by Philip GOLDSMITH

Man has been observing the Earth- his planet-for many hundreds of years. Cartography dates back to the ancient Greeks and Romans. The first networks of meteorological stations were set up nearly 200 years ago. Earth observation from space is, on the other hand, comparatively young and recent. Its impact during the course of the last 25 years has nonetheless been widespread. For example, global data from polar-orbiting and geostationary satellites are now fed routinely into operational meteorological models, significantly enhancing the quality of weather forecasts around the globe. And remote sensing from space is being routinely applied to areas as diverse as vegetation monitoring, agriculture, forestry, geology, hydrology, land use, cartography and glaciology.

However, during the last few years a new element has appeared: it has become more and more widely accepted that some of the major challenges facing mankind today are environmental in. nature. Mankind is now aware that its activities may disturb the delicate balance which determines the Earth's environment. Linked to this concern is the realisation that the Earth has only limited resources.

Today, it is obvious that industrial development on the one hand and both global and regional environmental issues on the other are closely linked together. However, this finding is very recent indeed. When Europe's industrialisation started about 200 years ago, environmental considerations did not play any role in the transition from an agricultural to an industrialised society. In addition, rapidly growing consumption required supply of renewable and non-renewable . resources from outside the industrialised nations.

It was not until after 1945 that genuine development policies were established. In parallel, as pollution began to be perceived as a serious local by-product of industrialisation, most of the developed countries began to look for and implement measures to contain it. A few years ago, another turning point was reached: the recognition of global environmental challenges, such as climate change caused by greenhouse gas emission, deforestation and destruction of the stratospheric ozone layer. The discussion since then has established the point that there is no such thing as a local or regional solution to global problems. Thus, development policies need to be tied in with environmental considerations.

This requirement is now largely accepted and the problem of how to satisfy it has moved to the top of the agenda. Solutions to environmental problems are increasingly sought on a global level, and progress has been made recently with, for example, the Montreal protocol on the limitation of CFC emissions and the subsequent agreements.

This article does not set out to propose actual political solutions or suggestions for reconciling development and the environment. That is a political task. Satellite data cannot take the burden from the shoulders of those who bear the political responsibility for finding political solutions to political problems. But satellite data can constitute the indispensable tool to assist them in making decisions or concluding agreements, and will help them implement political measures. This article describes contributions that remote-sensing satellites will make to both the environmental and the developmental aspects of these problems.

Looking at this issue in depth now is very relevant. The forthcoming United Nations Conference on the Environment and Development in June 1992 will be a forum where these issues will be discussed in depth and with unprecedented political impact. The European Space Agency, ESA, has requested observer status at the conference. It is ESA's objective to make its expertise in space technology in general, and observation of the Earth and its environment from space in particular, an asset at the conference for the benefit of all the nations of the world.

The dimensions of the problem

The global dimension of some environmental problems, in terms of their causes and consequences, has already been mentioned. It is now also apparent that only a global scientific effort, making extensive use of satellite data, can meet these daunting challenges. In parallel, remote sensing from space can be put to work immediately to solve a certain number of regional environmental and development problems. This section outlines these two aspects of the observation of the Earth and its environment.

International research into the environment: the global dimension

Today, the Earth is perceived as a single, intricate system with physical, chemical and biological interactions occuring between the atmosphere, the oceans, the land and the ice regions. A big step forward in our understanding of the planet was taken when international, interdisciplinary research programmes were set up. These are now rapidly gaining momentum. Most notably, the International Geosphere-Biosphere Programme and the World Climate Research Programme seek to advance the understanding of basic processes that control the Earth system. Essential to both is the provision of global data.

Since most of the processes involved; cannot yet be adequately modelled in the laboratory, only observation of the Earth, both on the ground and from space, can deliver the comprehensive sets of data needed for the required research. Indeed, all disciplines of environmental research are making rapidly growing demands on governments and space agencies for Earth observation data, in particular from space. That is why the European Space Agency (ESA) has put observation of environmental factors to the forefront of its Earth observation programme.

Trends and changes can be monitored, and their driving forces understood, only if continuity of data over long periods is guaranteed. Here, decades rather than years are suitable units of observation : time. To this requirement must be added , timely delivery and coordination of data streams.

