The invention belongs to the field of fire alarm technologies, and discloses a forest fire prediction and forecasting system. The forest fire prediction and forecasting system comprises a control tower, an unmanned aerial vehicle, a control room host and monitoring stations, and is characterized by further comprising a signal composite, an operational amplifier, a high-voltage circuit, a violet-ray tube, a time base, monitored objects, a monitoring point 1, a monitoring point 2 and GPRS. The control room host transmits a control signal to the control tower through the GPRS and controls the unmanned aerial vehicle. The unmanned aerial vehicle arrives at the monitoring stations, and transmits various kinds of information to an information center via the GPRS through analyzing the monitored objects at the monitoring point 1 and the monitoring point 2. In key regions, flame sensor are installed for detection fire, once a fire is detected, the fire extinguishing unmanned aerial vehicle performs fire extinguishment autonomously, and can further monitor forest growth technical indexes such as soil humidity. The forest fire prediction and forecasting system can record on-site personnel condition in real time, and provides powerful evidence for illegal and criminal activities such as illegal felling, illegal felling and arson.