People truly have short memories.

Driving out to Chulan, where the September quake wreaked the greatest damage along the Chelungpu fault, you pass Yutengping. The town is well known among train buffs for the scenic Yutengping Railway Bridge, which graces the many a postcard despite being half toppled and long out of service. Few people, however, know that after it collapsed during the 1935 Houli earthquake, the local government preserved it as an "earthquake memorial bridge."

The Houli quake was triggered by slippage along the Shihtan and Tuntzuchiao faults and caused more than 3,000 deaths. Houli elementary school, then a Japanese public elementary school, sat directly on a fault line and was completely destroyed. They built an earthquake memorial on the site, but after retrocession everyone forgot all about the quake, and the memorial was dragged behind Houli City Hall. Remarkably, the site was used to build yet another school, Houli Junior High! Fortunately, the September 21 quake didn't involve the local fault, but the school still suffered costly damages, with a number of its windows falling out.

Although the map of active fault lines prepared by the Central Geological Survey in 1998 is not entirely in accordance with the findings of the academic community, it is the most complete map of its kind currently available.

Sudden concern

"People have finally realized that it's dangerous to live on fault lines!" said one geologist at a seminar held after the Chichi quake.

Unless people have personally experienced a major quake, it's hard for them to understand how dangerous faults are. This is particularly so in Taiwan, where minor quakes are very common. For many years there had been little discussion here about how fault activity causes earthquakes.

The big Chichi quake finally reminded people of how destructive faults can be. With its hypocenter kilometers under Nantou County, the quake wreaked devastation along the 80-kilometer Chelungpu fault. It has caused people across the island to grow terrified of faults. When aftershocks occur, everyone immediately asks: Which fault caused it? Once an earthquake measuring 4.0 on the Richter scale occurred in Hsinchu that was in fact unrelated to surface fault activity, but everyone nonetheless turned their attention to a newly discovered fault line running under the Hsinchu Science Park: Could it cause earthquakes? Unfortunately, experts couldn't say. The long and the short of it is that we simply don't understand very much about geologic faults.

What exactly is the relationship between geologic faults and earthquakes?

Let's start with the basics. When tectonic plates collide, damage isn't restricted to the area of contact. Instead, energy can be displaced, causing ruptures to appear on the surface of the plates in any direction. For instance, by slamming into Asia, India has caused the rise of the Tibetan Plateau. But energy from this collision of plates has also been released all across mainland China. The 1976 earthquake at Tangshan in Hebei Province that caused 240,000 fatalities, for instance, was a result of this geologic movement. Fault lines are ruptures in the rock layers of the earth's crust caused by the collision of plates.

When plates are continually moving against each other, energy accumulates. Release of this energy results in the earthquakes that people feel, which may create new fault lines. Taiwan's 1946 Hsinhua earthquake, for instance, created the Hsinhua fault. Because the Chiayi-Tainan area hasn't had a earthquake with a magnitude of 7 or more for 30 years, scientists believed that the area was ripe for a big quake, and on October 22 it did indeed experience an earthquake with a magnitude of 6.4. Which fault was the culprit? According to the Central Weather Bureau, it was none other than the Meishan fault, which was created by the 1906 earthquake.

The energy created by plate motion can also cause existing fault lines to move or "slip." The energy from the Chichi quake, for instance, was released along the previously existing Chelungpu fault. When a lot of built-up stress is suddenly released, existing faults can move in unpredictable ways, resulting in major earthquakes. But plate movement doesn't necessarily cause earthquakes, or it may only create underground faults, with no ruptures appearing on the surface. As a result, some people in Taiwan, knowing of faults in connection with landslides during typhoons, may not have previously even connected them to earthquakes at all. Then there is the process called "fault creep," by which plates can gradually move a few centimeters a year without sudden release of huge amounts of energy in the form of earthquakes.

There is a waterfall where the Chelungpu fault crosses the Tachia River. Did you know that faults are responsible for many of the waterfalls in Taiwan?

Where aren't there fault lines?

