Sensor Web service integration for pandemic disease spread simulation

Abstract

Pandemic diseases, such as avian influenza, can be deadly and disastrous although they do not happen frequently. Simulation is called for to control the spread of such diseases in preparing for the outbreaks. The spread of pandemic diseases can be affected by many factors, including virus vectors (such as migratory birds), climatic changes, and environment en route of virus vector migration. Many datasets and observations need to be assimilated into the simulation systems. The data and observations can be grouped into three types, i.e. satellite remotely sensed observations, in-situ data, and simulated model outputs. Satellite observations provide a timely large-scale coverage of environment changes and weather conditions. In situ data gives detailed ecological observations of birds at different stations and sites for simulation model construction and validation. Global climate models provide predictions of climate/weather changes, especially useful for filling the gaps when actual observations are not available. In this study, the Self-adaptive Earth Predictive System (SEPS), a general framework of Sensor Web service coordination and integration, was used as the basic framework to develop the observation and data assimilation services for the simulation of avian influenza. With the framework, these three types of data and observations are served through three different Web services. Remotely sensed observations are served through either Sensor Observation Service or Web Coverage Service. Climate models and simulation models are wrapped as Web Processing Services. In-situ bird observations are extracted online and populated as Web Feature Services. The integration of these services is managed using the central piece of SEPS, Coordinate and Event Notification Service (CENS). Internally, a Business Process Execution Language (BPEL) engine is used to actually execute the service integration.