RTD info logoMagazine on European Research

N 44 - February 2005
  EARTH AND SPACE  -  A picture of precision

Earth observation from space generally generates two kinds of picture: high resolution (from 5 m to 0.50 m) with a narrower field (under 20 km at ground level) and medium resolution (from 40 m to 5 m) with a wide field (at least 100 km). Read on to discover how they work.

The Envisat sensors. © ESA
The Envisat sensors.
© ESA
Passive spectrometry, which employs optical sensors (cameras and scanners), studies the nature of the Earth’s surface on the basis of the reflection of solar light at various wavelengths. The equipment functions in the same way as a digital or television camera fitted with filters that vary depending on the details which are to be highlighted. Blue, green, yellow, red, near infrared and medium infrared are the most frequently used. The hyperspectral imager is a new instrument that permits a fine scanning of hundreds of spectral bands. Its use presents a technological challenge for the processing and interpretation of its observations from space.

Making waves
Active spectrometry using radar (SAR) systems probes the Earth’s surface by emitting radio waves through the cloud layers, night and day, and collecting the echoes, or backscatter, that are sent back. Similarly, active spectrometry using lasers, also known as light detection and ranging (LIDAR) systems, also probes the Earth’s surface but by emitting light waves instead of radio waves. This covers an area of the Earth’s surface that corresponds to the size of the beam and contains the elements likely to produce backscatter. 

Complex software is used to process the measurements taken in order to produce an image. For each of the millions of pixels in a radar image, hundreds of mathematical calculations (algorithms) must be carried out.

The frequency or wavelength at which the radar beam operates is one of the parameters that influence an object’s ‘radar signature’; it is in the region of 1 GHz to 10 GHz (corresponding wavelength to 30 cm to 3 cm), which places it   among the hyperfrequencies or microwaves. The higher the wavelength and the lower the soil humidity, the deeper the radar beam penetrates. Over a forest, for example, a band C radar (5.3 GHz) can ‘see’ the tree canopy whereas an L band (1.3 GHz) will penetrate to ground level.  

A question of altitude
Altimetry, which measures the reflection of radar waves by sea and land, makes it possible to measure height down to the nearest centimetre. This provides uniquely detailed information on the height of waves (from which wind speed can be calculated) and the influence of ocean currents (such as the caprices of El Niño in the Pacific). This helps us to understand better how ocean masses behave, enabling us to construct better forecasts of high tides and storms.

From a height of 1 300 km, the Franco-US Topex-Poseidon and Jason-1 satellites give a reading every ten days on the sea conditions beneath the same points. These measurements are combined with the data provided by the Doris (Doppler Orbitrography and Radiopositioning Integrated by Satellite) instrument and the altimeters on board the SPOT, ERS and Envisat satellites. Altimetric readings of this kind have become a necessity for oceanographers worldwide. 

Combining the signals
Interferometry, which consists of combining two radar signals from the same region at two different moments in time, is a technique practised at the time of successive flights over the same region or a tandem flight by European ERS satellites. If the signals are identical, the wave form of the combined signal will remain the same. But if changes have taken place on the Earth’s surface, the wave forms will show differences or ‘interference fringes’.

By processing these interferences it is possible to identify the least change in topography, providing valuable information on ground behaviour at the time of earthquakes.

France’s Centre National d'Etudes Spatiales (CNES) has developed the Diapason software package that, on the basis of radar satellite measurements (for INSAR applications with ERS-2), can automatically detect changes of just a few millimetres in 1 km² sections of the Earth’s surface. This method of differential interferometry makes it possible to determine the least variations in relief, surface humidity and plant cover.