BirdCast supports real action towards conservation with science, spearheading the still emerging scientific discipline of “aeroecology”, the study of ecological activities in the air. For example, we designed complex models – based on over 20 years of radar and weather data, processed and analyzed with machine learning tools, and published in a peer-reviewed journal – to forecast bird migration across the US. The ability to forecast migration two weeks in advance has tremendous potential for bird conservation – from informing siting and operation of wind turbines to addressing impacts of light pollution to providing information to mitigate bird collisions with aircraft and support biosecurity decision-making.