The work analyzes technical and experimental research in the field of fire monitoring. It proposes a technique of using cyber-physical systems for monitoring natural and technogenic emergencies, providing information and algorithmic support for a cyber-physical system of forest-fire monitoring. It also provides hierarchical structures of components for a cyber-physical system of forest-fire monitoring. The methodology is implemented as development of intelligent telecommunications structures on the base of the MB 7707 neuroprocessor systems that can undergo training in conditions of uncertainty. The work is accomplished within the framework of the Russian Foundation for Basic Research (RFBR) grant 14-07-00261A.