Analyst in the sky satellite-based remote sensing In April NASA scientists released remarkable new images the Ærst from the Terra satellite launched in December last year. The data Øowing back from Terra and similar missions to be launched over the next few years promises a revolution in our understanding of the Earth's climate and environmental systems. Scientists have been successfully using remote sensing (RS) from space to monitor and study the Earth for more than 30 years. The data sent back by these instruments have helped researchers build realistic models of environmental processes and have contributed signiÆcantly to our understanding of complex phenomena such as ozone depletion and climate change. Increasingly remote sensing1 technologies are employed alongside geographical information systems (GIS) for long-term monitoring of environmental parameters; short-term Box 1 The Terra Mission The Terra satellite was launched by NASA from the Vandenberg Air Force Base California on 18 December 1999.Flying in a sun-synchronous orbit it crosses the equator at 10.30 in the morning (hence the ofÆcial name EOS AM-1) when cloud cover over land is at a minimum. Terra is carrying Æve instruments (1) The Moderate Resolution Imaging Spectroradiometer (MODIS) an imaging spectrometer that observes in 36 different wavelengths. (2) The Advanced Spaceborne Thermal Emission and ReØection Radiometer (ASTER) is a collaboration between the US and Japan. With its 14 spectral bands extremely high spatial resolution and 15 m long-track stereo capability ASTER is Terra's zoom lens.(3) The Multi-angle Imaging Spectroradiometer (MISR) is a new type of instrument that has never been Øown in space before. MISR has nine digital cameras each looking in a different direction. An area can be imaged at all nine angles in only seven minutes. It has a track 400 km wide and can see objects as small as 275 m. (4) The Measurements of Pollution in the Troposphere (MOPITT) is a new type of instrument designed by an international team led by Canada. It is the Ærst instrument to make simultaneous measurements of carbon monoxide and methane. (5) The Clouds and the Earth's Radiant Energy System (CERES) is a scanning radiometer with twice the resolution of previous designs.Terra carries two identical instruments one operating in a cross-track scan mode and the other in a biaxial scan mode. Data from all Terra's instruments are fed back to the EOS Operations Centre at the Goddard Space Center Greenbelt MD where they are processed and made available as science data products for users worldwide. Data is available at NASA's Distributed Active Archive Centers within 90±120 days. Terra will be followed by a series of satellites to be launched by NASA over the period to 2012. The Ærst of these is a sister satellite called Aqua (EOS PM-1) to be launched later this year which will Øy in a sun-synchronous polar orbit ascending northward across the equator in the afternoon. These will be followed in four years time by EOS AM-2 and EOS PM-2.This journal is # The Royal Society of Chemistry 2000 assessment and mapping; and detecting seasonal climatic changes and changes in ecosystems. They offer an accurate and cost-effective means of gathering environmental data. Present remote sensing devices are still pretty primitive however the imaging equivalent of the box brownie. A new generation of more sensitive remote sensing instruments combined with dramatic increases in computing power will allow data to be obtained at much greater detail and resolution. Environmental analysts have a powerful new tool at their disposal. Environmental monitoring applications of satellite RS are numerous and include ozone and stratospheric chemistry; land ecosystems and hydrology; and the study of polar and glacier regions (``cryospheric systems'').This article focuses on three areas with a strong analytical component aerosols and tropospheric chemistry; clouds and climate change; and ocean remote sensing. Focus Next generation satellites Over the next 10±15 years environmental remote sensing applications will be dominated by two missions the NASA-sponsored Earth Observing System (EOS) and the European Envisat. NASA began planning the EOS in the early 1980s and it is now part of the US Global Change Research Program.2 EOS consists of three components a space-based observing system; a scientiÆc research programme; and a Data Information System (EOSDIS). EOS will comprise a series of satellites to be launched by NASA and its international partners through to the year 2012 the Ærst of which was the Terra satellite launched last December (see Box 1).EOS's instruments are designed to give complementary measurements of the Earth's processes and will be operational for at least the next 18 years. EOSDIS processes and stores incoming data and makes them 41N J. Environ. Monit. 