Optical Imagery is Eclipsed!

Solar eclipse across the USA captured by Suomi NPP VIIRS satellite on 21st August. Image courtesy of NASA/ NASA’s Earth Observatory.

Last week’s eclipse gave an excellent demonstration of the sun’s role in optical remote sensing. The image to the left was acquired on the 21st August by the Visible Infrared Imaging Radiometer Suite (VIIRS) aboard the NOAA/NASA Suomi NPP satellite, and the moon’s shadow can be clearly seen in the centre of the image.

Optical remote sensing images are the type most familiar to people as they use the visible spectrum and essentially show the world in a similar way to how the human eye sees it. The system works by a sensor aboard the satellite detecting sunlight reflected off the land or water – this process of light being scattered back towards the sensor by an object is known as reflectance.

Optical instruments collect data across a variety of spectral wavebands including those beyond human vision. However, the most common form of optical image is what is known as a pseudo true-colour composite which combines the red, green and blue wavelengths to produce an image which effectively matches human vision; i.e., in these images vegetation tends to be green, water blue and buildings grey. These are also referred to as RGB images.

These images are often enhanced by adjustments to the colour pallets of each of the individual wavelengths that allow the colours to stand out more, so the vegetation is greener and the ocean bluer than in the original data captured by the satellite. The VIIRS image above is an enhanced pseudo true-colour composite and the difference between the land and the ocean is clearly visible as are the white clouds.

As we noted above, optical remote sensing works by taking the sunlight reflected from the land and water. Therefore during the eclipse the moon’s shadow means no sunlight reaches the Earth beneath, causing the circle of no reflectance (black) in the centre of the USA. This is also the reason why no optical imagery is produced at night.

This also explains why the nemesis of optical imagery is clouds! In cloudy conditions, the sunlight is reflected back to the sensor by the clouds and does not reach the land or water. In this case the satellite images simply show swirls of white!

Mosaic composite image of solar eclipse over the USA on the 21st August 2017 acquired by MODIS. .Image courtesy of NASA Earth Observatory images by Joshua Stevens and Jesse Allen, using MODIS data from the Land Atmosphere Near real-time Capability for EOS (LANCE) and EOSDIS/Rapid Response

A second eclipse image was produced from the Moderate Resolution Imaging Spectroradiometer (MODIS) sensor aboard the Terra satellite. Shown on the left this is a mosaic image from the 21st August, where:

  • The right third of the image shows the eastern United States at about 12:10 p.m. Eastern Time, before the eclipse had begun.
  • The middle part was captured at about 12:50 p.m. Central Time during the eclipse.
  • The left third of the image was collected at about 12:30 p.m. Pacific Time, after the eclipse had ended.

Again, the moon’s shadow is obvious from the black area on the image.

Hopefully, this gives you a bit of an insight into how optical imagery works and why you can’t get optical images at night, under cloudy conditions or during an eclipse!

Night-time Treats

This image of Rio de Janeiro was acquired on the night of July 20, 2012 by the VIIRS instrument aboard the Suomi NPP satellite. Data courtesy of NASA/NASA’s Earth Observatory.

This image of Rio de Janeiro was acquired on the night of July 20, 2012 by the VIIRS instrument aboard the Suomi NPP satellite. Data courtesy of NASA/NASA’s Earth Observatory.

The Opening Ceremony of the Rio Olympics featured a plane taking off from the Maracanã Stadium and treating us to a fantastic night flight over Rio. It was a beautiful sequence to celebrate the famous Brazilian aviator Alberto Santos-Dumont, for us at Pixalytics it led to a conversation about the beauty of night-time satellite imagery!

Currently, the best source of night-time imagery comes from Visible Infrared Imaging Radiometer Suite (VIIRS) which is one of five instruments aboard the Suomi National Polar-orbiting Partnership satellite launched on 28 October 2011. Although, if you look on Twitter you’ll also see a huge number of night-time images taken by astronauts aboard the International Space Station. This data has been used as the basis of the Cities at Night citizen science project whose aim is to create a Google maps style map of the world – as the astronauts are using cameras to take photos of the places that interest them, and there is no georeferencing information, citizens identify the cities pictures.

In contrast VIIRS is an orbiting satellite and so continually collecting calibrated and georeferenced data of the whole globe. In the day VIIRS is collecting optical and temperature data over both the land and ocean, while at night it collects temperature data and the night-time imagery using the 750 m spatial resolution Day/Night Band (DNB). Working through both the night and day, the DNB needs to be calibrated through several orders of magnitude in brightness to accommodate the dramatic contrast between solar reflection and the darkness of night. Its forerunner was the uncalibrated Operational Linescan System (OLS) on the Defense Meteorological Satellite Program (DMSP) satellites, whose primary aim was to study clouds, but when its data was declassified in the 1970s it generated a lot of interest in low light night-time observations.

The DNB VIIRS images, like the one at the top of the blog, show hubs of human activity and the road arteries that connect them, and so are of special interest to the Campaign for the Protection of Rural England who use these types of maps to protect dark skies. It also enables calculations of light pollution to be made, together with indications of the associated carbon emissions. The DNB can pick up many different phenomena. For example, aurorae are visible, as well as gas flares, volcanic activity, the lights of ships, sea ice and climatological monitoring of clouds. It’s even possible to see thunderstorms, although individual lightning flashes are hard to make out in these snapshots, the glow inside clouds caused by them are evident as bright strips with DNB imagery as seen in this image from over Louisiana, USA on 4 April 2012 (Miller et al., 2013).

Another interesting discovery in 2012 was the presence of a faint ‘nightglow’ in the upper atmosphere on moonless night over the Pacific. The DNB team were aiming to collect scenes of complete darkness for calibration purposes, but they found clouds were still clearly visible. This was due to an assortment of photochemical reactions, especially of the molecule fragment hydroxyl, which allows this nightglow to pick up subtle atmospheric phenomena such as gravity waves and the tops of anvil clouds.

Here we’ve gone from an aviation image inspired from 1903 to modern satellites, all via the Rio Olympics. It’s amazing where space can take you!

 

Blog written by Dr Louisa Reynolds and Andrew Lavender from Pixalytics Ltd.