9. New York Hurricane

Major hurricanes have made direct hits on the boroughs before, but the interval between them is so long that people forget, and officials fear they might not take evacuation orders seriously. The larger problem: It would take nearly 24 hours to make a proper evacuation of New York City, but hurricanes move more swiftly as they race north, so real warning time could be just a few hours.


8. Asteroid Impact

Scientists can't say when the next devastating asteroid impact will occur. Odds are it won't be for decades or centuries, but an unknown space rock could make a sucker punch any time. Many experts say planning to deal with a continent-wide catastrophe should begin now.

7. Los Angeles Tsunami

An earthquake fault just off Southern California could generate a major quake and a $42 billion tsunami that would strike so fast many coastal residents would not have time to escape. Add to that the unprecedented destruction from the earthquake's shaking, and the situation would be reminiscent of Hurricane Katrina.

6. Supervolcano

It probably won't happen for hundreds or possibly even millions of years, but nobody really knows when Yellowstone will blow again, destroying life for hundreds of miles around and burying half the country in ash up to 3 feet (1 meter) deep.

5. Midwest Earthquake

It has been nearly two centuries since a series of three magnitude-8 quakes shook the then-sparsely populated Midwest, centered near New Madrid, Missouri. Another big one is inevitable. Now the region is heavily populated, yet building codes are generally not up to earthquake snuff. What's more, geology east of the Rockies causes quakes to be felt across a much wider region. Shelves would rattle from Boston to South Carolina. Some homes along the Mississippi would sink into oblivion.

4. Heat Waves

Heat waves kill more U.S. residents than any other natural disaster. As many as 10,000 people have died in past events. As urban areas get hotter, electricity systems are strained and the population ages, the risk grows.

3. East Coast Tsunami

It seems no coast is immune to the threat of tsunami. For the Eastern United States, the likeliest scenario is waves kicked up by an asteroid splashing into the ocean. Astronomers already have their eye on one rock that could hit in the distant future, but the cosmos could hold a surprise, too.

2. Gulf Coast Tsunami

A fault line in the Caribbean has generated deadly tsunamis before. Up to 35 million people could be threatened by one in the not-to-distant future, scientists say.

1. Total Destruction of Earth

Okay, so nobody is spending too much time worrying about what to do if the planet is annihilated, but at least one person has seriously pondered whether and when it could happen. From being sucked into a black hole to being blown up by an antimatter reaction, there are scientifically plausible risks of an event that would render this whole list moot.

The historical record for this zone, which has the longest recorded data about its earthquakes of any major fault in the world, shows that earthquakes occur in clusters of up to five events, with an average time interval of 300 years between quakes, said Chris Goldfinger, a marine geologist at Oregon State University. Goldfinger and other scientists have been studying this subduction zone for many years.

The two most recent quakes on this fault occurred in the year 1700 (a magnitude 9 event) and approximately the year 1500. It has now been 305 years since the last event. So is the Cascadia subduction zone finished for now or on the brink of event number five?

"We know quite a bit about the periodicity of this fault zone and what to expect," he said. "But the key point we don't know is whether the current cluster of earthquake activity is over yet, or does it have another event left in it."

The Cascadia subduction zone occurs where the relatively thin Juan de Fuca plate moves eastward and under the westward-moving North American Plate. When that collision results in a rupture, massive earthquakes occur. The other active subduction zone capable of producing a major earthquake-tsunami sequence is in Alaska, the site of a giant earthquake and subsequent tsunami in 1964.

Scientists say a rupture along the Cascadia fault would cause the sea floor to bounce 20 feet or more, setting off powerful ocean waves relatively close to shore. The first waves could hit coastal communities in 30 minutes or less -- too rapidly for the current warning systems to save lives.

A tsunami along the Atlantic Coast is considered extremely unlikely.

Tsunamis are the result of sudden rises or falls in a section of the earth's crust under or near the ocean, usually caused by earthquakes, volcanic activity or landslides. Earthquakes at subduction zones (rather than at other types of faults such as thrust faults) produce the highest energy tsunamis, especially when they occur in deep water. The seismic activity displaces sea water, creating a rise or fall in the level of the ocean above. This rise or fall in sea level initiates the formation of a tsunami wave. The wave's height increases in shallower water.

Geologists can track earthquakes back in time by radiocarbon dating deposits of sand called turbidites, which come from undersea landslides.

Major studies on the Cascadia fault zone have identified 19 to 21 major earthquake events during the past 10,000 years. During at least 17 of these events, the entire fault zone probably ruptured at once, causing an earthquake around magnitude 9 and major tsunamis, such as those which savaged East Asia last week.

The Asian event happened where the India plate was subducted beneath the Burma microplate. It ruptured, for the first time since 1833, along a 600-mile front just about the same length as the Cascadia Subduction Zone.

The Asian event may provide a shocking demonstration of the geologic future of the Pacific Northwest, Goldfinger said. For hundreds of years, subduction zone plates remain locked in place, releasing little tension. Every few centuries, in a few minutes of violence, forces are released as the upper plate moves seaward, producing a massive tsunami following earthquake shaking.

"In the case of the Cascadia Subduction Zone, you could have an area of ocean floor that's 50 miles wide and 500 to 600 miles long suddenly snap back, causing a huge tsunami," Goldfinger said. "At the same time, we could expect some parts of the upper, or North American, plate to sink one to two meters. These are massive tectonic events. Subduction zones produce the most powerful earthquakes and tsunamis in the world."

The question is not whether, but when the Cascadia Subduction Zone will break again.

"One possibility is that we could be done with this cluster and looking at a period of many hundreds of years before the next earthquake," Goldfinger said. "The other distinct possibility is we could still be in a cluster of events. If that's the case, the average time interval between earthquakes within a cluster is already up. We would be due just about any day."

The Associated Press contributed to this report

With another hurricane season officially underway and lots of activity expected, forecasters are armed with more data and more powerful computer models than ever before. There would be some warning.

Because of the population density, a successful evacuation of vulnerable low-lying areas in and around New York City and metropolitan New Jersey would actually have to start sooner than what is typically ordered in Florida and elsewhere, officials have determined based on studies by the Army Corps of Engineers.
But hurricanes move more quickly and become very difficult to predict when they head north of the Carolinas. In a likely scenario, experts say, there might only be hours of warning.

History reveals that New York and the Northeast have been hit hard before, and with little warning. Scientists say the next major hurricane to strike the city is a question of when, not if.

History forgotten

Late in the summer of 1938, a wave of energy moved off the coast of Africa. Winds were sucked into the atmospheric depression and curved by the Earth's spin into a counterclockwise rotation.

The storm churned unnoticed across the open sea, gathering strength from warm tropical waters. On Friday, Sept. 16, a Brazilian freighter reported the storm.

