WORLD METEOROLOGICAL ORGANIZATION
COMMISSION FOR BASIC SYSTEMS
WORKING GROUP ON DATA-PROCESSING
TASK GROUP ON WMO/CTBTO MATTERS
Geneva, Switzerland, 15-17 July 1998
1.1 The Conjoint WMO/CTBTO Expert Meeting on Meteorological Data, Products and Services, and on WMO/IAEA Environmental Emergency Response Activities (Vienna, 1-5 December 1997) proposed to establish a small WMO task group and to invite CTBTO PTS experts to participate. The group would explain, discuss, and identify mutual needs and respective benefits and explore areas of possible future collaboration in areas related to meteorological measurements and transport modeling. In accordance with the terms of reference of the task group, most of the work in developing the draft report was accomplished by correspondence/e-mail.
1.2 The report was reviewed, finalized and relevant recommendations developed by the task group meeting held in Geneva from 15 to 17 July 1998. The meeting was chaired by Mr Roland Draxler (USA) the convenor of the task group and attended by two members, one alternate member of the task group and the WMO Secretariat as indicated in the list of participants given at the end of the report.
1.3 In general, the CTBTO has a number of requirements for meteorological data and services which could be fulfilled by WMO, such as observed and forecasted global meteorological data for 3D trajectory and dispersion calculations, transport and dispersion products from the RSMCs in the event of a nuclear event detection, meteorological observations, and routine weather forecasting to support on-site inspection teams. On the other hand, the WMO would be interested in obtaining meteorological observations from some of the CTBTO stations, using the CTBTO telecommunications system at certain locations to transmit data from WMO observing stations, and obtaining radionuclide measurement data that can be used for transport model development and verification.
1.4 Four major areas were defined in the group's terms of reference, which are attached to the Annex of this report and are dealt within the first four major sections of this report. The report then concludes with a set of recommendations for further action.
2. METEOROLOGICAL MEASUREMENTS AND DATA
2.1 General CTBTO Needs
Support the radionuclide measurement program
2.1.1 Meteorological data are needed by the CTBTO analysts in conjunction with the radionuclide signal to improve the differentiation between potential local interference or signals from more distance sources. In the case of no significant radionuclide signal at a site, but significant above-background measurements at other sites, meteorological data can be used to determine if local meteorological conditions may have prevented that sampler from measuring a radionuclide signal. The meteorological data will be used to deduce how representative local meteorological conditions are as compared with the local meteorological conditions forecasted by meteorological models, which are used for trajectory and dispersion calculations. The data will further be used by the CTBTO analysts to judge the possible influence of weather conditions on the performance of the station and to estimate the disturbance to satellite communications. Some of these requirements can be satisfied by data from nearby WMO meteorological sites, but some of the requirements can only be satisfied by meteorological measurements collocated with the radionuclide station.
Support to infra-sound source location calculations
2.1.2 Surface meteorological measurements will be used to estimate the effective speed of sound in the direction of wave propagation at the surface and to provide a quantitative measure of background noise due to wind-induced turbulence in the atmospheric boundary layer. These measurements are required to be collocated with the infra-sound array and very near to the ground surface.
Support to radionuclide source location calculations
2.1.3 If additional meteorological observations taken at CTBTO locations follow WMO observing standards, they should be inserted into the WMO Global Telecommunication System (GTS) network, and then could be assimilated by various national meteorological center's models to improve the gridded fields that are then used by trajectory and dispersion models in source location calculations. If not available over the GTS, but only at the CTBTO's International Data Centre (IDC), then those observations would still be available to CTBTO National Data Centres (NDC).
Support to the on-site inspection (OSI) program
2.1.4 OSI activities are weather dependent: ranging from general activities such as deployment of instruments and collection of data, to over-flights, sampling (atmospheric conditions can affect radioactive gas seepage from the soil) and even drilling. The inspection team would need weather data support during its mission of 25 to 130 days for the inspection area of 1000 km2 of any location. The meteorological information and data may include current weather conditions and forecast (wind velocity and direction, temperature, humidity, precipitation, atmospheric pressure), climatic conditions, historical meteorological records (for instance, to indicate the general weather background and transport pattern).
