants from the investigator's home country, or from ESA or the EC under their normalresearch programmes. However ESA kindly makes a grant to TROPOSAT to support part of the co-ordination of theproject.

TROPOSAT TASK GROUPS

1. Development of Algorithms for the Retrieval of Tropospheric Species and Parameters.

Leaders: Andreas Richter, IEP, Bremen, Thomas Wagner, IUP, Heidelberg

Albert P.H. Goede, SRON, NL Gabriele P. Stiller FZK, Karlsruhe, DHennie Kelder , KNMI, NL Arnolds Ubelis, Riga, LatviaJohannes Keller, PSI, CH Jean Verdebout, JRC, IspraMartin Riese, Wuppertal, D Mark Weber, IEP, Bremen, DM. van Roozendael, BIRA, Brussels, B R. de Winter-Sorkina,IMAU, Utrecht, NLThe immediate requirement in the field is the improvement of existing algorithms for extracting concentration profilesfrom measured reflectances and the development of new ones for the quantitative determination of tropospheric speciesfrom space. The work will focus on development and refinement of algorithms for:

the spectral de-convolution and the better understanding of instrumental effects; radiation transport modelling, which is important in converting the recorded radiances or ‘apparent (or slant) trace

gas column densities’ into vertical column densities or trace gas concentration profiles; cloud or haze detection and correction; deriving further quantities from the primarily measured and evaluated data, such as the calculation of UV levels

and photolysis frequencies [e.g. Meerkoetter et al. 1997] or trace gas source strengths.

2. Use of Satellite Data for Understanding Atmospheric Processes.

Leader: Martin Dameris, DLR Oberpfaffenhofen

Jann Forrer, EMPA, Duebendorf, CH J.-F. Müller, BIRA, Brussels, BRainer Friedrich, IER, Stuttgart, D Ulrich Platt, IUP, Heidelberg, DGarry Hayman, AEA Technology, UK Franz Rohrer, Forschungszentrum Jülich, DD. P. Jeker, EMPA, Duebendorf, CH Hans Schlager, DLR, Oberpfaffenhofen, DJean-C. Lambert, BIRA-IASB, Brussels, B Andreas Stohl, TU, München, DMark G. Lawrence, MPI, Mainz Wilfried Winiwarter, Seibersdorf, AData on tropospheric concentrations is now becoming available and the object of this task group is to demonstrate howsatellite data can best be combined with model results and data from terrestrial measurements to investigatetropospheric problems. The activities will include:

case studies with chemical transport models including inversion of transport processes; validation of chemistry-climate models; comparison and interpretation of model results and observations with satellite data.

The studies within the task group will address the following scientific problems: identification of sources and sinks of pollutants; export of pollutants out of Europe into the free troposphere and to the East; assessment of the role of tropospheric-stratospheric exchange in the ozone budget; study of the NOx emissions from aircraft in the upper troposphere and from lightning; accurate characterisation of the atmosphere and the change in its composition with time using data assimilation

analysis.

Fig. 2 shows the potential of satellite data for addressing problems in our understanding of the troposphere. The totaldisappearance of ozone on the fringes of the arctic was a mystery when it was first discovered ten years ago. It is nowthought to be due to interactions with bromine compounds produced in the arctic spring. The satellite data in Fig. 1 canbe expected to throw much light on this problem, both in the Arctic and the Antarctic.

Fig. 2: Vertical column density of BrO measured by GOME in the Arctic (left) and Antarctic (right) during events of stronglyenhanced tropospheric BrO concentrations. These events are indicated by the red and white areas while green and yellow coloursrepresent typical 'background' values (BrO mainly located in the stratosphere and possibly also in the free troposphere). The circlesindicate the location of Ny-Alesund (Spitzbergen) and the Neumayer station (Antarctica). At the time of the GOME BrO maps, verylow ozone concentrations were reported from both stations [14].

Fig. 3: The lower part shows observations of the total ozone column with GOME from 1 July 1995 to 30 June 1999 (prepared byDeutsches ernerkundungsdatenzentrum, DLR-Oberpfaffenhofen). The upper part shows four individual years of a multi-yearsimulation using the three-dimensional interactively coupled chemistry-climate model ECHAM4.L39(DLR)/CHEM [15]. The free-running climate model including chemistry is not driven by the observed dynamics, and therefore the calculated dynamical andchemical parameters do not exactly reflect observations.

There is great potential in the comparison of satellite data with model results to the benefit of both, as Fig. 3 shows.

