Satellites for observing the Earth’s surface have been used since the 70’s and ever since advanced our understanding in science. Using satellites allows us to observe and detect changes in the most remote regions of the Earth. The first land cover satellite named Landsat 1 was launched by the United States on 23 July 1972. This mission and many more have continued providing an enormous collection of satellite imagery. According to the United Nations Office for Outer Space Affairs (UNOOSA), almost 600 earth observation satellites are currently in the orbit around our planet– operated by various countries and fulfilling different tasks. This article will give you an introduction into remote sensing, covering the electromagnetic spectrum, active and passive satellite systems, types of resolution that need to be considered and open access satellite depositories.
The Electromagnetic Spectrum
Satellite sensors detect emitted or reflected energy from the Earth’s surface. This energy or electromagnetic radiation is often expressed in frequency (Hz) or wavelength (λ). The following image shows the range of waves including long wavelengths such as microwaves or broadcast radio waves and short wavelengths such as gamma or x-rays. The spectrum also includes the visible spectrum, which we can see with our eyes. This fraction is very small compared to the large range of wavelengths. Some gases in the atmosphere (carbon dioxide, water vapor and ozone) absorb energy at certain wavelengths, which constrains the wavelength ranges that can be used for remote sensing. The regions of the spectrum where electromagnetic energy can travel through the atmosphere are called “atmospheric windows”, and this is where satellite sensors operate.
Electromagnetic Spectrum, Credit: Engineering ToolBox, (2016). [Accessed 30 July 2018]
Active and Passive Satellites
Satellite sensors can work in two different ways: active and passive. Active systems illuminate the area of interest and measure the wavelength that is reflected or backscattered back from the surface. Active systems mostly use the microwave region of the electromagnetic spectrum, they are independent of weather conditions and can also operate during night time. In contrast, passive sensors use the sun as illumination source and measure the energy that is naturally emitted from the Earth’s surface. Most passive systems require a clear sky and daylight to operate. During night-time, thermal energy can be recorded when the signal is strong enough.
Different surfaces of the Earth, for example bare soil, vegetation or water bodies have a “spectral signature”. Energy of the sun reaches the surface and is then absorbed, transmitted or reflected by the material in certain regions of the spectrum. Satellite sensors are tuned to detect the reflected energy. The near infrared region is for example very suitable for detecting how healthy or unhealthy vegetation is or the ultraviolet (UV) region allows to distinguish rocks and minerals.
All satellite sensors have their strengths and weaknesses and are useful for different problems. We distinguish between spatial, temporal and spectral and radiometric resolution of sensors. (1) Spatial resolution refers to the smallest possible detail that can be captured by the sensor. A high spatial resolution is required when for example detecting separate houses. It is commonly expressed in metres – a spatial resolution of 30 m will result in pixel sizes, which are 30 x 30 m each.
Illustration of high, medium and low spatial resolution. Credit: GIS Geography
(2) Temporal resolution describes how often the sensor will revisit the same object and is often reported in days. For example, the Landsat satellite will pass by the location every 16 days in its orbit, while the SPOT satellite can revisit a location every 1 to 4 days. A high temporal resolution is for example needed when mapping the impacts of an extreme weather event, for example how flooding impacts a region over several days. Monitoring changes in vegetation over the summer require a lower temporal resolution. (3) Radiometric resolution refers to the ability of the sensor to detect grey-scale values. It is reported in bit. The higher the bit value of an image, the more variation in reflection is captured, but also the larger the image. Finally, (4) spectral resolution means the width within the electromagnetic spectrum in which a sensor records information. High spectral resolution describes a narrow wavelength range. For example, multi-spectral satellite systems such as Landsat detect several discrete bands (3 to 10) at different wavelength intervals. Hyperspectral instruments can consist of hundreds or thousands of narrow bands. This high spectral resolution is useful when a fine discrimination for example between minerals or vegetation species is needed.
Every satellite image consists of a number of bands or channels each representing different wavelength intervals. Satellite images can be displayed as a “true colour composite”, which is exactly how we would see the Earth with our eyes. Similar to raw images from a DSLR, the channels red, green and blue of the visible spectrum are used for that. To visualize multi-spectral images, image processing software allows to display other, for the human invisible wavelengths as red, green or blue. This is called a “false colour composite”.
Following snapshot of Iran shows a true colour composite on the left and a false colour composite on the right. In the false colour composite, red represents shortwave infrared, green represents the near infrared and blue represents green. Take a look at the image and compare both. The shortwave infrared helps to distinguish wet from dry areas. You may notice the red colour along the dark ridges on the western edge of the image. This represents water flowing down the ridges, which would be invisible in the natural colour composition. With this false colour composition, it is also much easier to distinguish between different rock layers and formations. The green shading throughout the image is not representing vegetation, but is due to the spectral signature of a particular rock. To allow an informed interpretation of satellite imagery, detailed knowledge of the spectral signature of the Earth’s surface is required. Scientists also collect “ground truthing” data in the field with handheld spectrometers to identify spectral signatures specific to a material (for example a mineral) to support their interpretations.
True and false colour composites of the Yazd and South Khorosan provinces of Iran based on Landsat-8 imagery. Credit: Visible Earth, NASA
Sometimes, satellite images can be extremely beautiful or even seem artistic like following image capturing the Lena River delta in northern Russia by the Landsat-7 satellite. The false colour composite is composed of near infrared (red), shortwave infrared (green) and red (blue). If you want to learn more about false colour band combinations and try them out yourself, you can look here.
Lena River Delta, Russia acquired by Landsat-7. Credit: NASA
A lot of satellite imagery is freely available to the public. The USGS’s database Earth Explorer for example provides access to the 40 years archive of the Landsat mission, digital elevation models and many more products of NASA’s missions. The European Space Agency (ESA) also launched their Copernicus Open Access Hub, providing access to imagery from the Sentinel mission.