There are many satellites that orbit our Earth. You may not realise, but there is at least one satellite passing over your house every day. These satellites continually view our planet, creating images from the light or heat that it gives out. From these, scientists can derive really useful measurements of the land, oceans and atmosphere that help us to forecast the weather, monitor our climate, and gain knowledge of the planet.

 

Except of course when it's cloudy...

 

Clouds obstruct our view of the Earth’s surface and atmosphere. This happens a lot. After all, two-thirds of the Earth's surface is covered in cloud at any one time. But scientists who use these images need to know if there is cloud present. Otherwise, those clouds will affect the measurements that they are trying to make.

 

There are many sophisticated ways of identifying cloud, of course. We teach our computers to spot when there is a cloud in an image or not. We call this cloud masking. Most of the time the computer gets it right. But not always. That is where we need you. Help us to classify lots of images. In most cases, the human eye is better than a computer! We want to build up a 'human-cloud mask' that can be used to say how accurate the computer-generated mask is and point us to the conditions where the needs to be improved most. We can even use our results to help train brand new algorithms that use machine learning, but the primary goal of the project is to gather a good set of validation data.