The need for a truly international global research effort is now largely accepted. A main objective of ESA's Earth observation programme is to make a significant contribution to this effort. It is of the utmost importance for everyone involved in Earth observation to seek collaboration and try to take full account of similar programmes around the globe. In space-based Earth observation, complementarity of instruments and orbits can now generally be taken for granted. Close cooperation has been established between ESA and bodies such as NASA and NOAA in the US, NASDA in Japan, the other national space agencies in Europe and Canada and user entities such as WMO, the EC and Eumetsat.

Regional monitoring and management: the local dimension

Remote-sensing data are now used on an operational basis for a number of applications, such as agriculture, forestry, geology, hydrology, land use and cartography. These applications are of immediate interest to a growing number of Third World countries as aids to the management and exploitation of their resources. In addition, satellite data can be used in support of certain local environmental tasks such as pollution assessment.

Established procedures for urban planning from the developed countries often fail in developing countries because of lack of data on which decisions could be based. For example, the rapid, often chaotic, development of some of the big cities in the world can only be monitored continuously by observations from space.

Satellites: reconciling local and global dimensions

Contrary to occasional false impressions, satellites are not a costly source of data that could be obtained otherwise by Earth-bound instruments. In many cases, this option is not open and satellites are ideally suited to covering both the global and local dimensions: their measuremeets are global, but can still be used to monitor and manage local phenomena inremote areas or in developing countries with limited planning and statistical capabilities.

Global coverage is of prime importance for research into the environment. In addition, if a single type of instrument or mission provides global coverage, this ensures the coherence of data sets and helps to minimise systematic and statistical errors. This is why coordination between space agencies to achieve interchangeability of data sets is so important.

In addition to providing global coverage, only satellites have the flexibility to carry out repeated measurements over a particular location at short notice. Consecutive observations can be as little as three days apart in the case of the ERSI satellite, and range from 15 to 30 days in the case of satellites carrying imaging radiometer payloads, such as SPOT and Landsat, while geostationary meteorological satellites, Meteosat for instance, I deliver an image of the Earth with a resolution of 2.5 kiLométres at intervals as short as 30 minutes.

Finally, satellites are the most efficient means of obtaining comprehensive data from very extensive or remote areas such as polar regions, oceans, deserts and rain forests. For urban or more densely populated areas, or for management of natural resources, combining satellite data with aerial photography data or data gathered on the ground can be very advantageous.

Developing countries' requirements and opportunities for using environmental and resource management data

Developing countries generally do not sustain space programmes on their own, apart from a few notable exceptions such as China, India or Brazil. Nonetheless, developing countries can participate in observations of the Earth and its environment from space by using the data.

First, all countries should have the right of access to civil space data and, I most importantly, data from remote-sensing satellites. It was this requirement that led to the emergence of what is now called the 'open sky policy'.

However, the generally accepted open sky policy has to be distinguished from free availability of data for use in development activities. Data from meteorological satellites are usually made available free or at marginal cost. In general, however, data from remote sensing satellites are widely regarded as commercial goods for which a market price has to be paid. This situation worries not only developing countries but also a section of the worldwide scientific community which often has difficulties purchasing the data required for basic science projects. This is why it is now generally the practice to refer to 'ready access' to remote-sensing data as an adjunct to the open sky policy.

The European Space Agency has drawn up a data policy which seeks to reconcile the commercial exploitation of remote-sensing data and their ready availability for scientific and development purposes.

This policy is founded on three cornerstones. First, ESA sets up joint ventures with large institutions fostering science or development. Here the ESA contribution is either remote-sensing data as such or consultancy support on the use of data. One of the most advanced of these joint ventures in the field of the environment and development is the TREES tropical rain forest mapping project. In this case, the EC and bodies in numerous states in the tropical belt have become ESA's partners. Other projects have been started or are under consideration.

Secondly, the global environmental research effort is assisted by the appointment, by ESA, of scientists, called Principal Investigators, who are given privileged, free access to the data they require for their basic research projects. These scientists do not have to be from ESA Member States. Their results have to be published and are thus available to the entire scientific community.

Thirdly, pilot projects explore the viability of future applications of remote sensing data. Once a pilot project has been accepted by ESA, the participating institutions and bodies obtain privileged access to remote-sensing data. Several ESA pilot projects concern ventures in developing countries.

