History of earth observation

History of Earth observation

Aerial Photographs

During World War I reconnaissance aircraft flew over enemy forces in order to observe troop movements. By using ordinary cameras mounted onto the aeroplanes, the precursors of modern remote sensing systems were developed. Based upon the photographs it was possible to observe the position and strength of enemy forces.

During World War II the technique was further developed. As part of preparations for the Normandy Invasion (D-day), aerial photographs were used to map coastal conditions to identify the most suitable sites on which to land. By measuring waves close to the coast it was possible to determine wavelength and thereby calculate water depth. Furthermore, infrared film was used to identify green vegetation and distinguish it from camouflage nets.

The 1960s witnessed the development of film sensitive to wavelengths that could be used to map features like different vegetation types.

Development of other remote sensing technologies continued apace. Mapping experiments were performed with airborne radar systems. TIROS 1, the first weather satellite, was sent into orbit in 1960. It provided the US Weather Bureau with daily images of cloud formation and represented a milestone in weather forecasting.

The development of non-photographic remote sensing technology progressed rapidly after the first mapping satellite, Landsat 1, was put in orbit in 1972. It was equipped with a new type of sensor known as a multispectral scanner (MSS). With this new technology, data were produced in the form of digital chorological matrices enabling substantial advances in image processing.  

Satellite with multispectral scanner
Satellite with multispectral scanner

Multispectral scanners

Today the scanner is a very important instrument in remote sensing. It is used on land, in aircraft and onboard satellites. The detectors on each scanner are designed to receive radiation in specific channels. The number of channels, their width, and their location in the electromagnetic spectrum vary for each sensor, resulting in different spectral resolution characteristics. This combination of factors determines the uses for which the sensor’s imagery are most suited.

A diagram of a scanner in a rotating satellite is shown on the right. The radiation from the scanned area on the Earth strikes a mirror, from which an optical filter separates the various wavelengths. The filtered radiation strikes various detectors, each of which measure the amount of radiation within their particular sensitivity (channel). The result of this measurement is a number quantifying the amount of radiation in each channel, meaning that the scanner records digital data. For each scanned area a number is delivered for each channel producing a chorological matrix. If all numbers from all channels are considered together they represent the spectral signature of the scanned area.

The mirror receives the radiation from a unit-square area on the surface of the Earth. The spatial resolution depends on the scanned unit-area and on the height of the aircraft or satellite.

The satellite rotates and moves forward in its orbit at the same time. For every rotation a new line is scanned on the Earth. Because the satellite is moving along its track, an aperture mechanism ensures that light is admitted and excluded in a set pattern so that the scan lines are divided into scanning areas. In this way, data is collected for a chorological matrix. The numbers in the matrix, the digital data, are transmitted to Earth-based stations by ordinary radio communication.

In pushbroom scanning, a linear array of detectors is orientated perpendicularly to the direction of movement. Satellites utilising pushbroom scanning do not rotate, and therefore the array of detectors detect the chorological matrix as the satellite moves along its track.

The most advanced military satellites can scan areas of 10 cm by 10 cm or less. The actual spatial resolution is kept secret, but it is sufficient to detect very small details such as individuals, vehicles and small installations.

The weather satellite Meteosat has a resolution of 5 by 5 km. Fewer details are visible, but it is possible to obtain a complete survey of one hemisphere in a single image.

Hubbard glacier in the southeast of Alaska
Hubbard glacier in southeast Alaska

Radar sensors

A radar sensor system emits the radiation that it ultimately records and is therefore classified as an active sensor. Passive sensors, on the other hand, are dependent upon receiving reflected sunlight or thermal infrared emissions. Examples of these passive systems are the multispectral sensors discussed in the above section.

In simple terms, the radar sensor sends pulses of energy down towards the surface of the Earth. A portion of the energy is reflected and returns as an ‘echo’ signal. The strength of the returned ‘echo’ will depend on the roughness and moisture content of the surface and the degree and orientation of sloping in relation to the radar beam. The delay of the ‘echo’ reveals the distance to the reflecting surface.

The emission of radar pulses makes substantial demands on the power supply of the satellite, which, as a consequence, becomes both very expensive and complicated. However, the potential which lies in radar technology is so great that substantial investments are being made for its continued development.

Radar sensors use energy emitted at longer wavelengths which can penetrate clouds and haze effectively, and can therefore acquire imagery at night. This provides a significant advantage over passive satellites that are hampered by clouds and require sunlight to acquire detailed imagery.

Radar sensor systems are used both in aircraft and satellites. Their images can reveal topographical details, and if the same area is sensed from two different angles, the object’s distance from the satellite can be calculated, and consequently, its height above sea level inferred (interferometry). These data can then be used in three-dimensional mapping. Such terrain models are used, for example, in the control system of missiles which can find their own way to targets. The missile control system can compare the landscape over which it is passing with the installed terrain model and can navigate automatically to its target. The data can also be used for a range of other applications, such as assessing the impact of flooding.