2000 2 Focus available to the research community and other users as data products. The European Space Agency will launch Envisat-1 an advanced polarorbiting Earth observation satellite in June 2001. Continuing the data measurements begun under the ERS-1 and ERS-2 satellites Envisat will support Earth science research and allow monitoring of environmental and climatic changes (see Box 2).The ERS-1 mission was ended on 10th March this year by a failure in an on-board control system after nine years of service over three times its planned lifetime.3 ERS-2 launched in 1995 took over the operational services of ERS-1 in 1996 and remains in excellent condition. Both missions feature an advanced imaging spectrometer as their Øagship instrument. These are the Moderate Resolution Imaging Spectrometer (MODIS) on EOS and the MEdium Resolution Imaging Spectrometer (MERIS) on Envisat. Imaging spectrometers are Øexible instruments with a wide range of potential remote sensing applications including measurements of ocean colour cloud height water vapour total column and aerosol loads over land and water.4 Aerosols and tropospheric chemistry Satellite remote sensing has been applied to the study of the stratosphere (upper atmosphere) for more than 30 years.Our understanding of the troposphere (lower atmosphere) is much less evolved Box 2 Envisat Envisat is the next generation of European earth observation satellite due to be launched in June 2001. The Envisat-1 satellite will carry ten instruments of which the most signiÆcant for environmental analysis are (1) Advanced Along Track Scanning Radiometer (AATSR) will continue the precise sea surface temperature (SST) measurements begun under the ERS-1/2 missions. This will produce a unique 10 year near-continuous dataset at the levels of accuracy required for climate research (0.3 K or better).(2) Microwave Radiometer (MWR) has evolved from instruments Øown previously on ERS-1/2. It will be used to measure the integrated atmospheric water vapour column and cloud liquid water content as correct terms for the radar altimeter. Data will also be used in surface energy budget investigations. (3) The Michelson Interferometer for Passive Atmospheric Sounding (MIPAS) is a Fourier transform spectrometer for measuring high-resolution gaseous emission spectra in the upper atmosphere. It operates in the near to mid-infrared where many of the atmospheric trace gases have important emission features. (4) Medium Resolution Imaging Spectrometer (MERIS) is an imaging spectrometer that measures the solar radiation reØected by the Earth.It has a ground spatial resolution of 300 m in 15 spectral bands and is programmable in width and position in the visible and near infrared. It will be used primarily in ocean colour measurements. (5) The Scanning Imaging Absorption Spectrometer for Atmospheric Chartography (SCIAMACHY) will perform global measurements of trace gases in the troposphere and stratosphere. It will continue and enhance the measurements made by the Global Ozone Monitoring Experiment (GOME) which has been Øying on ERS-2 since 1995. Adapted from the Envisat website at http://envisat.estec.esa.nl 42N J. Environ. Monit. 2000 2 however partly because the troposphere is so complex. Phenomena such as storms forest Æres lightning and manmade pollution emissions all contribute to this complexity.There are currently few data sets of tropospheric chemical components an important requirement in understanding the contribution of tropospheric chemical processes in air quality and climate change. Ground-based datasets are of little use for modelling studies because of the troposphere's great variability. Only space-based measurements can provide comprehensive global time series for troposphere components. Over the next Æve years data from EOS's MOPITT and Envisat's MERIS combined with concerted modelling efforts will help create the Ærst global and long-term picture of the composition of the lower atmosphere. The Canadian MOPITT instrument is the Ærst to simultaneously monitor carbon monoxide and methane two important tropospheric gases.5 MOPITT aims to determine the total amount of carbon monoxide and methane within a vertical column and for carbon monoxide measurements at different altitudes as well.Infrared radiation originating from the Earth's surface is measured and the components radiated from carbon monoxide and methane are isolated using gas correlation spectroscopy. Since methane has an atmospheric concentration of around 1.7 parts per million and carbon monoxide only around 1 part in 10 million the instruments need to be extremely accurate. MOPITT collects data from pixels about 22 km square small enough to detect emissions from individual cities. With a 600 km wide track it can cover the whole of the Earth's surface every 5 days.Since the measurements are in the infrared range data can be collected during the night as well as during daylight. The European MERIS when it is launched next year will also have the capacity to evaluate tropospheric aerosol properties including optical thickness and type6 Satellite data will provide atmospheric chemists and other scientists with highly detailed maps of tropospheric composition. Comparisons of these data over weekly monthly and annual timescales are expected to provide signiÆcant