By detecting data patterns, improving crisis prediction and disaster control

Two researchers say some applied math can help us detect earlier similar patterns in mega-disasters such as earthquakes, homicides and economic recessions.

Feeling unprepared for the next major disaster? ( It's likely you are .)

According to a new theory, disasters and crises -- as disparate as earthquakes, murder rates in megacities, magnetic storms and economic performance -- all share a precursory development pattern, and have indicators that can be tracked.

According to a theoretical framework published in the journal CHAOS, which is published by the American Institute of Physics, detecting this pattern of a change of scale can help improve crisis prediction.

The theory is the work of Vladimir Keilis-Borok of the University of California, Los Angeles and Alexander Soloviev of the Russian Academy of Sciences. They argue that these four types of disasters are likely "a manifestation of a certain general feature of complex systems."

The researchers took a mathematical approach to the complex challenge. They say a given system may give off signals that a major change is coming -- a "premonitory pattern." As the pattern shifts, infrequent events begin to occur more frequently.

They write:

These patterns may not trigger extreme events but merely signal the growth of instability, making the system ripe for extreme events. Finding premonitory patterns creates a basis for development of algorithms for prediction of extreme events and is pivotal for fundamental understanding of relevant complex systems. Specifically, the following types of premonitory patterns have been identified formerly: intensity, clustering, range of correlation in space, and change of scaling.

For now, it's just a theory, but if applied, it could serve as the basis for earlier warnings that could help save lives.

Image: Promotional poster for the film 2012. (Columbia Pictures)

This post was originally published on Smartplanet.com