Earth viewed from above

15 July 2020
3 min read
15 July 2020
3 min read

Remote satellite sensing: what it is and what it is used for

In the scientific field, the images produced by the sensors on board Earth observation satellites make it possible to observe vast territories and investigate beyond simple human vision. Thanks to repeated orbits of the satellite, it is possible to study the evolution of natural phenomena such as the occurrence of unexpected catastrophic events. The branch of science that deals with this is Remote Sensing


The blue planet: one of the first images of the Earth seen from space taken on July 12, 1972.

Remote sensing data collection takes place via special sensors that record information transported by the electromagnetic energy emitted, reflected or diffused by the bodies observed, succeeding in making radiations that the human eye would otherwise not be able to perceive on the contrary “visible”.

Why observe from above

It seems that Socrates (5th century BC) already intuited that, in order to fully understand the world in which we live, man would have to rise above the Earth, to the apex of the atmosphere and beyond.


Felix Tournachon, known as Nadar, one of the first photographers. Credits: Wikipedia

Modern Earth Observation satellites are the modern answer to the philosopher’s suggestions. The images transmitted by satellites allow us to observe our world beyond everyday experience. In fact, images taken from above using Remote Sensing tools show a reality that we recognise, but that coincides only partially with what we usually experience.

Synoptic images

Remote photography offers a view over vast territories: we can follow the path of a river that crosses several countries; observe the differing distribution of forests in different territories; control the spread of fires over an entire continent or monitor the conditions of entire seas.


The image was acquired on 29 May 2001 (CNR-IREA, Milan © NASA).

Multispectral images

Human colour vision allows observation of only a small part of the spectral behaviour of objects. Remote sensing observes the behaviour of the surfaces and various portions of the electromagnetic spectrum, viewed by the sensor bands, thus succeeding in distinguishing the types and states of the surfaces and objects that compose a territory. Sensors therefore allow us to go beyond the “visible” thanks to the possibility of detecting and measuring data not perceivable by the human eye.


Example of Multispectral imaging by the TM Sensor onboard the LANDSAT satellite.

The image taken by the TM sensor shows the confluence of the Ticino and Po rivers near the city of Pavia. Different band combinations make it possible to detect different features of the land.

The first combination (RGB 321) reproduces what the human eye can see, i.e. colours. Band 3 shows red data, band 2 green and band 1 blue. It is worth noting how the colour of the water in the two rivers differs: the Po is lighter because it is rich in sediments.

The second combination is known as Infra Red (IR) false colour. Red channel information is carried by near-infrared, band 4. This combination is used to analyse the vegetation. It can be seen that all cultivated fields and forests are more easily recognisable and distinct from other areas.

The third image again shows a combination in false colours in which infrared bands 4 and 5 are loaded. This latter falls within the mid-infrared spectrum. This combination facilitates recognition of water bodies since water absorbs infrared wavelengths. In addition to the rivers, in black, other specific elements appear in the farmland: they are the flooded rice fields.

Multi-temporal imaging

Finally, remote sensing techniques make it possible to capture the same scene at regular intervals. This feature allows us to follow environmental dynamics over time, as the seasons change or to monitor other factors, such as human intervention.


The deforestation process in Rhodonia (Brazil) from 1975 to 1992.

This sequence of images shows the deforestation process in some areas of Brazil viewed by the Landsat-MSS sensor on 19-06-1975 (a), Landsat -TM on 1-08-1986 (b) and 22-06-1992 (c) (Source: USGS).

by Mirco Boschetti – CNR-IREA Milan section