Towards a typology of integrated functional early warning systems

Lack of common measures in evaluating, comparing and contrasting Early Warning Systems (EWSs) is a persistent problem. As such, there is a growing need for an abstract technical definition of EWSs and a framework to classify them, which would help developers set standards. The standards can be the basis for effective EWS deployments. Practitioners and researchers have been using EWSs for detecting and communicating events for risk aversion. Some EWSs focus on the detection and decision aspects, and others focus on the warning and response aspects, without a precise foundation to follow in the design, deployment and evaluation of an end-to-end 'integrated functional EWS'. In an attempt to construct a theoretical framework for EWS classification on the basis of the system's operation, complexity and entropy, this paper sets the foundation by first defining the essential components of an EWS; namely, the sensor, detection, decision, broker and response subsystems. This paper discusses the abstract operational characteristics and the measures associated with the five subsystems, with an attempt to establish an abstract definition of an EWS to be used in future work on EWS classification.