These two pictures acquired by Copernicus Sentinel-2 spotlight how the mission can assist distinguish between clouds and snow.
The Copernicus Sentinel-2 satellites every carry a high-resolution multispectral imager to offer optical pictures in 13 spectral bands, from the seen to the shortwave-infrared area of the electromagnetic spectrum. These two pictures, from August 2024, seize the identical space over the Australian Alps in southeast Australia, however have been produced utilizing the instrument’s completely different spectral channels.
On the left, the true-colour picture reveals the realm because the human eye would see it. As we will see, the realm is roofed with clouds. The lengthy, straight traces are contrails from plane, almost definitely travelling alongside the busy route between Canberra and Melbourne.
The picture options quite a few snow-covered and cloud-covered areas, however how can we distinguish them? Each seem white, as they each replicate seen gentle. That is the place completely different band combos can assist.
The false-colour picture on the best consists of the shortwave-infrared bands, which spotlight snow and clouds in another way. Snow and ice seem in electrical blue, and water vapour, or clouds, will be seen in white.
By evaluating the 2 pictures, it’s now straightforward to establish snow in blue on the best and thicker clouds in white in the direction of the centre of the picture. Thinner layers of clouds and contrails then again are barely seen.
The duty of separating clouds from snow can also be a vital classification challenge when utilizing Synthetic Intelligence (AI) to course of satellite tv for pc pictures – a discipline that’s quickly rising. By analysing multispectral pictures, the latest AI algorithms are studying to tell apart between these two parts.
Satellite tv for pc missions corresponding to Copernicus Sentinel-2 that may present observations in a number of spectral bands are key to this activity, offering a wealth of data in numerous views of our planet.
