The Earth's land surface is a critical component of the Earth system as it carries over 99% of the biosphere. It is the location of most human activity and it is therefore on land that the human impact on the Earth is most visible. Within the biosphere, vegetation is critical as it supports the bulk of human and animal life and largely controls the exchanges of water and carbon between the land and the atmosphere.
Observations of the land surface by ENVISAT allow the characterisation and measurement of vegetation parameters, surface water and soil moisture, surface temperature, elevation and topography. Global scale measurements (1 km resolution) provide critical data sets for improved climate models, in particular estimates of albedo, vegetation productivity and land surface fluxes.
ENVISAT also provides managers of local natural resources with a capability to monitor their land with detailed (selective) observations on a monthly basis. In particular, ASAR provides 30 m spatial resolution multi-look images for monitoring economically important land units, such as agricultural fields and forest compartments. Natural resources can also be monitored at global and regional scales every few days using the low-resolution imaging of MERIS, AATSR and the ASAR Global Mode.
The relatively high frequency of global coverage provided by ENVISAT is also of great value for hazard monitoring, in which locally infrequent events such as earthquakes, volcanic eruptions, floods and fires, require intensive observation over short periods. The beam steering mode of ASAR (in conjunction with its independence from cloud and illumination conditions) also permits (at least) 3-day repeat observation of certain localised events at high spatial resolution. Although locally rare, certain natural hazards are frequent events on a global scale, thus they can have substantial effects on climate, especially large vegetation fires and volcanic dust clouds. Hazard monitoring is therefore an important component of the ENVISAT mission.
Global Land Cover
A major scientific uncertainty in global change research is the cycling of carbon in the Earth system. It is well known that CO2 contributes to the greenhouse effect and that over the last few centuries increased human activity, especially the burning of fossil fuels and deforestation, have resulted in an increase in the release of CO2 into the upper atmosphere. Much of the estimated anthropogenic CO2 emission cannot currently be accounted for, indeed there is an order of magnitude uncertainty in the global carbon budget.
Critical to this carbon accounting activity is global vegetation monitoring. Figures below show a global land cover product, and a forest map of S.E Asia, both derived from 1 km AVHRR data, The narrow bands of MERIS make it possible to derive more accurate global maps and more effective vegetation indices than have previously been available. From physically based vegetation indices, it is then possible to retrieve key variables in modelling plant productivity (and thus carbon sequestration), surface-atmosphere gas exchanges and energy transfers at the land surface.
Figure 1.12 - Global MERIS land cover map.
Figure 1.13 - Forest map of Southeast Asia (Acknowledgment: JRC/ESA TREES Project.).
For practical reasons (e.g., obtaining sufficient cloud-free coverage on a seasonal basis), global vegetation monitoring is based on low resolution (1 km) data. However, vegetation products (such as land cover, leaf-area index and biomass) at this resolution cannot be validated directly and this is usually done by scaling up data collected at higher resolution on a sample site basis. The availability of contemporary data sets at resolutions of 1000 m from AATSR, 300 m from MERIS and 30 m from ASAR, is thus of key importance in producing and validating global vegetation products.
The control of subsidies at the field level using satellite remote sensing has become an operational activity in Europe. Inventory and estimation of agricultural yields at a national and international level has not been so widely used, but is becoming more operational, particularly in developing countries. ASAR provides important data for this, with supporting products also coming from MERIS and AATSR.
The ERS programme has demonstrated the ability of satellite radars, independent of weather conditions, to identify crops and monitor seasonal land cover changes. Multi-temporal techniques are used, which involve the collection and analysis of SAR data on a series of different dates over the period of interest.
figure 1.13 below shows a sequence of 9 ERS SAR images (each 3.75 km x 3.75 km) taken over the crop growing season (January to November 1993) in Flevoland, The Netherlands. Figure 1.14 shows the corresponding backscatter temporal profiles for the three winter wheat fields highlighted. Research carried out with ERS data has shown that many crop types have distinctive temporal profiles which can be used successfully for crop classification purposes. ERS data are now being used operationally within major European programmes concerned with agricultural statistics (MARS STAT) and the control of agricultural subsidies (MARS CAP). Within MARS STAT the use of ERS data has improved the estimation of crop area early in the crop growing season. ERS data are used as a substitute for optical data in the MARS CAP control activity when cloudy conditions are encountered at key times during the crop growing season.
Figure 1.14 - Time Series.
Figure 1.15 - Temporal Backscatter Profiles.
Time series of ERS-1 images covering (a) the crop growing season, and (b) temporal backscatter profiles for the 3 winter wheat fields highlighted, Flevoland, The Netherlands. (Acknowledgment: M. Borgeaud, ESTEC.)
Mapping the area of crops, for the policing of subsidies and crop area inventories, is expected to continue as a primary application to be supported by ENVISAT. However, yield estimation techniques also are improved through the availability of ENVISAT data. Regional yield estimation has been previously accomplished by exploiting the temporal curve of vegetation response obtained from satellite borne instruments such as AVHRR which have low spatial resolution but frequent revisit capability. This information is compared with previous years and combined with meteorological data and crop growth modelling to predict year-on-year yield variations.
MERIS greatly improves the quality of the spectral information compared to AVHRR, as its spectral bands are narrower and less sensitive to atmospheric effects. MERIS calibration and atmospheric correction is also more accurate than AVHRR, and the spatial resolution is improved while still providing regional revisit every 3 days. Although the VIS/NIR bands on AATSR are broad, the superior atmospheric correction capability and multi-angle view assist in improving estimates of bidirectional reflectance distribution functions for crop growth modelling. ASAR could also contribute to improved crop yield estimates and the identification of stress and disease in crops, particularly in very cloudy areas, but considerable research still needs to be undertaken to prove the potential of ASAR data in these areas.
Hydrology in particular benefits from ENVISAT, as detailed spatial and temporal information on a wide range of land surface parameters is required in order to run physically based models and management scenarios. Major variables that can be derived from ENVISAT observations include land surface temperature, vegetation state, soil moisture, surface roughness, and terrain. Much hydrological modelling is based on gridded data at around 1 km. Because of its narrower bands and improved radiometry, MERIS may be better suited to providing vegetation parameters for hydrology than other instruments such as AVHRR. Some hydrological applications require information on snow cover distribution and snow-water equivalent. Research is required on how to derive these variables from ASAR data at high incidence angles, particularly in mountainous terrain.
Forestry information is important as an aid to formulating land use, forest and environment policy, for long-term regional, national and local planning, and for the assessment and monitoring of natural resources, ecosystems and the environment. The qualitative information requirements include vegetation and forest type, species composition, vigour and health, as well as site conditions. The quantitative requirements include area, volume, age and density of forest stands as well as growth forecasts for sustained production. For large parts of the world, and specifically in developing countries, forest maps and statistics are outdated, unreliable, or sometimes nonexistent. For regional inventories, available national maps and statistics are difficult to compare.
Global and regional forest inventory for change detection, particularly of tropical forests is an important aspect of global vegetation mapping to which ASAR, AATSR and MERIS provide improved capabilities.