Another requirement is the setting-up of a global network of ground stations to receive and process data, most notably from synthetic aperture radar instruments, SAR, on Earth observation satellites. Since SAR data rates are so high that tape recorders cannot be applied for on-board storage, the data stream has to be directed to a ground station as it is generated by the SAR. Then, dense networks of ground stations ensure that data relevant to regional or local issues actually reach the recipients at the least possible cost.

Decentralised data reception and processing are particularly important since data rates are expected to rise even further in the medium term. This is why ESA closely cooperates with operators of many ground stations in developing countries, for example in Brazil, Ecuador, India and Thailand. For the longer term, the data relay satellite systems will be a more cost-effective means of dealing with the high data transmission rate requirements.

A final, very important point concerns training of experts and the transfer of technology or expertise in data archiving, processing and interpretation. Much effort has been devoted to these areas over the past few years. Many universities and research institutes in developing countries are now offering remote-sensing courses and have purchased equipment to handle and interpret remote sensing data.

In this area too, it cannot be the task of a space agency to organise training and transfer of technology on its own, since these activities need to be carried out jointly with bodies directly involved in development issues. ESA has sponsored remote-sensing training courses in the past and plans to step up this activity over the coming years. This activity is being developed in close cooperation with the UN, the FAO, the EC and various entities from ESA Member States.

The following parts of this section expand on these general considerations by discussing the prospects for developing countries in three important areas of environmental observation by satellite: climate change and pollution, resource monitoring and management, and meteorology.

Climate change and pollution

It has already been stressed that understanding and monitoring large-scale environmental change necessitates a global research effort to which satellite data will contribute significantly. For example, plans are now in hand to set up a Global Climate Observation System to build on the established, WMO-inspired World Weather Watch, in which extensive use will be made of satellites.

No comprehensive model capable of explaining the Earth's climate exists as yet. On the contrary, we are only just starting to become aware of the full range of factors which actually influence our climate. In the important field of air-sea interaction, for example, the European Space Agency's ERS-I satellite started last year to deliver global data sets of sea surface wind fields, ocean wave fields, ocean currents and highly accurate sea surface temperatures. In atmospheric chemistry, sounding of ozone and other trace gases in all layers of the atmosphere is of the utmost importance. However, the development of suitable measuring instruments is still in its infancy. Global data on precipitation and evaporation, or the emission and absorption of radiation, can only be delivered by satellites.

The technical, operational and financial demands of running these satellite systems are very heavy, both for the satellites themselves and for the ground segment. The actual research effort aimed at a comprehensive understanding of the global Earth system is therefore principally a matter for the developed countries. Development issues are all the more important in the subsequent stage, i.e. the formulation of policies to minimise the impact of large-scale or global environmental change. The burden of decision making and the consequences of meeting environmental challenges have to be distributed equitably. In particular, these policies must not militate against development objectives.

Although the scientific community has only limited influence on these political processes, it can contribute to the formulation of the terms of reference for political discussion in which due weight will be given to development issues: first, the relevant research programmes have to be accessible to every interested scientist; secondly, the results have to be made available to the entire scientific community and, thirdly and most importantly, the results have to be published in a comprehensible and readily accessible form, free from bias in their interpretation.

Resource management and monitoring

The monitoring of agricultural and renewable resources has already made considerable progress in the use of data from satellites, such as the French SPOT and the US Landsat satellites. It is now possible, for example, to identify different types of vegetation, enabling the deforestation of the tropical rain forests and acid rain damage to the European and US forests to be monitored. ESA itself is collaborating with the European Community in developing satellite-based methods for identifying vegetation types.

This research is of the utmost importance for crop yield forecasts and hence managing food supply on a regional or global scale.

In the general area of agriculture, forest inventories and water management are further items high on the list of desirable applications. It is also highly desirable to develop systems for crop control and early warning of crop failure, for example due to diseases, in order to initiate remedial action.

The monitoring of non-renewable resources is another important activity, touching on geology, mineralogy, hydrogeology and tectonics. The principal objectives are to use satellite data to prospect for minerals and hydrocarbons and to help manage groundwater supplies. Surface water resources outside agricultural use (lakes, ice and snow) could also be monitored and useful information provided for studies on soil erosion, volcanic activity and earthquakes.