Gordon Dunn and Grady Norton, U.S. Weather Bureau forecasters in Florida, issued a hurricane warning for Miami, expecting the storm to hit Tuesday. Miami residents stockpiled supplies, boarded windows, secured boats.



But on Monday evening, the storm turned north and sped up to 20 mph, a swift forward speed that scientists of the time knew little about. It followed a typical path of recurvature -- resembling a giant C -- around the Bermuda High, an area of high atmospheric pressure in the mid-Atlantic that pushes air outward from its center, bouncing hurricanes off its edges like bubbles.

Dunn and Norton lost track of the storm off Cape Hatteras, North Carolina, and assumed it would curve eastward. The national weather map for Wednesday, Sept. 21, showed no hurricane, only a storm moving out to sea.

But winds had increased to 140 mph inside the beast. It moved north and picked up more forward speed.

The Bermuda High had moved to 44 degrees North latitude -- from its normal September position of thirty to thirty-five degrees -- blocking the hurricane's path and deflecting it northward. Abnormally warm water fed the storm as its forward speed increased to 60 mph! Waves tore up boardwalks in New Jersey.

No warnings were issued. It had been 117 years since the last great storm hit New York, on Sept. 23, 1821. History had been forgotten.

The 'fog bank'

By Wednesday afternoon, shingles were flying from roofs on Long Island. The sky grew dark. Trees were uprooted and telephone poles snapped like matchsticks. Three hours before high tide, residents reported a thick bank of gray fog, twenty-five to forty feet above the water, rolling in toward the south-facing coast.

Some residents fled to relative safety across the bridge. Many did not. Most of them died as the "fog bank" turned out to be a wall of water known as a storm surge.

Created partly by the vacuum of reduced atmospheric pressure and more so by the wind blowing over the water, the storm surge was highest in an area just east of the eye -- the right side of the storm when looked at from above. There, in the right-side eyewall, the counterclockwise winds combined with the storm's forward speed to create gusts exceeding 200 mph.

The eye of the hurricane passed some 55 miles east of Manhattan, a near miss in meteorological terms. Had it been a few miles west of Manhattan, forensic hurricanologists agree it would have devastated the island.

Drowned at work

The storm marched north. In Rhode Island, residents joked about putting up storm shutters. Finally, at least a warning had been issued. But the ferocity of the hurricane was no match for any last-minute preparations that might have been made.

At the head of Narragansett Bay, 30 miles from the Atlantic, the city of Providence awaited a fate its inhabitants could not have imagined. As the hurricane barreled toward the city, it pushed a surge of water up the bay. As the channel narrowed and became more shallow, the storm surge had nowhere to go but up.

The mountain of water carried boats and houses into the capital, flooding the first floor of buildings in downtown, where workers were just preparing to go home. Having no idea what a storm surge was, many of them were trapped and drowned where they worked. The surge of water caused tides (and even non-tidal sections of rivers) to rise 30 feet and more above normal levels.

Later that night, the Great New England Hurricane reached Canada, weakened but still potent enough to cause damage.

Over a four-day period, the 1938 storm dropped an average of 11 inches of rain over a 10,000-square-mile area, according to the National Weather Service. Flooding inflicted major damage through Connecticut, Massachusetts, New York, and Vermont, causing more than $300 million in losses.

In all, 600 people died. Ten of those deaths were in New York City. Flooding knocked out electrical power in all areas above 59th Street in Manhattan and in all of the Bronx. A hundred large trees in Central Park were destroyed.

Being somewhat of a voyeur when it comes to natural catastrophe, I couldn't resist running some scenarios through the new catastrophe calculator.

If you read on, please keep in mind that the odds of a serious impact occurring in any year are extremely low. A civilization-ending impact, while possible, almost surely won't happen within our lifetimes (90 percent of all asteroids big enough and close enough to do the job will be found by 2008) and is extremely unlikely even over the next millennium.

But hurling big virtual rocks at the planet is admittedly kind of fun. And in this case it's at least more scientifically meaningful than the average video game. I started by dropping a 9.3-mile-wide (15-kilometer) asteroid -- the estimated size of the suspected dinosaur killer -- on San Francisco.

The Bay Area doesn't do so well.

The resulting crater, at 113 miles (181 kilometers) wide, pretty much tells the story. The entire metropolis vanishes faster than you can say where you left your heart. What isn't consumed is knocked over in an earthquake of magnitude 10.2, bigger than any in recorded history. Heat from a scorching fireball would turn much of the state, and parts of others, into toast.

The quick end to the Bay Area turns out to be a blessing compared to what Los Angeles residents face.

About 10 seconds after impact, radiation from the fireball sears Southern California, igniting clothing and even plywood. Within two minutes the ground under Hollywood begins to shake. Weak brick structures crumble. Concrete irrigation ditches are damaged. Frame houses not properly bolted to their foundations are knocked off. Even tree branches fall.

And then it gets nasty.

'Bad things happen'

Six minutes after impact, much of earth that used to be under San Francisco has soared high into the atmosphere and begins to fall on the City of Angels (and just about everywhere else). Ultimately, a blanket of ejected material nearly 18 feet (5.5 meters) deep is deposited in and around LA.

Within a half-hour of the initial cosmic impact, on comes a 66-mph (30 meters per second) wind to rake what's left of Los Angeles.

"If you're close to the site of a major impact, some pretty bad things happen," said Jay Melosh, an expert in impact cratering at the University of Arizona.

Melosh oversaw the computer program's development. It's intended to help scientists further their research and journalists to report on the risks of asteroid impacts. "We could have put in some worse stuff, but it started getting grisly," he told SPACE.com .

I continued with my West Coast disaster scenario by heading across the country. Denver, about a third of the way over, is coated with nearly a foot of ejected material, called ejecta. Winds reach 22 mph.

New York City residents are spared enough that they can at first sit comfortably and watch all this on TV (to the extent there are any video feeds). Still, about 13 minutes after impact, windows and doors rattle in the East. After about 21 minutes, stuff shot out from the world's largest new crater starts to rain down on the East, depositing a half-inch blanket across the city. About 3-1/2 hours later an 8-mph wind arrives.

Object as big as the one that presumably wiped out the dinosaurs -- some investigators aren't convinced it was the sole cause -- are extremely rare, slamming into the planet perhaps just once every 300 million years.

Presently there are no asteroids known to be on a collision course with Earth. And experts agree that one this big would almost surely come with decades or centuries of warning time, being fairly easy to spot well before it becomes an imminent threat. By then, experts hope, methods will be developed to divert or destroy such a rock.

More likely scenarios

But what about something the size of the asteroid that dug out Meteor Crater in Arizona? That 600-foot-deep (180-meter) hole in the ground, popular now with tourists, is less than 50,000 years old and was created by a space rock so small that another one like it might not be noticed until it hits.

The Arizona culprit was only about 66 feet (20 meters) wide. But instead of fragile stone, it was composed mostly of iron. Examples linked with the online impact-effect program show how to specify this harder, more destructive material. A rock this size (though not of this density) is thought to hit Earth every 158 years, on average.