3. Synergistic Use of Different Instrumentation and Platforms for Tropospheric Measurements

Leaders: Hermann Smit, Jülich, Paul Monks, University of Leicester

C.A.M. Brenninkmeijer, MPI, Mainz, D Martine De Mazière BIRA, Brussels, BJohn P. Burrows, IEP, Bremen, D Ralf Sussmann, IFU, Garmisch-P., DA. R. MacKenzie, Lancaster, UK Julian Wilson, JRC, IspraAlain Marenco, OMP, Toulouse, F

Satellite measurements often complement work done at individual ground stations and in campaigns. Carefulcombination of the results should yield much more understanding than would be possible separately. Task group 3 willaddress the following issues.

Different satellite data sets will be combined to explore large scale NOx sources and assess the contributions fromlightning and biomass burning.

The comprehensive record of ozone and water vapour measurements with quasi-global coverage made since 1994from commercial aircraft within the European MOZAIC program will be used together with satellite data fromdifferent O3 and H2O profiling instruments to explore the processes governing the global tropospheric O3 and H2Obudgets [16, 17].

Exploration of the possibility of including space-borne measurements as an operational part in national air pollutionmonitoring networks.

Improved FTIR-retrieval strategies from ground-based techniques will be exploited for synergistic use with satellitedata (composite profiles).

The experience and the know-how achieved during the project are the first steps towards the establishment of an“Integrated Observation System” as part of an integrated research approach to explore the troposphere.

4. Development of Validation Strategies for Tropospheric Satellite Data Products

Leader: Hennie Kelder, KNMI, de Bilt

Ilse Aben, SRON, Utrecht, NL Bo Galle, Chalmers U. Göteborg, SD.W. Arlander, NILU, Kjeller, N Alex. Meister, DLR, Oberpfaffenhofen, DCornelis Blom, FZK, Karlsruhe, D Rudi Zander, IAGL, Liège, BQuality assurance and quality control of the satellite data is crucial to their use, and a substantial effort is being devotedto validation. Strategies proposed for the geophysical validation of tropospheric satellite data products are: collection ofcomparative measurements (ground-based, aircraft, and other satellite measurements); intercomparison of thesemeasurements with the satellite data; comparison of the satellite data with model results; analysis of different retrievalmethods; analysis of different measuring techniques; use of data assimilation methods.

A B

Fig. 4A: NO2 as observed by GOME in March 1997. A stratospheric (or geometric) air-mass factor (AMF) has been used totransform the slant columns to vertical NO2 amounts. Only GOME pixels with small cloud fractions have been used. UnfortunatelyGOME is much less sensitive to the boundary layer than to the stratospheric concentration of NO2, and thus the AMF is verysensitive to the assumed shape of the profile, as well as the surface albedo, cloud fraction and aerosols.

Fig. 4B: Here the AMF is computed on the basis of NO2 profiles as predicted by a chemistry-transport model. The result is probablya more realistic estimate of the true vertical NO2 column. In clean regions the main contribution to the NO2 column is stratospheric,and both figures show similar values. Above industrialised, polluted areas a large correction is needed in this approach to compensatefor the small sensitivity of GOME to tropospheric NO2.[18].

Among the validation activities envisaged are the following.

Validation of the tropospheric output from the ENVISAT instruments, SCIAMACHY, MIPAS and GOMOS.

Ground-based data to be used for validation will include the ozone sonde data available from the NILU/NADIRdata base for ozone in the free troposphere, the long term NO2, ozone, SO2, and aerosol measurements available inthe EMEP data base.

The validation comparisons with aircraft measurements will include those produced from the EC project,CARIBIC, data from the high-altitude aircraft M-55 Geophysica, and the MOZAIC data.

Attempts at validation using model results include comparing the SCIAMACHY CO and CH4 troposphericcolumns with results from the 3-D chemistry transport model, TM3.

Fig. 4 illustrates some of the problems in reconciling satellite and modelling work.

Other Underpinning and Derived Activities

A number of other aspects will be explored in the course of the project. These will include:

The development of appropriate data assimilation techniques combining satellite measurements with modelling.

The specification of the requirements for future satellite instruments for tropospheric work.

POTENTIAL APPLICATIONS OF SATELLITE DATA TO ENVIRONMENTAL POLICY DEVELOPMENT

As Fig. 5 shows for NO2, it is now possible to measure tropospheric concentrations of trace constituents ofenvironmental importance directly from space and, as algorithms and techniques are improved, concentrations of manyspecies at low altitudes will be observable. Combined with suitable modelling it will be possible to monitorconcentrations on a continuous basis.

Also, satellite measurements are likely to provide the only way to observe developments in the upper troposphere on acontinuous basis. Finally satellite measurements are in essence very reliable since, when launched, they are largely freefrom external influence. Thus, in the long term, satellite measurements are ideal for the determination of trends, such asis required for Montreal/Kyoto monitoring.