contributions to our understanding of weather seasons and long-term climatic change. Early results from MOPITT have shown signiÆcant variations in carbon monoxide within the Indian sub-continent for example with the Himalayas having low values and lower altitude regions to the south having high values that stream out across the Bay of Bengal.7 Clouds and climate change Radiation and clouds strongly inØuence our weather and climate.4 The effects are so signiÆcant that clouds are second only to greenhouse gases in terms of their effect on long-term climate change.The surface temperature of the Earth depends not only on absorbed solar radiation but also on the rate at which energy is reradiated back into space. The rate of reradiation is controlled by both the amount and the vertical distribution of clouds aerosol particles and greenhouse gases of which the most important is water vapour. Low thick clouds reØect incoming solar radiation back into space causing cooling.High clouds trap outgoing infrared radiation and produce global warming. Since the mid-1980s it has become apparent that overall clouds have a net cooling effect on the Earth. But we have yet to explain the different effects observed in different parts of the world. The Clouds and the Earth's Radiant Energy System (CERES) experiment is one of the highest priority scientiÆc satellite instruments developed for EOS.8 CERES is investigating both solar-reØect and Earth-emitted radiation from the top of the atmosphere to the Earth's surface. Cloud properties are determined using simultaneous measurements by other EOS instruments such as MODIS. For climate change analysis CERES will build on observations begun under previous missions such as the Earth Radiation Budget Experiment (ERBE) of the 1980s.Measurements of radiative Øuxes at the top of the atmosphere (TOA) are analysed using the same algorithms as for ERBE data. CERES is expected to double the accuracy of radiative Øuxes at TOA and the Earth's surface. It will also provide the Ærst longterm global estimates of the radiative Øuxes within the Earth's atmosphere. The consistency of cloud property data and radiative Øux estimates should also be improved. Initial data from Terra and a similar instrument aboard the Tropical Rainfall Measuring Mission (TRMM) launched in 1997 suggest that CERES instruments are a substantial improvement over those of ERBE.9 Similar information on cloud properties (volume height optical thickness albedo) and on aerosols will be available from MERIS.AATSR another Envisat instrument will provide information on cloud coverage water/ice discrimination and particle size distribution while the MWR will produce total column measurements of water vapour and liquid water (see Box 2). Over the next Æve years satellitebased remote sensing should give us a much better understanding of the radiative and physical properties of clouds which in turn should provide a clearer picture of their role in climate change. Global observation will generate new data for improving seasonal-to-interannual climate forecasts including the cloud and radiative aspects of periodic events such as El Ninƒo.Global data on the surface radiation budget can also be used to evaluate the radiative effects and climatic impact of natural events such as volcanic eruptions and major Øoods and droughts. And by providing better estimates of radiative Øuxes within the atmosphere and the earth's surface remote sensing will help us understand the trends and patterns of changes in regional land cover biodiversity and agricultural production. How blue is the ocean? Major uncertainties remain about the amount of carbon stored in the ocean and the biosphere and about the Øuxes between these reservoirs and the atmosphere.4 In particular there is an important need for better information on the spatial distribution of biological activity in the upper ocean and its temporal variability especially in the case of oceanic phytoplankton biomass which has an important role in Æxing CO2 through photosynthesis.On a global scale marine phytoplankton consume Æfty thousand million tonnes of carbon per year–enough to be a signiÆcant factor in climate models. While phytoplankton distribution cannot be seen directly from space it can be calculated from concentrations of chlorophyll-a which can be measured in the visible part of the spectrum. Hence satellite measurements of ocean-colour data have emerged as an important tool for studying the ocean carbon cycle and the role of the ocean in climate change. Ocean colour data also has applications in meteorology and the scientiÆc analysis and measurement of the coastal zone including Æsheries management monitoring of algal blooms and coastal pollution.The Ærst observations of ocean colour from space were carried out by the experimental US Coastal Zone Colour Scanner (CZCS) from 1978 to 1986 aboard NASA's Nimbus-7 satellite. This instrument provided a wealth of new information about the distribution and seasonal variability of biological activity. After a considerable gap due to lack of an operational instrument ocean