I duplicated the Meteor Crater script in Newark, New Jersey, 10 miles (16 kilometers) from Manhattan as the crow flies.

Location, of course, is everything. A nearly mile-wide crater is dug in New Jersey. A moderate earthquake of magnitude 4.7 rattles the region. New York City has handled such temblors in the past and largely endures. No significant ejecta hits Manhattan, and winds reach an insignificant 3.7 mph.

So I cranked it up a bit.

An asteroid the size of two football fields, even if stony in nature, is another story. It carves a crater in New Jersey that's 2 miles wide (3.4 kilometers). The resulting earthquake is magnitude 6.4, enough to frighten the pants off New York City emergency planners. A rock this size hits the planet about every 14,000 years.

Just estimates

Melosh developed the program along with fellow researcher Gareth Collins and undergraduate Robert Marcus. Melosh said the program should prove valuable to scientists who otherwise would have to wade through complicated calculations to generate impact scenarios for their research.

It's important to point out that what actually happens when an asteroid crashes into the surface is not well known, because scientists have never actually witnessed an impact event of any scale. But they have used computer models to explore the probable effects of impacts, and they have some craters to study on Earth.

Researchers assume the blast of heat generated by a cosmic collision would be similar to nuclear bomb detonations. So for these thermal radiation effects, Melosh's team relied on a 1977 publication, "The Effect of Nuclear Weapons," by the U.S. Defense Department and the Department of Energy.

"We determine at a given distance what type of damage the radiation causes," said Marcus, who did most of the grunt work developing the computer program. "We have descriptions like when grass will ignite, when plywood or newspaper will ignite, when humans will suffer 2nd or 3rd degree burns."

For earthquake predictions, the researchers culled data from actual California events and knowledge of how earthquakes propagate like sound waves through the planet.

The global effects of an asteroid impact are harder to predict. Many scientists think the larger impacts throw so much dust into the atmosphere that a global winter would ensue, lasting for months or years. Melosh is skeptical about the dust.

"Nuclear explosions are good at raising dust that's already there, but they're not good at creating it," Melosh said. He figures most of the ejected material is melted and then hardens into tiny, hardened and almost microscopic spheres. "Most of the ejecta comes down within an hour." He calls the massive dust creation "a media myth."

More to come

There is much left out of the new online impact program. Melosh wanted it to be practical to use, and he's open to suggestions for improvement.

Important details of an impact -- exactly how it penetrates the atmosphere and the surface, are simplified. And one major effect is not considered: the tsunami that would result from an ocean impact (Earth's surface is two-thirds water). In addition, the results of a user's input do not mention that a rush of wind can't blow down a tree that's buried under ejecta, nor does either matter if the tree were already burned to a crisp.

Minor and major revisions to the program may be made based on input from other scientists, Melosh said. And in coming weeks, his group will post background information explaining the complex calculations behind the program.

The Impact Effects program.

And there might not be enough time for a warning to be very useful.

The one-minute warning

The scientists say an underwater landslide just offshore, triggered by an earthquake, could generate a tsunami whose tallest waves would arrive "only one minute after the slide."

Other studies have the Pacific Northwest could be hit swiftly by devastating tsunamis, too. And researchers recently warned of a similar danger to Gulf Coast states. Prior to the tsunami last December, other geologists warned of the risk of mega-tsunamis around the entire Pacific basin from underwater landslides.

The new research was done by geologists, tsunami experts, engineers and urban planners at the University of Southern California. It is detailed in April edition of Civil Engineering magazine.

The group cites recent studies that suggest a large earthquake under the ocean off Los Angeles is likely at some point in the future. And it does not take a 9.0 "great quake," such as the one last December in Indonesia, to do the trick. A magnitude 7.0 earthquake near Papua New Guinea in 1998 generated deadly waves thought to be caused by a submarine landslide, which the earthquake triggered.

"The shaking from an earthquake of magnitude 7 or greater on an offshore thrust or reverse fault would undoubtedly be damaging to coastal communities [in Southern California], and its effect could be greatly magnified if it were to generate a tsunami," the scientists write.

Repeating history

The conclusions draw on evidence from several locally generated small tsunamis that have been recorded over the past 200 years. A Santa Barbara earthquake in 1812 spawned a moderate tsunami that affected more than 37 miles (60 kilometers) of coastline. Evidence of prehistoric events suggest submarine slumps that could have generated tsunamis up to 66 feet (20 meters) tall.

"There will be others," the researchers say, though of course they don't know when or how tall any tsunami might be.

Various scenarios in the study generate loss estimates of between $7 billion and $42 billion directly related to the tsunami. Costs include direct damage as well as the economic costs of lost shipping opportunities and traffic delays owing to damaged freeways.

"We chose not to model fatalities because we were being deliberately conservative, and because we wanted to avoid contentious assumptions about the economic value of life," said study member James Moore II. "The Papua New Guinea tsunami of 1998 was generated by a mechanism similar to the one modeled here, and that event cost over 2,000 lives. The toll here could be much higher."

"Super-eruptions are up to hundreds of times larger than these," said Stephen Self of the United Kingdom's (U.K.) Open University.

"An area the size of North America can be devastated, and pronounced deterioration of global climate would be expected for a few years following the eruption," Self said. "They could result in the devastation of world agriculture, severe disruption of food supplies, and mass starvation. These effects could be sufficiently severe to threaten the fabric of civilization."

Self and his colleagues at the Geological Society of London presented their report to the U.K. Government's Natural Hazard Working Group.

"Although very rare these events are inevitable, and at some point in the future humans will be faced with dealing with and surviving a super eruption," Stephen Sparks of the University of Bristol told LiveScience in advance of Tuesday's announcement.

Supporting evidence

The warning is not new. Geologists in the United States detailed a similar scenario in 2001, when they found evidence suggesting volcanic activity in Yellowstone National Park will eventually lead to a colossal eruption. Half the United States will be covered in ash up to 3 feet (1 meter) deep, according to a study published in the journal Earth and Planetary Science Letters.

Explosions of this magnitude "happen about every 600,000 years at Yellowstone," says Chuck Wicks of the U.S. Geological Survey, who has studied the possibilities in separate work. "And it's been about 620,000 years since the last super explosive eruption there."

Past volcanic catastrophes at Yellowstone and elsewhere remain evident as giant collapsed basins called calderas.

A super eruption is a scaled up version of a typical volcanic outburst, Sparks explained. Each is caused by a rising and growing chamber of hot molten rock known as magma.

"In super eruptions the magma chamber is huge," Sparks said. The eruption is rapid, occurring in a matter of days. "When the magma erupts the overlying rocks collapse into the chamber, which has reduced its pressure due to the eruption. The collapse forms the huge crater."

