Computer Vision of Cloud

Data flows from satellites like water from a firehose. We receive about 10 Megabytes per second: about one terabyte per day. Computer vision algorithms can help to transform this data deluge into descriptive, spatially organized information. Such information could help to increase the reliability of weather surveillance, increase the accuracy and timeliness of short-range weather predictions, and possibly facilitate the assimilation of data for numerical weather prediction.