colour Focus data is now again available from the Japanese OCTS and the French POLDER instruments both launched in 1996 and from NASA's SeaWiFS satellite launched in 1997.This instrument is now providing complete global coverage of the oceans every two days. In addition to MERIS and MODIS other key missions for ocean colour data over the period 2000±2005 will be GLI a mission being prepared by the Japanese Space Agency; and the French POLDER-2. MERIS will provide data with improved spectral and spatial performance. This results from the use of several near infrared channels to perform atmospheric corrections and several narrow visible channels to compute radiance values. By including advanced aperture radar radar altimeter ocean colour and ocean temperature measurements together on the same platform Envisat offers exciting opportunities for synergistic measurements. MERIS's measurements of ocean colour will be combined with those of AATSR on sea surface temperature and ASAR (Advanced Synthetic Aperture Radar) on sea surface roughness.Ocean observations require global spatial coverage at a resolution of 4±8 km a revisit frequency (temporal resolution) of 3±5 days and a minimum of three visible and two infrared channels. The temporal resolution of these instruments is typically 15% per day and integration of data from several instruments on different satellites is needed to achieve the resolutions required. With three instruments around 60% of the ocean surface can be observed every four days. Inland water bodies and marine coastal zones have characteristics at the spatial scale ranging from several tens to several hundreds of metres.10 This requires much higher resolution than present ocean colour instruments such as SeaWiFS.As well as the EOS's MODIS major advances in coastal zone imaging are expected from MERIS and GLI. Seeing through the data As the next generation of earth observation satellites comes on-stream attention is now shifting from collecting the data to how to use it. EOS Envisat and similar missions will generate huge amounts of data of the order of terabytes per day. Scientists are keen to 43N J. Environ. Monit. 2000 2 Focus ensure they utilise this data effectively so that they don't miss any pearls of wisdom enclosed within. Three trends are apparent. Firstly researchers are looking to increase coordination between missions.Although satellites such as EOS and Envisat compete to some extent especially in the supply of commercial data products it is in everyone's interests to ensure that the datasets are as comprehensive as possible. International forums are now emerging to coordinate data collection efforts so as to ensure satellite resources are used effectively. In ocean colour monitoring for example the International Ocean Colour Coordination Group (IOCCG) has taken on this role.10 Secondly with an increasing array of instruments available researchers are focusing on how to integrate and interpret data from multiple sensors using so-called data fusion techniques. NASA scientists demonstrated the approach at their Ærst showing of the Terra data in April.Combining images from MODIS MOPITT CERES and other Terra instruments they clearly 44N J. Environ. Monit. 2000 2 showed how concentrations of gases and aerosols over India are clearly related to topography and population and alter the way sunlight is reØected over the region.7 Thirdly the increasingly large and detailed data sets demand continuing investment in analysis and product development. Algorithms are essential to the accurate and timely interpretation of the data collected. In some cases backward compatibility with previous data sets is also a key consideration as for example in the ERBE/CERES analyses. As satellite coverage and resolution increases and algorithms become more sophisticated more specialised data products are being produced focusing for example on particular parameters or geographical locations (Tables 1 and 2). References and Notes 1 Remote sensing is used here to mean satellite remote sensing. Earth-based remote sensing techniques are not covered. 2 Moderate Resolution Imaging Spectroradiometer NASA 1998. See also MODIS website http:// modis.gsfc.nasa.gov 3 European Space Agency Press Release No 18-2000 13 March 2000. http:// earth.esa.int 4 EOS Science Plan Michael D. King Reynold Greenstone eds NASA Goddard Space Flight Centre 1999. Available at http://eospso.gsfc.nasa.gov 5 Measurement of Pollution in the Troposphere Canadian Space Agency 1998. See also MOPITT website http:// www.eos.ucar.edu/mopitt/home.html 6 Characterisation of aerosols over land with space sensors Centre for Earth Observation 1999. Available at http:// www.ceo.org 7 NASA News Release First Images Press Conference 19 April 2000 Available at http://terra.nasa.gov/Events/FirstImages/ 8 Clouds and the Earth's Radiant Energy System NASA 1998. 9 For information on TERRA data products see http://eosweb.larc.nasa.gov 10 Characterisation of inland and coastal waters with space sensors Centre for Earth Observation 1999. Available at http://www.ceo.org Mike Sharpe