2.2 Summary of CTBTO technical requirements
Transport calculations at radionuclide stations
2.2.1 A radionuclide site requires wind direction and speed to estimate transport, temperature for pollutant mixing estimates, precipitation measurements to estimate particulate deposition, and atmospheric pressure to judge the representativeness of the observation with respect to the general synoptic pattern. Generally WMO stations reporting 4 to 8 times per day satisfy the requirements for transport modeling. The attached Table 1 and Table 2 show the current network of infra-sound, radionuclide, and meteorological stations. Out of a total of 80 radionuclide stations, 76 have a meteorological SYNOP station nearby that can meet the reporting requirements. One option is to ask those countries with nearby stations, but inadequate reporting frequency, to enhance their observations. At radionuclide stations with no nearby meteorological station (only 2 locations) it would be essential for CTBTO to assure that meteorological observations at those sites meet WMO standards and are injected into the GTS.
On-site meteorological measurements at radionuclide stations
2.2.2 Specifications for meteorological equipment at radionuclide sites are not very strict primarily because only qualitative deductions are drawn from the measurements. For example, heavy rain would explain a very low background of radon decay products, and strong winds might explain high concentrations of deposited radionuclides as a result of resuspension. In order to be able to explain equipment behavior, CTBTO determined to have a data and reporting frequency of the order of 10 60 minutes.
2.2.3 Although on-site meteorological measurements to support interpretation of the radionuclide data do not need to meet WMO standardization requirements, it would be desirable to follow the standard anyway so data can be fed to the GTS, and assimilated into meteorological models. Commercial vendors supply meteorological systems costing in the range of 5000 to 10000 USD that meet WMO standards. That is considerably less than the cost of a standalone meteorological station, because much of the station infrastructure and communications are already in place to support the radionuclide program. It would be more cost effective to use meteorological stations collocated with radionuclide monitoring sites.
Requirements at an infra-sound station
2.2.4 In many cases infrasound stations will be located inside a forest. An infra-sound site requires temperature, humidity, and pressure to estimate air density and wind speed to estimate turbulent noise. Measurements at a height of 10 m in the canopy of a forest will not accurately reflect conditions near the surface. CTBTO has determined that these observations will be made at a height of 2 m above the surface. Infrasound array stations require surface meteorological observations within 100 m of at lease one element in the array. There are no WMO stations that meet these requirements.
2.2.5 Infrasonic waves are propagated around the globe by refraction from layers in the stratosphere and lower thermosphere. The properties of these waveguides depend strongly on upper atmospheric wind and temperature profiles and this in turn can have a strong influence on the rate of attenuation, the degree of horizontal refraction along the ray path and the morphology of the observed waveform. The results of propagation models that are based on observed tropospheric, stratospheric and thermospheric wind and temperature profiles will be required by analysts to assist in the interpretation of observed infrasound events and to reduce the error on source location.
2.3 Summary of WMO reporting standards and observing practices
2.3.1 Although the CTBTO requirements for meteorological data can be achieved over a broad range of sensor siting, calibration, and reporting standards, WMO standards and practices are more specific for the measurement and reporting of wind, temperature, humidity, precipitation, and other meteorological variables. A condensed version of some of the principal WMO requirements is given below, primarily for information. It is recommended that for any observations reference should be made to WMO recommendations as reflected in several relevant manuals and guides.
2.3.2 Observation requirements
The following table is based on WMO recommendations, such as contained in the Guide to Meteorological Instruments and Methods of Observation (WMO-No. 8, 1996, World Meteorological Organization, Geneva).
|Variable||Unit||Range||Accuracy requirements||Reported resolution||Averaging time|
|Wind direction||°||0 to 360°||± 5%||10o||2 minutes|
|Wind speed||m/s||0 to 75 m/s||±
0.5m/s for £ 5m/s
± 10%for > 5m/s
|Air-temperature||°C||-60 to +60°C||± 0.1°C||0.1°C||1 minute|
|Dewpoint temperature||°C||-60 to +35°C||± 0.5°C||0.1°C||1 minute|
|Atmospheric pressure||hPa||100 hPa
(within the range from 920 to 1080 hPa)
|± 0.1hPa||0.1hPa||1 minute|
2.3.3 Siting criteria for sensors
The basic recommendations for siting sensors can also be found in the Guide to Meteorological Instruments and Methods of Observation (WMO-No. 8). The following guidelines are based on this Guide. In addition to this, it is highly recommended to consult the National Meteorological Service concerned regarding the specific nationally adapted instructions to widely guarantee homogeneous measurements on national/regional level, especially if standard conditions cannot be achieved for the observing site.