As a first step, ESA and the FAO have started setting up a system to disseminate Earth observation data. The primary objective of this DIANA project is to help improve food security in Africa by transmitting forecasts and early warnings of imminent natural disasters through the FAO's African infrastructure. These will | be based on pre-processed and analysed l remote-sensing data, supplied by various European and US satellites and distributed by ESA's own telecommunications infrastructure.

Another project concerns data from the AVHRR instrument on board the US TIROS satellite. These data are highly relevant for a number of resource monitoring applications. ESA has initiated, together with the EC, reception of AVHRR data at three African ground stations in Kenya, Niger and La Reunion (France). A coordinated network of these three African, five European and two ground stations in the Antarctic has since been generating standardised AVHRR data products and a single catalogue. There are plans to extend the network jointly with the EC and the FAO, by , ground stations in Egypt, Malaysia, Indonesia, the Philippines and Brazil.

Requirements in the area of resource management and monitoring are generally complementary to those for environmental studies. The relevant research or operational effort is usually provided locally or regionally. The number of projects is large, making it reasonable to decentralise them as far as possible. Decentralisation is also justified by the fact that individual projects are usually small in comparison with an international effort in environmental research. In most cases, resource management and monitoring exploit spectrally resolved satellite imagery, in the near future supplemented by radar imagery. The space segment is thus provided by the existing space agencies. The requirements of the developing countries in terms of ready availability of data, ground processing capacity and training have been discussed above.

Meteorology

Meteorology was the first civil application of Earth observation from space. Nobody can afford to neglect weather forecasting, which has an enormous impact on many areas of economic activity: agriculture, transport, leisure, etc. Weather forecasts have improved considerably over recent years, thanks to satellite imagery and steadily increasing computer power.

All weather forecasts depend on weather observation data coupled with numerical calculations. Current numerical methods provide forecasts for up to seven days ahead. But, even with more computing power available, input data of higher precision and better spatial resolution, and covering more variables, are required if further improvements are to be made. These objectives as discussed in the frame-work of the WMO will be taken into account by the space agencies.

For regional and local forecasts, satellite imagery and meteorological products have to be made available around the globe. This is the data dissemination function of the satellites of World Weather Watch, which equally ensures that all meteorological services are able to produce forecasts of a quality similar to that in developed countries. ESA and Eumetsat have contributed a series of six Meteosat satellites to this system, the first of which was launched nearly 15 years ago. Since weather forecasting is largely seen as a public service, and since reliable forecasts for all areas of the globe are vital to safe and reliable transport worldwide, there are few, if any, obstacles to the provision, to developing countries, of the data required to give reliable weather forecasts.

Current and future programmes

Earth observation is now occupying a prominent place in the space programmes of all space-faring nations. The space segment comprises well-known programmes such as Meteosat, GOES, SPOT, NOAA and Landsat, ERS and WARS, Topex/Poseidon and MOS. Cooperation is high on the list of priorities, particularly where the exchange of data, complementarily of orbits and exchange of instruments are concerned. Over the past three years, ESA has formulated an Earth observation strategy and a longterm plan. Both are centred on environmental questions, with operational meteorology, solid Earth and land applications as further cornerstones. Similar plans exist in the United States, based on the EOS-system being developed by NASA and NOAA, and also in Japan.

This section focuses on current and future European systems. However, it should be borne in mind that the other space-faring nations' efforts are comparable to ESA's, and are being coordinated with it.

Meteosat

The European Meteosat satellites were developed by ESA in the 1970s as part of an experimental programme. This programme became fully operational a few years ago and the financing and management of Meteosat-4, Meteosat-5 and Meteosat-6 was entrusted to a dedicated organisation, Eumetsat. ESA retains responsibility for the development, launch and orbital operations of these spacecraft. The Meteosat satellites image the Earth's disc every 30 minutes in three spectral bands-visible, water vapour and thermal infrared-with a resolution of between 2.5 and 5 kiLométres.

Many products are routinely generated from these images: cloud motion vectors, sea-surface temperatures, cloud top height maps etc. These data are used not only by meteorologists but by many other disciplines. The Meteosat database of over 550 000 images spanning nearly 15 years amounts to an impressive climate archive.

The present Meteosat system will provide operational service until at least 1995, and an extension probably until the end of the century. Studies are already in hand for an improved system, Meteosat Second Generation, which is envisaged as a joint ESA-Eumetsat programme.

