Billions of dollars in damage and decades of research, but quakes still are hard to predict

In less than a week, on opposite sides of the world, earthquakes have killed hundreds and brought billions of dollars in property damage and economic losses.

But despite decades of research and advanced monitoring and modeling technologies, the science of predicting earthquakes remains elusive.

Back-to-back earthquakes this week in Ecuador and Japan have devastated those regions, inflicting heavy loss of life and property.

In Ecuador, President Rafael Correa declared a national state of emergency for six of the country's worst-hit provinces after a 7.78 magnituide quake struck the country's coastal region Saturday. As of Thursday, the death toll had risen to 570, with more than 4,600 injured. More than 1,100 buildings were reported destroyed and more than 800 damaged, according to catastrophe modeling firm AIR Worldwide.

The company said its initial estimate of insured losses was between US$325 million and US$850 million. But those estimates don't include the far larger losses from uninsured property, infrastructure damage and relief efforts to help thousands of displaced households.

Correa on Monday said the rebuilding effort would cost billions of dollars and could inflict a major toll on the nation's oil-dependent economy, which has already suffered from the drop in crude prices

In Japan, the death toll hit 48 Thursday, a week after a series of quakes rocked the Kumamoto area in the country's southwestern Japan, the largest of which hit magnitude 7.3. The economic losses from economic property damage alone is expected to top US$3 billion.

Despite efforts to minimize their impact, some of the deadliest earthquakes on record have occurred within the last two decades.

More than 200,000 earthquakes are recorded around the world every year, according to the European Geosciences Union. But seismic monitoring and data analysis can only offer general estimates of the long-term odds of a quake striking a giving region.

"There is yet no reliable way of predicting in the short term (days to weeks or months) when an earthquake of a given size will occur in a specific location," according to the group's website.

Geologists have been able to identify quake-prone areas and provide longer-term predictions about the odds of a quake striking a specific region. In the U.S. The National Seismic Hazards Mapping project has an online tool that calculates the probability of a large earthquake within 50 miles of a specific location over a certain time period.

So far the most effective technologies in saving lives and property have been applied to building design and construction to try to minimize the damage from quakes when they strike. But those measures can be costly, and in many parts of the world, building codes in quake-prone areas remain lax.