Exactly how many fault lines are there in Taiwan? Because of the loss of life and destruction of property along the Chelungpu fault during the Chichi quake, the Central Geological Survey's map of fault lines has become a hot seller. The 51 fault lines on the map, however, are only drawn based on information that scholars, government departments and various large construction projects passed along to the CGS during the 1940s and 1950s. From an academic standpoint, the map is more like a book report than hard scholarship. Chen Wen-shan, an associate professor of geology at National Taiwan University, says that in 50 years the government probably hasn't spent more than NT$10 million on fault line surveys. Although the CGS publishes the fault line map, it doesn't dare distribute it, for fear that it will adversely affect the construction industry.

If the number of faults is larger than 51, then exactly how large is it? "With plates moving for six million years, where aren't there fault lines?" So writes one earth scientist on the Internet. Some scientists use the term "myriad"; others say the faults are "multitudinous." In any case, "there are too many to survey completely," writes Professor Shih Reuy-chyuan from the Institute of Seismology at National Chung Cheng University on an Internet site.

But just because Taiwan has countless fault lines doesn't mean that we are living above a fiery pit of hell. While all the lines in this geologic spider's web were active at some point, their "friction coefficient," to use the jargon of science, has increased as they have gradually been pushed together. The faults, in other words, become harder to push into motion over time. You could say that they get more and more stubborn as they age. Like old scars, they become "inured to pain." The plates turn to bully younger faults, because scabs on new wounds are easier to reopen.

Personnel from Taipei City assisted in surveying fault lines after the quake. In facing this highly fractured fault, the technicians performing the tests couldn't help but shake their heads and sigh. The irregularity of the ruptures makes it particularly difficult to determine the exact location of the fault lines.

Keeping still or staying in motion

The CGS divides faults into three groups according to current information about the history of fault line activity. The first group consists of faults that have been active sometime over the past 10,000 years. These are the ones most likely to move again and are the most dangerous. The second category includes faults that have been active sometime in the last 100,000 years, while faults that were last active more than 100,000 years ago fall into the third category.

Where is the highest concentration of active faults in Taiwan? As the Philippine plate smashes into the Eurasian plate off Taiwan's east coast, the energy is displaced from east to west. There are many fault lines in the Haulien-Taitung area. Moving west, the Central Mountain Range has been pushed out of the sea by the faults, but the old fault lines under it can no longer move, so the energy is transferred farther west, to where the mountains meet the western plains. This phenomenon is explained by what is called "the bulldozer theory." As a bulldozer pushes forward, what's behind it never moves, but the earth and rocks in front of it are constantly being upturned. In the Chichi earthquake, the energy from the colliding plates moved westward, but because the Central Mountain Range is very stable, the foothills west of the range in Nantou, Taichung and Miaoli gave way instead. And it's a fact that highly active faults often run through foothills.

But don't assume that the western plains are any safer. In the aftermath of the Chichi quake, people are dreading the destructive force of the Chelungpu fault. But examine a vertical section graphic of faults formed by the pressure from the movement of the plates. From east to west, you see the Lishan fault, the Shuangtung fault, the Chelungpu fault and the Changhua fault, which runs close to the Taiwan Strait. "This time the energy was displaced only as far as Chelungpu, but it's quite possible that in the future the stress will be released along the Changhua fault," says Lin Chin-wei, an associate professor in the Department of Earth Sciences at Chengkung University.

Yet people in Changhua needn't be too nervous. Geologists have a different sense of time from most people. Who knows what "in the future" means? More than a thousand years from now? More than ten thousand years from now? Geologists really don't have any idea. Some geologists, based on plate activity and stratum movement, have predicted that major quakes should occur more frequently in Taiwan than mainland China or Japan. "But unless we have detailed information about a fault, such as when it has been active in the past and whether there is any rhythm to its activity, then we're not going to be able to make any kind of educated guess about the timing of its next quake," says Shih Ruey-chyuan. Simply put, there is a lack of research about fault lines.