ERS-1

ESA's first remote-sensing satellite, ERS-1, represents Europe's biggest step forward in this field since the launch of the first Meteosat satellite. ERS-1 was launched on 17 July 1991 and carries three active all-weather microwave instruments: a synthetic aperture radar, a wind scatterometer and a radar altimeter. These are complemented by an alongtrack scanning infrared radiometer giving global sea-surface temperature measurements of unprecedented precision.

This complementary set of instruments provides data for research into phenomena of the greatest relevance to environmental research: interaction between the oceans and the atmosphere; ocean circulation and energy transfer processes; establishment of the mass balances of Arctic and Antarctic ice sheets; coastal processes and pollution; and land use change.

The last two of these areas also provide examples of areas of practical applications, mainly fisheries, offshore activities and resource management.

To make the best use of ERS-1 instruments, much of the data is processed and disseminated in near-real time, i.e. within three hours of observation. Cataloguing, handling of requests, data quality control, scheduling etc. are performed at a central facility at ESA's ESRIN establishment in Frascati, near Rome. Five ERS-1 ground stations have been set up in Europe and Canada and provide the fast-delivery services with global coverage. In addition, four processing and archiving facilities in Europe generate a multitude of off-line precision products and are responsible for archiving data and products. Numerous foreign ground stations, notably in Japan, Thailand, Indonesia, India, Brazil and Ecuador, ensure the greatest synthetic aperture radar coverage possible and ready access for these countries to ERS-1 data.

The full potential of ERS-1 can only be realised if continuity of data is ensured to the users. A second satellite, ERS-2, will take over when the operational life of ERS-1 ends in 1994. However, ERS-2 is not just a 'copy' of ERS-1, but will in addition carry a Global Ozone Monitoring Experiment (GOME) to measure the content of ozone and some related gases in the upper atmosphere. The relevance of this experiment to the current concern over the Arctic and Antarctic ozone holes is obvious.

Polar platforms

As Earth observation data come to be used by more and more disciplines, future missions will be broader in scope and Earth observation satellites will have to carry ever more instruments. The trend is illustrated by the following figures: the Meteosat payload weighs just over 100 kilograms, that of ERS-1 about 800 kilograms, and the next generation of missions is being planned for payloads of about 1,700 kilograms.

These future missions will involve a series of polar-orbiting platforms, based on the Columbus polar platform now under development and due for launch in 1998. Each satellite has an expected lifetime of about 4.5 years.

The exact composition of the payload has not been finalised. It will probably comprise three packages: first, a meteorological package of instruments similar to those currently flying on the American Tiros-N operational meteorological satellites; an atmosphere/ocean/ ice package, including improved ERS-1 instruments; finally, an announcement of opportunity package, consisting of instruments selected on a competitive basis according to topical research. Lastly, the payload will include a set of instruments for studying ozone and related atmospheric trace gases.

Earthnet

It is important to realise that the European ground-based system for acquisition, archiving, cataloguing, preprocessing and dissemination of huge amounts of data has to be substantially expanded to match the development of new generations of satellites. There is little point in flying instruments without ensuring that the data reach the users. That is why ESA set up the Earthnet programme as early as 1977.

The Earthnet policy is based on the recognition and assurance of data availability over long periods of time. Another regular service to users has just started with the launch of ERS- 1 and will continue into the next century.

In addition to data from ESA's own, satellites, data from a variety of other satellites are handled by Earthnet. These include SPOT, Landsat, MOS-1, Tiros-N and, in the past, Seasat, Nimbus-7 and the Heat Capacity Mapping Mission. Earthnet operates ground stations at five I European sites and acquires data from I three others. The programme office is located at ESA's ESRIN establishment in Frascati.

Intensive cooperation with the EC has started with a view to creating an Environmental Data Network, EDN, which will ensure that data from the various sources are made available in the highest quality possible to all users.

Conclusion

The course of events has led to a situation, in Europe, in which the organisational structures concerned with satellites and space technology development are separated from those in charge of development policy and action. Raising or keeping barriers between the two is to be avoided, since ever closer cooperation is required. The European Space Agency has focused its Earth observation policy and programmes on environmental issues and has opened up its activities to joint projects in the field of development. Contributing to the reconciliation of environment and development policies is now firmly among the high priority objectives of ESA. P. G.