The various features commend themselves for future environmental monitoring and it is essential that the agenciesresponsible for atmospheric policy development in Europe have knowledge of and direct access to the rapiddevelopments expected in this field. Within the framework of EUROTRAC-2, TROPOSAT should provide such initialaccess.

Fig. 5: Tropospheric NO2 over Europe in 1997. The main VOC-limited regions are easily seen [19].

REFERENCES

[1] J:P. Burrows. “Current and future passive remote sensing techniques used to determine atmospheric constituents”,in Developments in Atmospheric Sciences 24: Approaches to Scaling Trace Gas Fluxes in Ecosystems, A. F.Bouwman, Ed. Elservier Amsterdam pp 315–347, 1999.

[2] J:P. Burrows. "Current and future passive remote sensing techniques used to determine atmospheric constituents",in preparation, 2000.

[3] M. Eisinger and J.P. Burrows. "Tropospheric sulfur dioxide observed by the ERS-2 GOME Instrument", Geophys.Res. Lett. vol. 25, pp. 4177-4180, 1998.

[4] J. Fishman, C.E. Watson, J.C. Larsen and J.A. Logan. "Distribution of Tropospheric Ozone Determined FromSatellite Data" J. Geophys. Res. vol. 95, pp. 3599-3617, 1990.

[5] J. Fishman and V.G. Brackett. "The climatological distribution of tropospheric ozone derived from satellitemeasurements using version 7 Total Ozone Mapping Spectrometer and Stratospheric Aerosol and Gas Experimentdata sets" J. Geophys. Res. vol. 102, pp. 19275-19278, 1997.

[6] E. Hegels, P.J. Crutzen, T. Klüpfel, D. Perner and P.J. Burrows. "Global distribution of atmospheric brominemonoxide from GOME on Earth-observing satellite ERS 2" Geophys. Res. Lett. vol. 25, pp. 3127-3130, 1998.

[7] C. Leue, T. Wagner, M. Wenig, U. Platt and B. Jähne. "Determination of the tropospheric NOX source strengthfrom GOME data" Proc. ISAMS Conference, Noordwijk Jan. 18-22, 1999.

[8] A. Richter, F. Wittrock, M. Eisinger and J.P. Burrows. "GOME observation of tropospheric BrO in northernhemispheric spring and summer 1997" Geophys. Res. Lett. vol. 25, p. 2683-2686, 1998.

[9] T. Wagner and U. Platt. "Mapping of Polar Tropospheric BrO by GOME, Earth Observation Quarterly, 58, March,21-24; Observation of Tropospheric BrO from the GOME Satellite" Nature vol. 395, pp. 486-490, 1998.

[10] J.P. Burrows and C. Leue, unpublished work.[11] EUROTRAC-2: TROPOSAT Subproject Description, EUROTRAC-2 ISS, München, in press, 2000: see also

http:members.gaponline.de/p.borrell/troposat[12] EUROTRAC-2 Project Description Handbook, EUROTRAC-2 ISS, München, 1999[13] R. Meerkoetter, B. Wissinger, G. Seckmeyer. "Surface UV from ERS-2/GOME and NOAA/AVHRR data: a case

study" Geophys. Res. Lett. vol. 24, pp. 1939-1942, 1997.[14] Wagner, T., C. Leue, M. Wenig, K. Pfeilsticker, U. Platt, Spatial and temporal distribution of enhanced boundary

layer BrO concentrations measured by the GOME instrument aboard ERS-2, Geophys. Res. Lett., 2000, submitted.[15] Hein, R., M. Dameris, C. Schnadt, C. Land, V. Grewe, I. Koehler, M. Ponater, R. Sausen, B. Steil, J. Landgraf, C.

Bruehl, Results of an interactively coupled chemistry-general circulation model: Comparison with observations,Ann. Geophysicae, 2000, submitted.

[16] M. Helten, H.G.J. Smit, W. Sträter, D. Kley, P. Nedelec, M. Zöger and R. Busen. "Calibration and performance ofautomatic compact instrumentation for the measurement of relative humidity from passenger aircraft" J. Geophys.Res. vol. 103, pp. 25643-25652, 1998.

[17] A. Marenco, V. Thouret, P. Nedelec, H.G.J. Smit, M. Helten, D. Kley, F. Karcher, P. Simon, K. Law, J. Pyle,G. Poschmann, R. Von Wrede, C. Hume and T. Cook. "Measurement of ozone and water vapour by Airbus in-service aircraft: The MOZAIC airborne program, An Overview" J. Geophys. Res. vol. 103, pp. 25631-25642,1998.

[18] H. Eskes, unpublished work.[19] A. Richter and J. P. Burrows, Tropospheric NO2 from GOME measurements, Advances in Space Research, 2000

submitted.