Mathematical Patterns Found to Forecast Spanish Earthquakes

(NASA)

Researchers in Spain have found patterns of tectonic behavior that occur before an earthquake on the Iberian peninsula. The team, from University of Seville and Pablo de Olavide University in Seville used mathematical clustering techniques to forecast medium to large seismic movements when certain conditions coincide.

The research will be published this month by the journal Expert Systems with Applications.

The investigators based their findings on data compiled by Spain’s Instituto Geográfico Nacional on 4,017 earthquakes, registering between 3 and 7 on the Richter scale that occurred on the Iberian Peninsula and in the surrounding waters between 1978 and 2007. They then applied clustering techniques to the data, which allowed them to find similarities between the earthquakes and discover patterns that will help forecast new tremors.

The team concentrated on the two regions with the most data — Alboran Sea and the Western Azores-Gibraltar fault regions — analyzing three attributes: (1) magnitude of the seismic movement, (2) time elapsed since the last earthquake, and (3) the change in a parameter called the b-value that measures the tectonics of the region from one earthquake to another.

The technique allows geologists to make forecasts summarized to two statistical factors …

• Sensitivity: Probability of an earthquake occurring after the patterns detected occur
• Specificity: Probability of an earthquake not occurring when no patterns have occurred

The results reflect a sensitivity of 90 percent and specificity of 82.6 percent for the Alboran Sea region and 79.3 percent sensitivity and 90.4 percent specificity for the Western Azores-Gibraltar fault region. These calculations translate to a high probability of an earthquake in these regions immediately after the patterns discovered occur (high sensitivity) and, on most of these occasions, they only occur after the patterns are discovered (high specificity).

*     *     *