Pressure sensor: The difference between sensor elevation and field elevation should be less than 30 m.
Wind sensor: Should be mounted at 10 m height above ground of open terrain. Open terrain is defined as an area where the distance between anemometer and any obstruction is at least 10 times the height of the obstruction. Wind observations that are made in the direct wake of e.g. tree rows, buildings or any other obstacle are of little value and contain little information about the unperturbed wind. Since wakes can easily extend downwind to 12 or 15 times the obstacle height, the requirement of 10 obstruction heights is an absolute minimum.
Sensors for air- and dew point-temperature: The observed temperature should be representative of the free air conditions surrounding the station over as large an area as possible, at a height between 1.25 m and 2.00 m above the ground level. The height above ground level is specified because large vertical temperature gradients may exist in the lowest layers of the atmosphere. The best site for the measurements is therefore over level ground, freely exposed to sun and wind and not shielded by, or close to, trees, buildings and other obstructions. To ensure that the sensor is at true air temperature protection from radiation by a screen or shield is needed which also serves to support the sensor. This screen also shelters it from precipitation while allowing the free circulation of air around it and prevents accidental damage.
Precipitation gauge: The height of the exposure of the gauge should be selected according to the practice of the National Meteorological Service to widely guarantee homogenous observations. It should be noted that the effects on the wind field of the immediate surroundings of the site can give rise to local excesses and deficiencies of precipitation. In general, objects should not be closer to the gauge than a distance twice their height above the gauge orifice. Sites on a slope or on the roof of a building should be avoided. Sites selected for measurement of snowfall and/or snow cover should be in areas sheltered from the wind as much as possible. The negative effects of the wind, and of the site on the wind, can be reduced by using a ground-level gauge for liquid precipitation or by making the airflow horizontal above the gauge orifice by using wind shields around the gauge.
2.3.4 Reporting requirements
The basis for reporting observations should be the relevant WMO recommendations, such as laid done in the Manual and Guide on the Global Observing System (WMO-Nos. 544 and 488, respectively) and in the Manual on Codes (WMO-No. 306).
The following specific requirements should be considered:
Wind: In case of variable wind direction, the wind is variable when the mean wind speed is 2 m/s or less. Winds are also variable for directional variations of 60° or more with speeds greater than 2 m/s.
2.3.5 Field checks and calibration requirements
Regular field checks and calibrations of the instruments used are essential for guaranteeing the high quality and homogeneity of the observations. For field checks suitable travelling reference instruments should be used while sensor calibrations can best be done in a calibration laboratory. The time interval mainly depends on quality and long term stability of the sensors used and on the environmental conditions at the observing point. The instructions of the manufactures for maintenance and calibration of the individual sensors should be applied. However, it is strongly recommended that checks/calibrations should be performed at least annually. More detailed information on this issue especially related to the individual sensor types and their principal of work can be found in the Guide to Meteorological Instruments and Methods of Observation (WMO-No. 8).
3. PROVISION OF ATMOSPHERIC TRANSPORT MODEL PRODUCTS
3.1 Transport model requirements
A suspected event, first detected real-time from seismic or infra-sound network
3.1.1 The source location has been determined and atmospheric transport models are used to indicate when and which samplers may detect radionuclide products. Predictions are used to assure that the radionuclide network is operating at those locations where the radionuclides are forecast to go. Information may be used to obtain higher time resolution measurements to improve source location estimates if the initial location has large uncertainty.
A suspected event, first detected from (a) radionuclide sample(s) days after the event
3.1.2 No other information is available and the source location is unknown. Transport model products are then used to estimate potential source regions and event times that may have produced such a measurement distribution. Based on the location of existing measurements, transport model products may also be used to estimate the location and times of potential future radionuclide measurements.
The location and time of an event is known
3.1.3 Above-background radionuclide measurements have been made at one or more locations, then the transport model predictions can be used to estimate the total amount of material released by the event.
3.2 Data Requirements for the generation of transport model products
3.2.1 Trajectories (an indication of transport direction) and air concentrations (an indication of transport direction and pollutant dispersion) require the specification of meteorological fields of wind, temperature, moisture, that vary in space and time. Historical calculations, for events prior to the current time, require an archive of these fields, while future predictions require the same fields as output from a meteorological forecast model. Transport model calculations would then use these fields as the basis for their computations.