The eruption pumps dust and chemicals into the atmosphere for years, screening the Sun and cooling the planet. Earth is plunged into a perpetual winter, some models predict, causing plant and animal species disappear forever.

"The whole of a continent might be covered by ash, which might take many years -- possibly decades -- to erode away and for vegetation to recover," Sparks said.

Yellowstone may be winding down geologically, experts say. But they believe it harbors at least one final punch. Globally, there are still plenty of possibilities for super volcano eruptions, even as Earth quiets down over the long haul of its 4.5-billion-year existence.

"The Earth is of course losing energy, but at a very slow rate, and the effects are only really noticeable over billions rather than millions of years," Sparks said.

Human impact

The odds of a globally destructive volcano explosion in any given century are extremely low, and no scientist can say when the next one will occur. But the chances are five to 10 times greater than a globally destructive asteroid impact, according to the new British report.

The next super eruption, whenever it occurs, might not be the first one humans have dealt with.

About 74,000 years ago, in what is now Sumatra, a volcano called Toba blew with a force estimated at 10,000 times that of Mount St. Helens. Ash darkened the sky all around the planet. Temperatures plummeted by up to 21 degrees at higher latitudes, according to research by Michael Rampino, a biologist and geologist at New York University.

Rampino has estimated three-quarters of the plant species in the Northern Hemisphere perished.

Stanley Ambrose, an anthropologist at the University of Illinois, suggested in 1998 that Rampino's work might explain a curious bottleneck in human evolution: The blueprints of life for all humans -- DNA -- are remarkably similar given that our species branched off from the rest of the primate family tree a few million years ago.

Ambrose has said early humans were perhaps pushed to the edge of extinction after the Toba eruption -- around the same time folks got serious about art and tool making. Perhaps only a few thousand survived. Humans today would all be descended from these few, and in terms of the genetic code, not a whole lot would change in 74,000 years.

Sitting ducks

Based on the latest evidence, eruptions the size of the giant Yellowstone and Toba events occur at least every 100,000 years, Sparks said, "and it could be as high as every 50,000 years. There are smaller but nevertheless huge eruptions which would have continental to global consequences every 5,000 years or so."

Unlike other threats to mankind -- asteroids, nuclear attacks and global warming to name a few -- there's little to be done about a super volcano.

"While it may in future be possible to deflect asteroids or somehow avoid their impact, even science fiction cannot produce a credible mechanism for averting a super eruption," the new report states. "No strategies can be envisaged for reducing the power of major volcanic eruptions."

The Geological Society of London has issued similar warnings going back to 2000. The scientists this week called for more funding to investigate further the history of super eruptions and their likely effects on the planet and on modern society.

"Sooner or later a super eruption will happen on Earth and this issue also demands serious attention," the report concludes.

Click HERE to see image of the effects of a supervolcano.

Midwest record-setters

The infamous series of three New Madrid quakes in 1811-1812 occurred a few weeks apart, from Dec. 16 to Feb. 7. They measured 8.1, 8.0 and 7.8 and represent three of the four strongest earthquakes ever recorded in the lower 48.

"Strong earthquakes in the New Madrid seismic zone are certain to occur in the future," states a USGS fact sheet. "There is a 9-in-10 chance of a magnitude 6 to 7 temblor occurring in the New Madrid Seismic Zone within the next 50 years."

The New Madrid fault's remarkable spree of shifts devastated the sparsely populated area at the time. Amazingly, the quakes were so strong they were felt in much of the country -- as far away as Boston. Damage was reported as far away as Charleston, South Carolina, and Washington, D.C.

Differences in geology east of the Rockies account for the widespread effects, geologists say.

The strongest of the New Madrid events was 10 times more intense than the 1906 San Francisco earthquake, scientists say. Earthquakes of similar intensity occurred around the year 900 and again in the middle 1400s along the New Madrid fault, named for a small town in Missouri.

Good-sized temblors, the sort that can topple brick facades or bring down poorly designed bridges, rattled the New Madrid seismic zone in 1843 (6.3-magnitude) and 1895 (6.6-magnitude).

All this taken into account, plus the much larger population in the danger zone now, suggests a magnitude 7 quake on the New Madrid fault today "could be far worse than those of the 1989 magnitude 7 Loma Prieta, California, earthquake," according to the USGS. That rattle of the bay, which struck the San Francisco area during the World Series, killed 63 people and caused $6 billion in damage.

Not prepared

In California and other parts of the West, many old buildings and roadways have been retrofitted to largely endure earthquakes in the magnitude 6 to 7 range. But buildings in the New Madrid fault region constructed before modern building codes were put in place typically are not retrofitted, according to a USGS statement.

"This earthquake is an opportunity to gather more knowledge about the region's seismic risk and incorporate it into safer buildings," Groat said.

The regional Center for Community Earthquake Preparedness issued a statement Thursday saying the quake "underscores the need for Mid-South emergency management officials to have a seismic response plan ready." The CCEP is staffed by engineers and geologists from the University of Mississippi.

The Top 10

Here are the Top 10 U.S. earthquakes for all 50 states, according to the USGS:

1. Prince William Sound, Alaska 1964, 9.2
2. Andreanof Islands, Alaska 1957, 9.1
3. Rat Islands, Alaska 1965, 8.7
4. East of Shumagin Islands, Alaska 1938, 8.2
5. New Madrid, Missouri 1811, 8.1
6. Yakutat Bay, Alaska 1899, 8.0
7. Andreanof Islands, Alaska 1986, 8.0
8. New Madrid, Missouri 1812, 8.0
9. Near Cape Yakataga, Alaska 1899, 7.9
10. Fort Tejon, California 1857, 7.9

The largest earthquake in the world since measurements have been made was a 9.5-magnitude event that shook Chile in 1960. The 1964 Alaskan quake had held the No. 2 spot globally until this week, when scientists provided an updated estimate of the Dec. 26, 2004 tsunami-causing Sumatran earthquake. That subsea event has now been pegged at 9.3.

Scientists note that all of the magnitude numbers are estimates, especially the historical temblors for which there exist little or no reliable instrumented measurements.

It is important to note that there is no adjustment made for inflation when determining what events initially qualify for the list -- they had to cost $1 billion at the time they occurred. This results in a slightly greater probability that a more recent event will make the list.

However, in comparing the total cost from a year's disasters once they've made the list, the dollar amounts are normalized to reflect inflation. The total damages for all 62 events combined are over $390 billion in 2002 dollars.

Beach property

In an accompanying report, NCDC researchers Tom Ross and Neal Lott comment on one noticeable trend in the data: an increase in losses due to hurricanes.

Accounting for almost a third of the billion-dollar list, tropical storms and hurricanes are the most frequent type of disaster in the years studied. This is partly due to an apparent upswing recently in the frequency of major hurricanes, which are defined by sustained wind speeds above 110 mph. The frequency of hurricanes is thought to ebb and flow in long-term cycles related to global climate factors.

