Often system identification is used to attack problems corresponding in those system which exists a perception of type input-output. We brought this methodology to apply it to the concrete case of analyze the risks that are continuously expected due to the climatic variations as consequence of the arrival of phenomena such as "El Niño". In this paper we have associated the Volterra'a master equation to a one interpretation in the territory of probabilities. The resulting Volterra output is therefore seen as a kind of risk probability. For this end we used Google images by which we have focused our attention to the populations located near to rivers that are in permanent risk in summer times. This methodology can be finally seen as a scheme for disaster anticipation. We paid attention to the zones which have been affected by river overflow along the coast of Peru.