Remote Sensing as a Convergent Discipline: A Path Forward

ABSTRACT Since its inception, remote sensing has evolved from informing military and natural resource management decisions to addressing a broad spectrum of information deficits about environmental phenomena in support of myriad sciences, industries, and management priorities. Driven by changing information requirements and the advancement of technology, geographers, physicists, statisticians, and many others developed a shared understanding of the problem of extracting reliable information from the patterns of electromagnetic radiation observed by radiometers and other electromagnetic sensors, a shared understanding that is now manifest in professional societies, academic journals, and specialized training programs around the world. The discipline we call remote sensing is now uniquely positioned to play a principal role in a new convergent project oriented around the vexing problem of informing humanity's response to a rapidly shifting climate and the 6th mass extinction. Remote sensing, artificial intelligence, computer vision, and robotics each make substantial contributions to this problem, but language, standards, and practices vary widely. The proliferation of sensors now deployed in the environment as part of the internet‐of‐things (e.g., modern cars, satellite constellations, smartphones, CCTV, and trail cameras), coupled with rapid advancements in machine learning and the proliferation of communication networks, makes the potential for persistent, robust, and largely autonomous environmental monitoring systems clear, but also points to the pressing need to converge around a shared understanding of the problem.