The thirty years between 1961 and 1990 saw very few major hurricanes make landfall on the United States. This is in contrast to the 18 major hurricanes that struck between 1941 and 1960.

Starting with Hurricanes Hugo (1989) and Andrew (1992), there now appears to be a return to the active hurricane seasons of the '40s and '50s.

But in assessing the damages that these storms wrought, Ross and Lott stress that other factors must be included, such as population changes.

Americans have been moving to the coastlines in droves. Between 1940 and 1990, the coastal population of Florida has grown four times faster than the rate of the nation as a whole. The 426 counties along the Atlantic Ocean and Gulf of Mexico comprise 11 percent of the land in the continental U.S., but account for 45 percent of the population.

If a hurricane hits, there is more now to destroy. This is also true for other disasters, like flooding.

Heat danger

As dramatic as the recent hurricanes have been, it is perhaps surprising that droughts and heat waves cause the most damage and deaths.

In both 1980 and 1988, severe drought and heat ravaged the central and eastern parts of the country. They were the only billion dollar events on the list for those two years, yet they alone made 1980 and 1988 the two costliest years on the list - $48 and $62 billion, respectively, in 2002 dollars.

These heat waves also were the deadliest weather or climate events of the study period - with estimated deaths due to heat stress approaching 10,000 in each case.

In comparison, 1998 saw the most events - seven - but it was not an especially high year in total damages, with just over $20 billion in 2002 dollars.

The United States sustained 62 weather-related disasters between 1980 and 2004 in which overall damages and costs were $1 billion or more at the time of the event. Any event under that amount is not on the list, even if it would be based on an inflation adjustment. That means some earlier events that might have been as financially severe in relative terms may not be represented here.

Some inflation adjusting has been done for events that initially met the billion dollar threshold. Two cost figures are given for events prior to 2002: the first represents actual dollar costs at the time of the event; the second value in parenthesis (if given) is the dollar cost normalized to 2002 dollars, allowing for more accurate comparison of damage figures over time.

The total normalized losses from the 62 events exceed $390 billion.

2004

Hurricane Jeanne September 2004. Category 3 hurricane makes landfall in east-central Florida, causing considerable wind, storm surge, and flooding damage in FL, with some flood damage also in the states of GA, SC, NC, VA, MD, DE, NJ, PA, and NY. Puerto Rico also affected. Preliminary estimate of over $6.5 billion in damage/costs; at least 28 deaths.

Hurricane Ivan September 2004. Category 3 hurricane makes landfall on Gulf coast of Alabama, with significant wind, storm surge, and flooding damage in coastal AL and FL panhandle, along with wind/flood damage in the states of GA, SC, NC, VA, LA, MS, WV, MD, TN, KY, OH, DE, NJ, PA, and NY. Preliminary estimate of over $12 billion in damage/costs; at least 52 deaths.

Hurricane Frances September 2004. Category 2 hurricane makes landfall in east-central Florida, causing significant wind, storm surge, and flooding damage in FL, along with considerable flood damage in the states of GA, SC, NC, and NY due to 5-15 inch rains. Preliminary estimate of over $9 billion in damage/costs; at least 38 deaths.

Hurricane Charley August 2004. Category 4 hurricane makes landfall in southwest Florida, resulting in major wind and some storm surge damage in FL, along with some damage in the states of SC and NC. Preliminary estimate of over $14 billion in damage/costs; at least 34 deaths.

2003

Southern California Wildfires Late October to early November 2003. Dry weather, high winds, and resulting wildfires in Southern California. More than 743,000 acres of brush and timber burned, over 3700 homes destroyed; over $2.5 billion damage/costs; 22 deaths. October and November.

Hurricane Isabel September 2003. Category 2 hurricane makes landfall in eastern North Carolina, causing considerable storm surge damage along the coasts of NC, VA, and MD, with wind damage and some flooding due to 4-12 inch rains in NC, VA, MD, DE, WV, NJ, NY, and PA; estimate of approximately $5 billion in damages/costs; at least 55 deaths.

Severe Storms and Tornadoes Early May 2003. Numerous tornadoes over the midwest, MS valley, OH/TN valleys and portions of the southeast, with a modern record one-week total of approximately 400 tornadoes reported; over $3.4 billion in damages/costs; 51 deaths.

Storms and Hail Early April 2003. Severe storms and large hail over the southern plains and lower MS valley, with Texas hardest hit, and much of the monetary losses due to hail; over $ 1.6 billion in damages/costs: 3 deaths.

2002

Widespread Drought Spring through early Fall 2002. Moderate to Extreme drought over large portions of 30 states, including the western states, the Great Plains, and much of the eastern U.S.; estimate of over $ 10.0 billion in damages/costs; no deaths.

Western Fire Season Spring through Fall 2002. Major fires over 11 western states from the Rockies to the west coast, due to drought and periodic high winds, with over 7.1 million acres burned; over $ 2.0 billion in damages/costs; 21 deaths.

2001

Tropical Storm Allison June 2001. The persistent remnants of Tropical Storm Allison produces rainfall amounts of 30-40 inches in portions of coastal Texas and Louisiana, causing severe flooding especially in the Houston area, then moves slowly northeastward; fatalities and significant damage reported in TX, LA, MS, FL, VA, and PA; estimate of approximately $5.0 (5.1) billion in damage/costs; at least 43 deaths.

Midwest and Ohio Valley Hail and Tornadoes April 2001. Storms, tornadoes, and hail in the states of TX, OK, KS, NE, IA, MO, IL, IN, WI, MI, OH, KY, WV, and PA, over a 6-day period; over $1.9 billion in damage/costs, with the most significant losses due to hail; at least 3 deaths.

2000

Drought/Heat Wave Spring-Summer 2000. Severe drought and persistent heat over south-central and southeastern states causing significant losses to agriculture and related industries; estimate of over $4.0 (4.2) billion in damage/costs; estimated 140 deaths nationwide.

Western Fire Season Spring-Summer 2000. Severe fire season in western states due to drought and frequent winds, with nearly 7 million acres burned; estimate of over $2.0 (2.1) billion in damage/costs (includes fire suppression); no deaths reported.

1999

Hurricane Floyd September 1999. Large category 2 hurricane makes landfall in eastern NC, causing 10-20 inch rains in 2 days, with severe flooding in NC and some flooding in SC, VA, MD, PA, NY, NJ, DE, RI, CT, MA, NH, and VT; estimate of at least $6.0 (6.5) billion damage/costs; 77 deaths.

Eastern Drought/Heat Wave Summer 1999. Very dry summer and high temperatures, mainly in eastern U.S., with extensive agricultural losses; over $1.0 (1.1) billion damage/costs; estimated 502 deaths.

Oklahoma-Kansas Tornadoes May 1999. Outbreak of F4-F5 tornadoes hit the states of Oklahoma and Kansas, along with Texas and Tennessee, Oklahoma City area hardest hit; over $1.6 (1.7) billion damage/costs; 55 deaths.