Meteorological archives and forecasts
3.2.2 All major meteorological centers maintain archives and corresponding model forecasts which, depending upon resolution, may generate a data volume of up to 200 Mb per day for global coverage. Although the meteorological observations that initialize these models are more compact, the need for transport model calculations over data sparse regions, requires the use of processed and gridded meteorological fields.
Centers for meteorological data and modeling
3.2.3 Although the International Data Centre (IDC) could maintain its own basic meteorological modeling capability to generate high resolution archives and forecasts, it would be more cost-effective for these data to be provided by the existing meteorological centers, considering that these models require a complex assimilation of meteorological data from a variety of different sources. However the IDC could easily maintain a suite of different transport models (appropriate for various space and time scales) that could be run by analysts at the CTBTO as required for each event using meteorological data obtained from one or more of the WMO centers. A third possibility would be that existing meteorological centers would provide pre-computed transport and dispersion estimates for a grid of source locations in their geographic area of responsibility and provide only the transport model results to the IDC.
3.3 Types of transport model products
3.3.1 Several different transport model products are summarized here that can be used in conjunction with data from the radio-nuclide system to estimate the location of the source and project future air concentrations downwind of existing measurement points.
3.3.2 A trajectory is the time-integration of the position of a particle in a spatially and temporally varying velocity field. Many of the simpler modeling approaches assume that the center of mass of a pollutant cloud will follow the trajectory path. Trajectories may be computed backward in time to indicate the upwind path that the pollutant has followed to reach a receptor location or forward in time to indicate the future position of the particle.
3.3.3 The turbulent mixing of pollutants into regions of large variations in wind speed and direction with height, and pollutant interactions with weather systems on a variety of different spatial scales, frequently preclude the use of simpler single trajectory methods to estimate pollutant transport. Simpler methods are more difficult to interpret when for instance, trajectories show significant curvature, severe changes in direction, and large variations in location with starting height or over the duration of the sample collection period.
3.3.4 In these situations more complex dispersion modeling approaches are required that simulate the evolution of the pollutant either through the use of many trajectories to describe it's horizontal and vertical distribution or through direct numerical integration of the advection-diffusion equation on three-dimensional grids. These models can provide direct estimates of air concentrations or normalized air concentrations, if the source strength is unknown. Normalized air concentrations are essentially the atmospheric dispersion factor between each source and receptor, or a measure of the probability of a concentration measurement at a particular point for a unit emission.
Estimating the location of a source
3.3.5 Locating the emission source requires inverse modeling methods. As noted by the Montreal meeting, this is an area that still requires more research and development to reach the same level of maturity as the other detection methodologies, the accuracy of which, in addition to atmospheric processes, will be determined by the number, geographical orientation, and duration of any samples collected as well as any information, such as the time of the event, that may be available from the other detection networks.
3.3.6 In general, atmospheric transport and dispersion is dominated by the transport processes - advection by the winds. Therefore in many situations, especially when complex weather systems do not interact with the pollutant, a simple backward trajectory calculation, using winds at appropriate levels of the atmosphere, can provide a reasonable estimate (at best with an uncertainty of 10% to 20% of the travel distance) of the track along which the source may be located. As noted above in paragraph 3.3.3, simple trajectory methods are most suitable when there is a consistency in space, height, and time, of their indicated transport pathway. This indicates that regardless of how the pollutant has mixed due to turbulent processes, it will most likely follow the centerline path of a single trajectory. With measurements and corresponding backward calculations from multiple locations (or times) it may then be possible to triangulate a source location.
3.3.7 The frequency at which simpler trajectory approaches versus more complex modeling scenarios would be required, can be estimated from climatological factors. For instance, trajectory methods could easily be applied at tropical latitudes, with directionally consistent trade-winds, than at the higher latitudes, with more frequent cyclonic storms. Appropriate seasonal differences should be considered.
3.3.8 Adjoint methods use a special linearized version of the dispersion model that is run backwards in time from measurements, to provide initial source conditions for the full forward dispersion calculation. The process is iterated until the differences between model predictions and measurements are minimized. The approach is very appealing and provides an integrated and physically consistent solution to the source estimation problem. Although transport is a fully reversible process and atmospheric dispersion is not, adjoint methods should yield comparable results to back trajectory methods, but in more rigorous and quantitative manner.