（face page） "Sincerity a hundred times over. Good fortune in myriad matters." The courtyard gate with this couplet is all that remains standing of what was a traditional style Chinese house in Puli. Only by respecting the power of Mother Nature and learning more about geology will people their homes.

The fault extends

Three years ago the CGS took action to rectify this situation, by starting new and intensive research into fault lines. Work the first year was devoted to research on three southern fault lines that concern geologists most. The second year it completed a report on four fault lines in Taipei. This year it is making a survey of fault lines in central Taiwan, including the Chelungpu fault line. No one had expected that not long after this work had begun, the Chelungpu fault would rock the island.

Because past surveys of Taiwan faults were mostly limited to examinations of surface geology, with no money to obtain more precise results by digging or drilling under the surface, major mistakes were common. Particularly in metropolitan areas where high real estate prices mean that all the land is covered by buildings, ruptures are impossible to see. After the Chichi quake, Lee Yuan-tseh, the president of the Academia Sinica, heard that a Taipei fault supposedly ran past the Academia Sinica, so he asked experts from the Institute of Earth Sciences to look for the fault both in front of and behind the Academia Sinica. Unable to see anything, they couldn't ascertain where the fault was running.

Now, in the wake of the disaster caused by the Chelungpu fault, the National Science Council has started carrying out its own research into fault lines in addition to the fault-line surveys being conducted by the CGS. But will appointing people to spend time and money researching faults really give us control over them?

After the Chichi quake rescue teams rushed to the scene of the disaster from across the island and around the world. Likewise geologists from north, south and central Taiwan mobilized to trace the fault. Yet these experts didn't all chart the fault lines the same way. Wang Chien-ying, a professor of earth sciences at National Central University, arrived at the scene only four hours after the quake. Five days later he came to the conclusion that the fault had bent, a notion that didn't mesh with the previous consensus that the Chelungpu fault directly connected to the Sanyi fault farther north. Separate groups of experts went to the area around Shihkangpa to investigate the turn in the fault. They discovered that there is no more apparent logic to how this fault changes course than to how hairs in an ink brush clump.

Faults can bend

After the Chichi earthquake, the government urgently wanted experts to chart the fault line accurately, so that it could delineate a surrounding zone in which building would be prohibited. But surface ruptures caused by a fault don't necessarily run in a straight line or fit in a regularly shaped area. Like a piece of glass, the earth along a fault doesn't shatter cleanly: the lines zigzag and the fragmentation is irregular.

Along the southern section of the fault, surface ruptures occurred within a belt ten or 20 meters across, whereas along the northern section of the fault, fractures branched off within a belt measuring as much as three kilometers across. In the provincial government's Chunghsing Village a new office building and an old office building toppled over and everything in between was completely undamaged. Lacking an understanding of faults' complexity didn't stop people from offering interpretations to match their political viewpoint. "Look," some would say, "even the gods aren't happy with the disbanding of the provincial government!" "Disbanding the provincial government wasn't enough," others argued. "See how the gods acted to destroy it."

"It is indeed very hard to draw an accurate and detailed map of geologic fault lines," explains Shih. Even if you know there's a fault line under your house, who can say that the next quake will strike there? And you can't always find clear surface evidence of an active fault. Without more extensive surveys, we may be able to say that a fault line is probably somewhere in a general area, but rarely can we say with any certainty that it's right there, under that road. While a fault can often be placed in a belt-like zone, to narrow its path down to one thin line is very difficult.

But no matter how hard it may be to trace a fault, you have to put the fault lines on a map, because there is a greater likelihood that geological stress will be released along already existing faults than anywhere else.

Even if an earthquake doesn't strike again in the near future, the earth around previously active faults is loose and the geological strata unstable, so it is best not to build there. The Council of Agriculture's Taiwan Animal Health Research Institute used to stand on the Hsinhua fault in Tainan. The director's office gradually sank and his nerves frayed until he finally couldn't take it any longer and went to geologists for help. When they told him the building was on a fault, he moved it. Laochuan Li in Mucha, Taipei City is the site of mudslides and rockslides whenever it rains. The CGS blames them on the Hsintien fault.