Arkansas-Tennessee Tornadoes January 1999. Two outbreaks of tornadoes in 6-day period strike Arkansas and Tennessee; approximately $1.3 (1.4) billion damage/costs; 17 deaths.

1998

Texas Flooding October-November 1998. Severe flooding in southeast Texas from 2 heavy rain events, with 10-20 inch rainfall totals; approximately $1.0 (1.1) billion damage/costs; 31 deaths.

Hurricane Georges September 1998. Category 2 hurricane strikes Puerto Rico, Florida Keys, and Gulf coasts of Louisiana, Mississippi, Alabama, and Florida panhandle, 15-30 inch 2-day rain totals in parts of AL/FL; estimated $5.9 (6.5) billion damage/costs; 16 deaths.

Hurricane Bonnie August 1998. Category 3 hurricane strikes eastern North Carolina and Virginia, extensive agricultural damage due to winds and flooding, with 10-inch rains in 2 days in some locations; approximately $1.0 (1.1) billion damage/costs; 3 deaths.

Southern Drought/Heat Wave Summer 1998. Severe drought and heat wave from Texas/Oklahoma eastward to the Carolinas; $6.0-$9.0 billion (6.6-9.9) damage/costs to agriculture and ranching; at least 200 deaths.

Minnesota Severe Storms/Hail May 1998. Very damaging severe thunderstorms with large hail over wide areas of Minnesota; over $1.5 (1.7) billion damage/costs; 1 death.

Southeast Severe Weather Winter-Spring 1998. Tornadoes and flooding related to El Nino in southeastern states; over $1.0 (1.1) billion damage/costs; at least 132 deaths.

Northeast Ice Storm January 1998. Intense ice storm hits Maine, New Hampshire, Vermont, and New York, with extensive forestry losses; over $1.4 (1.5) billion damage/costs; 16 deaths.

1997

Northern Plains Flooding April-May 1997. Severe flooding in Dakotas and Minnesota due to heavy spring snowmelt; approximately $3.7 (4.1) billion damage/costs; 11 deaths.

MS and OH Valleys Flooding and Tornadoes March 1997. Tornadoes and severe flooding hit the states of AR, MO, MS, TN, IL, IN, KY, OH, and WV, with over 10 inches in 24 hours in Louisville; estimated $1.0 (1.1) billion damage/costs; 67 deaths.

West Coast Flooding December 1996-January 1997. Torrential rains (10-40 inches in 2 weeks) and snowmelt produce severe flooding over portions of California, Washington, Oregon, Idaho, Nevada, and Montana; approximately $3.0 (3.4) billion damage/costs; 36 deaths.

1996

Hurricane Fran September 1996. Category 3 hurricane strikes North Carolina and Virginia, over 10-inch 24-hour rains in some locations and extensive agricultural and other losses; over $5.0 (5.8) billion damage/costs; 37 deaths.

Southern Plains Severe Drought Fall 1995 through Summer 1996. Severe drought in agricultural regions of southern plains--Texas and Oklahoma most severely affected; approximately $5.0 (6.0) billion damage/costs; no deaths.

Pacific Northwest Severe Flooding February 1996. Very heavy, persistent rains (10-30 inches) and melting snow over Oregon, Washington, Idaho, and western Montana; approximately $1.0 (1.2) billion damage/costs; 9 deaths.

Blizzard of '96 Followed by Flooding January 1996. Very heavy snowstorm (1-4 feet) over Appalachians, Mid-Atlantic, and Northeast; followed by severe flooding in parts of same area due to rain and snowmelt; approximately $3.0 (3.5) billion damage/costs; 187 deaths.

1995

Hurricane Opal October 1995. Category 3 hurricane strikes Florida panhandle, Alabama, western Georgia, eastern Tennessee, and the western Carolinas, causing storm surge, wind, and flooding damage; over $3.0 (3.6) billion damage/costs; 27 deaths.

Hurricane Marilyn September 1995. Category 2 hurricane devastates U.S. Virgin Islands; estimated $2.1 (2.5) billion damage/costs; 13 deaths.

Texas/Oklahoma/Louisiana/Mississippi Severe Weather and Flooding May 1995. Torrential rains, hail, and tornadoes across Texas - Oklahoma and southeast Louisiana - southern Mississippi, with Dallas and New Orleans areas (10-25 inches in 5 days) hardest hit; $5.0-$6.0 (6.5-7.1) billion damage/costs; 32 deaths.

California Flooding January-March 1995. Frequent winter storms cause 20-70 inches rainfall and periodic flooding across much of California; over $3.0 (3.6) billion damage/costs; 27 deaths.

1994

Western Fire Season Summer-Fall 1994. Severe fire season in western states due to dry weather; approximately $1.0 (1.2) billion damage/costs; death toll undetermined.

Texas Flooding October 1994. Torrential rain (10-25 inches in 5 days) and thunderstorms cause flooding across much of southeast Texas; approximately $1.0 (1.2) billion damage/costs; 19 deaths.

Tropical Storm Alberto July 1994. Remnants of slow-moving Alberto brought torrential 10-25 inch rains in 3 days, widespread flooding, and agricultural damage in parts of Georgia, Alabama, and panhandle of Florida; approximately $1.0 (1.2) billion damage/costs; 32 deaths.

Southeast Ice Storm February 1994. Intense ice storm with extensive damage in portions of TX, OK, AR, LA, MS, AL, TN, GA, SC, NC, and VA; approximately $3.0 (3.7) billion damage/costs; 9 deaths.

1993

California Wildfires Fall 1993. Dry weather, high winds, and wildfires in southern California; approximately $1.0 (1.3) billion damage/costs; 4 deaths.

Midwest Flooding Summer 1993. Severe, widespread flooding in central U.S. due to persistent heavy rains and thunderstorms; approximately $21.0 (26.7) billion damage/costs; 48 deaths.

Drought/Heat Wave Summer 1993. Southeastern U.S.; about $1.0 (1.3) billion damage/costs to agriculture; at least 16 deaths.

Storm/Blizzard March 1993. "Storm of the Century" hits entire eastern seaboard with tornadoes, high winds, and heavy snows (2-4 feet); $5.0-$6.0 (6.3-7.6) billion damage/costs; approximately 270 deaths.

1992

Nor'easter of 1992 December 1992. Slow-moving storm batters northeast U.S. coast, New England hardest hit; $1.0-$2.0 (1.3-2.6) billion damage/costs; 19 deaths.

Hurricane Iniki September 1992. Category 4 hurricane hits Hawaiian island of Kauai; about $1.8 (2.4) billion damage/costs; 7 deaths.

Hurricane Andrew August 1992. Category 5 hurricane hits Florida and Louisiana, high winds damage or destroy over 125,000 homes; approximately $27.0 (35.6) billion damage/costs; 61 deaths.

