, a team of researchers led by Lucas Kruitwagen, a climate scientist andresearcher at Oxford University, demonstrate another way to keep tabs on the green-energy revolution. Dr Kruitwagen and his colleagues have put together an inventory of almost 69,000 big solar-power stations all over the world—more than four times as many as were previously listed in public databases. This new inventory includes their locations, the date they entered service and a rough estimate of their generating capacity.
One is a growing abundance of cheap, easily available satellite imagery. In the 20th century, reconnaissance satellites were the jealously guarded property of a handful of governments. These days, a cottage industry of Earth-observation firms and agencies sells images on the open market. Dr Kruitwagen’s pictures came from two sets of satellites, Sentinel-2 and, run by the European Space Agency and Airbus respectively.
Computer vision is a hot field. But the specifics of orbital reconnaissance meant that off-the-shelf software was not suitable for the task the researchers had in mind. Machine-learning systems are taught what to do by examining a “training set”, which contains examples of what is being searched for. For common tasks such as facial recognition, pre-built training sets are often available. But Dr Kruitwagen’s team had to build their own.