These two photos acquired by Copernicus Sentinel-2 highlight how the mission may assist distinguish between clouds and snow.
The Copernicus Sentinel-2 satellites each carry a high-resolution multispectral imager to provide optical photos in 13 spectral bands, from the seen to the shortwave-infrared space of the electromagnetic spectrum. These two photos, from August 2024, seize the an identical house over the Australian Alps in southeast Australia, nevertheless have been produced using the instrument’s fully completely different spectral channels.
On the left, the true-colour image reveals the realm as a result of the human eye would see it. As we’re capable of see, the realm is roofed with clouds. The prolonged, straight strains are contrails from aircraft, most definitely travelling alongside the busy route between Canberra and Melbourne.
The image choices fairly a number of snow-covered and cloud-covered areas, nevertheless how can we distinguish them? Every appear white, as they every mirror seen light. That’s the place fully completely different band combos may assist.
The false-colour image on the suitable consists of the shortwave-infrared bands, which highlight snow and clouds in one other approach. Snow and ice appear in electrical blue, and water vapour, or clouds, may very well be seen in white.
By evaluating the two photos, it is now easy to determine snow in blue on the suitable and thicker clouds in white within the route of the centre of the image. Thinner layers of clouds and contrails nevertheless are barely seen.
The responsibility of separating clouds from snow may also be a important classification concern when using Artificial Intelligence (AI) to course of satellite tv for pc television for computer photos – a self-discipline that is shortly rising. By analysing multispectral photos, the latest AI algorithms are learning to inform aside between these two parts.
Satellite tv for pc television for computer missions similar to Copernicus Sentinel-2 that will current observations in plenty of spectral bands are key to this job, providing a wealth of data in quite a few views of our planet.