1991

Oakland Firestorm October 1991. Oakland, California firestorm due to low humidities and high winds; approximately $2.5 (3.5) billion damage/costs; 25 deaths.

Hurricane Bob August 1991. Category 2 hurricane--Mainly coastal North Carolina, Long Island, and New England; $1.5 (2.1) billion damage/costs; 18 deaths.

1990

Texas/Oklahoma/Louisiana/Arkansas Flooding May 1990. Torrential rains cause flooding along the Trinity, Red, and Arkansas Rivers in TX, OK, LA, and AR; over $1.0 (1.4) billion damage/costs; 13 deaths.

1989

Hurricane Hugo September 1989. Category 4 hurricane devastates South and North Carolina with ~ 20-foot storm surge and severe wind damage after hitting Puerto Rico and the U.S. Virgin Islands; over $9.0 (13.9) billion damage/costs (about $7.1 (10.9) billion in Carolinas); 86 deaths (57--U.S. mainland, 29--U.S. Islands).

Northern Plains Drought Summer 1989. Severe summer drought over much of the northern plains with significant losses to agriculture; at least $1.0 (1.5) billion in damage/costs; no deaths reported.

1988

Drought/Heat Wave Summer 1988. 1988 drought in central and eastern U.S. with very severe losses to agriculture and related industries; estimated $40.0 (61.6) billion damage/costs; estimated 5,000 to 10,000 deaths (includes heat stress-related).

1986

Southeast Drought/Heat Wave Summer 1986. Severe summer drought in parts of the southeastern U.S. with severe losses to agriculture; $1.0-$1.5 (1.8-2.6) billion in damage/costs; estimated 100 deaths.

1985

Hurricane Juan October-November 1985. Category 1 hurricane--Louisiana and Southeast U.S.--severe flooding; $1.5 (2.8) billion damage/costs; 63 deaths.

Hurricane Elena August-September 1985. Category 3 hurricane--Florida to Louisiana; $1.3 (2.4) billion damage/costs; 4 deaths.

Florida Freeze January 1985. Severe freeze central/northern Florida; about $1.2 (2.2) billion damage to citrus industry; no deaths.

1983

Florida Freeze December 1983. Severe freeze central/northern Florida; about $2.0 (4.0) billion damage to citrus industry; no deaths.

Hurricane Alicia August 1983. Category 3 hurricane--Texas; $3.0 (5.9) billion damage/costs; 21 deaths.

Western Storms and Flooding 1982 - Early 1983. Storms and flooding related to El Nino, especially in the states of WA, OR, CA, AZ, NV, ID, UT, and MT; approximately $1.1 (2.2) billion in damage/costs; at least 45 deaths.

Gulf States Storms and Flooding 1982 - Early 1983. Storms and flooding related to El Nino, especially in the states of TX, AR, LA, MS, AL, GA, and FL; approximately $1.1 (2.2) billion in damage/costs; at least 50 deaths.

1980

Drought/Heat Wave June-September 1980. Central and eastern U.S.; estimated $20.0 (48.4) billion damage/costs to agriculture and related industries; estimated 10,000 deaths (includes heat stress-related).

"You'd want to first decide how far away [from impact] the effects are going to be felt," Steven Ward, a research geophysicist at the University of California Santa Cruz (UCSC), said in a telephone interview. "And once you know that, maybe you could set up a zone of evacuation." According to Ward's research, about 120 million people live in coastal areas at elevations within 65 feet (20 meters) of sea level and just over a mile (two kilometers) of the ocean.

Ward is the lead author of the simulation study, which will appear in the June issue of the Geophysical Journal International.

Ward and co-author Erik Asphaug, an associate professor of earth sciences at UCSC, used 1950 DA as the asteroid in their simulation because of its history. Astronomers first discovered the asteroid in 1950 and tracked it for 17 days. The object was seen again in 2000, when researchers determined its radius and other characteristics. Its most recent Earth flyby last year led researchers to determine its future paths up through 2880, when it crosses the planet's orbit.

Asteroids are fascinating objects because they're the only disasters that you can prepare for, not like volcano eruptions, earthquakes or major storms," Asphaug told SPACE.com.

On March 16, 2880 the Earth's Atlantic Ocean is turned toward 1950 DA, and with 70-percent of the Earth covered in water an ocean impact would be the most likely to model.

In the simulation, 1950 DA slams into the Atlantic Ocean about 360 miles (579 kilometer) off the eastern coast of the United States. The asteroid, traveling at 38,000 miles an hour (61,155 kilometers per hour), is vaporized in the impact's 60,000-megaton blast and blows a hole in the ocean straight to the seafloor.

Seawater rushes into the cavity, which spans 11 miles in diameter and three miles deep (about 17 kilometers wide and five deep), and a halo of waves begin their journey outward from the impact site like overgrown ripples from a pond. Two hours after impact, waves 400 feet (121 meters) high hit the coastline from Cape Cod in Massachusetts to Cape Hatteras in North Carolina, and could ultimately penetrate more than two miles (about four kilometers) inland.

1950 DA is one of the larger asteroids that astronomers worry might hit the Earth. On the average, space rocks its size strike the earth every 100,000 years or so. Smaller asteroids up to 300 meters wide do so more frequently, about every 5,000 years, and so pose a larger tsunami danger for coastal residents, Ward said.

Asphaug, who led a NASA-sponsored workshop on asteroid hazards in September 2002, said that while not a lot can be done about larger, kilometer-sized objects that can cause global devastation, some action could be taken for smaller rocks. Observations, for example, could be fed into the already existing Pacific Tsunami Warning Center to track potential impacts.

"It takes time for these tsunamis to reach coastlines from the point of impact," Asphaug said, adding that once a satellite spots the flash of an impact, there will be a period of hours before the first big waves hit the shore. "It's possible to think that in 800 years, we might have enough technology to pinpoint which areas to evacuate and then do it."

The major source for past tsunamis in the northern Caribbean has been movement along the boundary between the North American and Caribbean tectonic plates. This fault line stretches 2,000 miles (3,200 kilometers) from Central America to the Lesser Antilles, brushing up against the north coast of Hispaniola (the island of Haiti and the Dominican Republic).

Nancy Grindlay and Meghan Hearne of the University of North Carolina and Paul Mann of the University of Texas identified 10 significant tsunamis that have resulted from movement along this plate boundary. Six of these caused loss of life.

In 1692, a tsunami destroyed Port Royal, Jamaica; another killed at least 10 Jamaicans on the island's south coast in 1780. The most recent tsunami in 1946 was triggered by a magnitude 8.1 earthquake in the Dominican Republic. It killed around 1,800 people.

Jian Lin of the Woods Hole Oceanographic Institution believes that this sort of historical analysis can indicate how frequent big tsunamis are in a geographic region. It also gives an estimate of how large such events can potentially be.