3.3.9 A third possibility would be to use a transport and dispersion model to estimate the air concentrations (or the probability of a source impact) at all receptor locations independently for all points in a source grid. The result would be a source-receptor matrix, where each receptor column for a specific source row would be the contribution to the air concentrations a that receptor from the selected source, while the each source row for a specific receptor column represents the contribution of each source to the air concentration at that receptor. These type of matrix methods have been developed in the 1970's to estimate contributions of various regions to transboundary air pollution. This method may require significant computer resources.
3.4 Role of WMO's RSMCs in supporting CTBTO operations
Providing gridded meteorological observations and forecasts
3.4.1 Meteorological observations could be provided to the IDC through a local link to WMO's GTS. However gridded meteorological analysis and forecasts are not available over the GTS at the spatial resolution and temporal frequency that would be required for transport model calculations. Global data may be most easily obtained through a dedicated circuit (or Internet connection) to the model products available at most European meteorological centers. A recommendation of the Montreal meeting was for designated WMO RSMCs to provide gridded meteorological fields to the IDC.
Providing dispersion modeling capabilities to the IDC
3.4.2 Transport modeling capability (trajectory and dispersion) could be provided by RSMCs who will install, train, and support the software at the IDC. Again in concurrence with the recommendation of the Montreal meeting, coarse or first-cut calculations can be made at the IDC, using these models and the appropriate gridded meteorological fields.
Providing higher resolution predictions over selected geographic domains
3.4.3 One role for the RSMCs and their respective meteorological centers is if an event is detected in their geographical region, they will have meteorological model fields, analysis and forecasts, that are of much higher resolution, that could be obtained from a global center. In that situation, the IDC could either request the higher resolution fields from the appropriate center or request the center to run its own transport model products for the event. Consistent with the Montreal meeting, WMO designated centres would respond with more sophisticated modeling results. Multiple centre results could be compared.
Organizational Requirements for Collaboration between WMO and CTBTO
3.4.4 One possible collaboration would be for WMO to arrange for one of the global meteorological centres to provide daily gridded global analysis fields and forecast fields to the IDC. The analysis fields would be archived at the IDC in case they are required for transport simulations. The IDC would have the capability for routine trajectories analysis or dispersion calculations. In the case of a confirmed nuclear explosion or detection of multiple fission products, the IDC may request predesignated RSMCs (similar to the current arrangements with the IAEA for nuclear accidents), to run their higher resolution meteorological models, to produce a set of predefined products, either from the source location, if known, or from designated sampling locations, if that is the only information. The results should be returned electronically to the IDC within 3 hours, again similar to the current arrangements with the IAEA. Data formats and other product details could be negotiated between IDC and WMO experts.
4. SHARING DATA AND MEASUREMENTS
Provision of WMO standard measurements to CTBTO
4.1 Regardless of whether the CTBTO decides to supplement their stations with meteorological measurements and transmit those to WMO, or assists WMO with communications at some remote WMO locations, the IDC will still require access to data from WMO's global meteorological observing network. This suggests that it will be necessary to establish an Internet or other telecommunications connection to a nearby National Meteorological Center, such as the one in Vienna.
Provision of CTBTO on-site measurements to WMO
4.2 Meteorological data from CTBTO sites with WMO acceptable site criteria and reporting practices would be transmitted from the station to the IDC using the CTBTO VSAT system and then to the international meteorological community through a linkage from the IDC to the nearest GTS hub.
Provision of other meteorological and gridded model products
4.3 Routine access to other meteorological products, analyses forecasts, and gridded model outputs, could be arranged between the CTBTO and one or more of WMOs RSMCs. These data would be used for transport model calculations as well as supporting on-site inspection teams.
Sharing radionuclide data
4.4 It should also be noted that radionuclide data, once associated with specific emissions or natural sources, can be used to improve and further develop the transport models, a valuable interaction between WMO and CTBTO. However, at the present time it is not clear if the CTBTO is in the position to release these data to the WMO centers.
5. SHARING TELECOMMUNICATIONS
5.1 Some WMO meteorological stations have poor communications linkages with the GTS resulting in long reporting delays or missing observations. It may be possible to use the CTBTO VSAT to transmit local observations at meteorological sites near CTBTO stations to WMO using the link between the IDC and the nearby GTS hub. One concern is that it might require an additional local communications link between the meteorological observation site and the CTBTO station.
