CloudCatcher: citizen science can be used to help to crowd source a satellite cloud validation dataset
Cox, Caroline ; Polehampton, Edward ; Hall, Vicky
Accurately identifying clouds in satellite images is an important stage in processing such images for scientific use. Cloudy image pixels are labelled using a pixel mask, to allow them to be filtered out and ignored. The accuracy at which 'cloud masking' can be performed needs to be determined and this is often done by expert manual classification. Given the public enthusiasm for weather, we tested whether this classification could be performed reliably by 'citizen scientists'. We found that the public could identify cloud with a 94% accuracy over ocean. Our analysis pointed to future improvements of the CloudCatcher project.</p>
Accès à la notice sur le site du portail documentaire de Météo-France