"The tectonic setting of the northern Caribbean is very similar to the Indian Ocean - except that the subduction zone is not as long," Lin told LiveScience in a telephone interview.

The subduction zone is where one plate dips below another. Lin, who was not involved in the recent research, explained that the longer a subduction zone is, the larger the earthquake that the zone is capable of producing.

"The [historical analysis] shows that the Caribbean zone is long enough to have greater than a magnitude 8.0 earthquake," Lin said.

In comparison, the Sumatra earthquake that unleashed last year's tsunami in the Indian Ocean had a magnitude of 9.3.

Besides the direct threat from plate movement, other research has shown that underwater landslides in the region - or even in the middle of the Atlantic - could trigger a giant tsunami.

"The recent devastating tsunami in the Indian Ocean has raised public awareness of tsunami hazard and the need for early warning systems in high-risk areas such as the Caribbean," Grindlay said in a statement.

There are meetings scheduled later this year to implement an Intra-Americas Sea Tsunami Warning Project, as approved by the Intergovernmental Oceanographic Commission. Such a warning system has been set up in the Pacific Ocean, and one is planned for the Indian Ocean. The United States has also proposed a global warning system.

A report by Grindlay and her colleagues will appear in the March 22 issue of Eos, the newspaper of the American Geophysical Union.

The research was supported by the National Science Foundation and the University of Puerto Rico SeaGrant program.

June 15, 1896: Waves as high as 100 feet (30 meters), spawned by an earthquake, swept the east coast of Japan. Some 27,000 people died.

April 1, 1946: The April Fools tsunami, triggered by an earthquake in Alaska, killed 159 people, mostly in Hawaii.

July 9, 1958: Regarded as the largest recorded in modern times, the tsunami in Lituya Bay, Alaska was caused by a landslide triggered by an 8.3 magnitude earthquake. Waves reached a height of 1,720 feet (576 meters) in the bay, but because the area is relatively isolated and in a unique geologic setting the tsunami did not cause much damage elsewhere. It sank a single boat, killing two fishermen.

May 22, 1960: The largest recorded earthquake, magnitude 8.6 in Chile, created a tsunami that hit the Chilean coast within 15 minutes. The surge, up to 75 feet (25 meters) high, killed an estimated 1,500 people in Chile and Hawaii.

March 27, 1964: The Alaskan Good Friday earthquake, magnitude between 8.4, spawned a 201-foot (67-meter) tsunami in the Valdez Inlet. It traveled at over 400 mph, killing more than 120 people. Ten of the deaths occurred in Crescent City, in northern California, which saw waves as high as 20 feet (6.3 meters).

Aug. 23, 1976: A tsunami in the southwest Philippines killed 8,000 on the heels of an earthquake.

July 17, 1998: A magnitude 7.1 earthquake generated a tsunami in Papua New Guinea that quickly killed 2,200.

The largest earthquake ever measured globally - a magnitude 9.5 in Chile - was on May 22, 1960. And the recent tsunami-causing earthquake off the coast of Sumatra happened the day after Christmas last year. It was first put at magnitude 9.0 but later raised to 9.3.

"Earthquakes occur randomly throughout the world, as well as the U.S.," said Waverly Person, a spokesman for the U.S. Geological Survey's National Earthquake Information Center.

Earthquake weather

There are claims that more earthquakes happen on hot, dry days - so-called "earthquake weather." But studies that have shown no tie between seismic activity and months or seasons.

The forces that drive weather are not likely to have much of an effect on the movement of continental plates several miles below the surface, geologists say.

Even the tide-causing gravity of the Moon and Sun apparently play no role in the onset of an earthquake. Scientists have looked for a seismic relationship to the tides, but nothing has been seen except for a small correlation with aftershocks in some volcanic regions.

Morning madness

Another misconception is that most rumblings occur in the morning. Although the likelihood of an earthquake is unrelated to the time of day, an earthquake could be more deadly - and therefore more memorable - at certain hours.

"In the early morning, most people are in their beds sleeping. You might have higher casualties with a lot of people in one place," Person told LiveScience.

As opposed to times, there are, of course, places that are more earthquake-prone than others. Alaska is one of the most seismically active regions in the world - experiencing a major earthquake (magnitude 7.0 or greater) almost every year, whereas no earthquakes of more than moderate intensity have occurred within the borders of North Dakota during historical times.

The geologists have observed a possible correlation between the tremors and the rate of small microquakes in the region.

"This is new information from an area deep down under the fault we have not been able to look at before," Nadeau said. "If these tremors are precursory to earthquakes, there is potential here for earthquake forecasting and prediction."

The town of Cholame is thought to have been the origin of the magnitude 8 Fort Tejon Earthquake of 1857. This was the last big quake to hit southern California, and some seismologists think the area is due for another, since the average time between big quakes is 140 years.

Although the study was no longer running at the time, a moderate, magnitude 6 earthquake erupted on Sept. 28, 2004, outside the city of Parkfield, which is 15 miles northwest of Cholame. Because this quake was near the tremor region, Nadeau believes it supports the tremor-quake relationship.


Parkfield describes itself as the "Earthquake Capital of the World" because for 20 years seismologists have been studying the fault-line that cuts through the city. Between the quakes in 1887 and 2004, the area has been hit by five other large events in 1881, 1901, 1922 and 1934.

Earthquake magnitudes are measured with seismographs and rated on the Richter scale:

* 2.5 or less: Usually not felt, but can be recorded by seismograph. 900,000 per year worldwide.
* 2.5 to 5.4: Often felt, but only causes minor damage. 30,000 per year.
* 5.5 to 6.0: Slight damage to buildings and other structures. 500 per year.
* 6.1 to 6.9: May cause a lot of damage in very populated areas. 100 per year.
* 7.0 to 7.9: Major earthquake. Serious damage. 20 per year.
* 8.0 or greater: Great earthquake. Can totally destroy communities near the epicenter. One every 5 to 10 years.

Small tremors have typically been ignored by seismologists, who are more interested in short bursts of activity rather than a sustained rumbling.

But tremors have been observed under volcanoes, and recently led to predictions of the eruption of Mount St. Helens in Washington State. They have also been discovered in Japan and the Pacific Northwest at sites called subduction zones, where one of the Earth's plates dips underneath another.

Nadeau and Dolenc's tremors are the first detected underneath a transform fault, which is where two plates scrape against each other in a horizontal direction. It had been thought that tremors result from fluids flowing deep underground, but Nadeau said that his findings challenge this theory.

"Transform faults like the San Andreas have no obvious source of fluid, so it's not clear what's causing the tremors," he said. "Either tremors don't need fluid, or there is another, unknown source of fluid, perhaps from the Earth's mantle."

By understanding this mechanism better, the researchers hope to uncover whether tremors really can predict earthquakes.

Nadeau and Dolenc describe their work today in the online version of the journal Science.