Making decisions in the face of a natural disaster like the one that recently struck Haiti is not always easy. A team of mathematicians from Madrid has developeda computer system that could clarify the best strategies to adopt after an earthquake, flood or tsunami.
?TheExpert System for Disaster Diagnosis (SEDD)is a computer tool for predicting and diagnosing natural disasters forassist NGOs in making strategic decisions?, explained to SINC Begoña Vitoriano, professor in the Department of Statistics and Operations Research of the Faculty of Mathematics of the Complutense University of Madrid (UCM) and co-author of a study published this month in the journalKnowledge-Based Systems. The program starts with information on the type of disaster (earthquake, flood, hurricane, tsunami, fire …), quantifiable units (Richter scale for earthquakes, wind speed for hurricanes, etc.) and a measure of the vulnerability of area. To obtain the latter, the most difficult to obtain, scientists use the human development index provided by the UN by country, and modify it to adjust it to the situation of the affected region.
With these data the computer applicationestimates the magnitude of the consequences of the catastrophe “in terms of deaths, injuries, homeless people, others affected and cost”, very useful information for NGOs. The choice of these variables was made from the database of the Center for Research in Disaster Epidemiology of the University of Leuven (Belgium), a WHO collaborator.
To cope with the high imprecision and uncertainty of the data handled in these cases, the researchers work withdiffuse logic, a mathematical tool that operates with intervals of numbers (not with exact figures) to try to quantify “how much or how little” of magnitudes such as the number of injured or affected. ? The decision support system that we proposeit could have been perfectly applied after the recent earthquake inHaiti, since it focuses on assessing the consequences of disasters like this one, where the first information is scarce, unreliable or, in general, of low quality ?, says Juan Tinguaro Rodríguez, also a member of the UCM team.
Aid deal
Another of the computer applications developed by these mathematicians and that the magazine has just publishedJournal of Multi-Criteria Decision Analysis, is a shumanitarian aid distribution system(HADS). The tool is based on the use of a logistic map of the territory, with nodes (localities) and connections (roads and highways). The “graph” includes the demand for aid in some nodes (affected populations), the supply in others (airports, ports or warehouses), the availability and characteristics of the vehicles (type, capacity, speed, cost), as well as data on the connections (distances, road conditions, risk of assault).