6.1 Considering the mutual benefit that can be drawn from their respective activities, the task group recommended that the CTBTO and WMO cooperate in areas related to meteorological measurements and transport modeling, by keeping each other informed on their respective activities in these matters. More specifically, the task group proposed the following recommendations.
Additional Meteorological Measurements
CTBTO radionuclide stations require on-site meteorological measurements. It is recommended that these measurements should conform with WMO observing standards. It is noted that WMO would benefit from CTBTO inserting into the GTS meteorological measurements from remote locations which have no nearby observing stations.
It is recommended that CTBTO investigate the option of co-using suitably located WMO meteorological stations for radionuclide monitoring.
Meteorological measurements at infrasound stations will not meet WMO observing standards nor can nearby WMO meteorological stations be used by the CTBTO.
Atmospheric Transport Model Products
It is recommended that WMO and CTBTO develop arrangements to provide transport model products by the RSMCs with activity specialization in Environmental Emergency Response.
It is recommended that WMO address CTBTO's request for RSMCs to provide transport models and other assistance to CTBTO to enhance their in-house capabilities.
Sharing Data and Measurements
It is recommended that WMO and CTBTO develop arrangements to provide the RSMCs gridded meteorological model output data to the CTBTO to support in-house modeling capabilities.
To provide meteorological support to the CTBTO on-site inspection program, it is recommended that CTBTO initially use the gridded forecast fields provided by the RSMCs and then develop other arrangements as necessary.
It is recommended that the CTBTO radionuclide observations be provided to WMOs RSMCs for validation and improvement of model products.
It is recommended that a data link be established between CTBTO and a suitable NMC (e.g. Vienna) for the purpose of routine access to WMO surface observations and to inject CTBTO meteorological measurements into the GTS.
It is recommended that WMO, in consultation with WMO Members concerned, compile a list of their meteorological stations that have poor data collection facilities and are near CTBTO stations and request that CTBTO consider the feasability of using their VSAT communications system to inject those observations into the GTS by establishing local links with such stations in cooperation with the meteorological service.
Terms of Reference of the WMO Commission for Basic Systems (CBS)/Working Group on Data Processing (WGDP)/Task Group on WMO/CTBTO Matters
The Task Group (TG) will study and make recommendations on matters related to the requirements and implementation of the CTBTO radionuclide and infrasound monitoring operations as regards the meteorological aspects involved.
Scope and issues:
The TG will:
Develop expert recommendations for formulating the requirements of the Provisional Technical Secretariat (PTS) of the CTBTO related to meteorology including the technical issues, standards and practices involved in making and reporting meteorological measurements;
Develop proposals for the generation and format(s) and for arrangements for the provision to the PTS of atmospheric transport model output products to meet its specific needs;
Study the opportunities for sharing meteorological measurements between the CTBTO and relevant WMO programmes and develop pertinent recommendations;
Identify possibilities for sharing telecommunication means for transmitting meteorological measurement data;
Working arrangements, consultation and reporting
The TG will work by correspondence/e-mail. The final report will be coordinated in a meeting. The report, preferably in a condensed version, will be submitted for consideration to the forthcoming extraordinary session of the CBS through the chairperson of the CBS/WGDP. The full report of the TG will be available to both Secretariats for their convenient use.
The TG shall be comprised of Roland Draxler (USA), convenor and Jean-Pierre Bourdette (France) and two invited experts from the PTS.
The TG shall begin its work as soon as possible and should complete its work with a report not later than June 1998.
LIST OF PARTICIPANTS
Mr Roland DRAXLER (Chairman)
Air Resources Laboratory
Office of Atmospheric Research
1315 East West highway
Silver spring, Maryland
Tel.: (001-301) 713 0295
Fax: (001-301) 713 0119
Mr Jean-Pierre BOURDETTE
Service Central d
Mr Lars-Erik DE GEER
P.O. Box 1200
Mr Matthias AUER
P.O. Box 1200
Mr Dieter C. SCHIESSL
41 ave Giuseppe-Motta
Case postale No. 2300
CH-1211 GENEVE 2
Tel: (41 22) 7308 369 / 7308 231
Fax: (41 22) 733 02 42
Mr Morrison E. MLAKI
41 ave Giuseppe-Motta
Case postale No. 2300
CH-1211 GENEVE 2
Tel: (41 22) 7308 231
Fax: (41 22) 733 02 42