Nature isn't a still life

What's more, Taiwan is very windy and rainy, and if you don't mark out a fault line right after a quake, all evidence of it soon disappears. It is already impossible to determine the exact location of most stretches of the Hsinhua fault, which was created just this century. Did you hear the joke about the Meishan fault? When National Chung Cheng University was being planned, some experts said that the Meishan Fault ran under the site for the campus, and others said that it didn't. When construction was to begin, neither side had given an inch. The university administration decided to just go ahead and build and resolve the matter later. The disagreement made them think that earthquakes were worth studying, so they resolved to open an institute of seismology. If nothing else, they thought, perhaps it would serve to press down on the fault and keep it from moving.

In fact, scholars have been studying the Meishan fault for many years. "There is still disagreement as to whether it passes under Chung Cheng University," says Lin Chin-wei. Does this show that earth science doesn't offer easy answers?

"This isn't unusual in the least," says Wang Chien-ying, who confesses that science has proven rather powerless when confronted with the earth. "Can you rely on people's eyes to describe the enormity of the earth or the sky? The object of study can't fit inside a laboratory." It is said that three geologists asked the same question will come up with six different theories, each keeping one as a spare. This is because so much about nature is unfathomable.

Many people criticize the CGS fault map for being riddled with inaccuracies. The CGS's Chang Hui-cheng explains that geological activity has a life of its own, with different stages of development. What everyone sees are just the phenomena of a specific time and place. Fault ruptures may very well have been different last time than they are this time. The major earthquake that rocked Turkey in August involved the Anatolian fault in northwestern Turkey. A new rupture zone extends out from the middle of the fault in a northeasterly direction. But this isn't causing the scientific community to doubt the reliability of the original map. If people are well informed, they won't blindly expect the scientific community or government to provide them with a 100% accurate, unchanging map of fault lines.

No fault on this side!

The emphasis on fault maps overseas is also rooted in tragedy. After the San Francisco earthquake of 1906, scientists began paying attention to faults. But it wasn't until after the Los Angeles quake of 1971, which resulted in surface ruptures, that California passed the Alquist-Priolo Act, which mandated that all known faults be put on a map at a scale of 1:24,000. After the agencies performing the survey had completed their geologic fault maps, they were required to publicize the fact for ten days, and then give academics 90 days to make revisions, before releasing the maps to local zoning offices. After this long involved process, the best they could do was to designate a 400-foot-wide "fault zone." If the construction industry wants to build within this zone, it must hire professionals to undertake a more detailed survey to determine the precise location of the fault. Within 50 feet, only non-residential structures such as warehouses can be built.

California is currently one of the few places in the world that has adopted a systematic plan to deal with faults. Though Japan is greatly admired by Taiwan for its earthquake preparedness, its zoning regulations don't currently take fault lines into consideration.

Shih believes that California legislators demanded only 1:24,000 maps because they understood the practical difficulties in tracing faults, and only intended these maps to be used as a general reference. More important was the requirement that construction firms be responsible for detailed on-site surveys before starting to build. There's a joke among geologists about the Taiwan construction industry taking the CGS 1:100,000 scale map and blowing it up to a size large enough to suit their purposes. After being blown up, there is a margin of error of several kilometers, which allows builders in the southern side of a fault zone to say the fault is in the north, and those on the northern side to say it is in the south. Thus, builders on both sides can conclude that their own sites are exceptionally stable.

Building smart

The September quake has given geologists more ammunition in their campaign for legislation that would mandate nationwide collection of geological data and surveys of underground water resources, as well as strengthen geological education and conservation and require builders to obtain geological safety certificates before starting construction.

The Chelungpu fault became famous overnight after destroying more than 10,000 homes and causing great loss of life. But there's more that we can do than merely pray for the gods to take pity on us and keep the fault from moving again. At the very least, let us hope that while living in peace we can stop and think before we build. In addition to hiring an architect to design an attractive building and a structural engineer to assure a sturdy structure, let us go a step farther and show concern about the geological integrity of the sites where we build. Only by so doing can we prevent further tragedy.