Exclusive | Geologist Warns Taipei's Dormant Fault Risk Is Rising

A Taiwanese geologist who pioneered the use of carbon dating to map the island's earthquake history is warning that a fault line beneath the Taipei Basin, silent since a major quake more than three centuries ago, becomes more dangerous with every year it stays quiet.

Chen Wen-shan (陳文山), an emeritus professor of geosciences at National Taiwan University, raised the warning in an interview with Storm Media's New News magazine following renewed earthquake activity in Kumamoto, Japan, that reignited public concern in both Taiwan and Japan. Chen taught at the university from 1981 until his retirement in 2022 and is credited in Taiwanese academic circles as the "father of fault dating" for introducing carbon-14 analysis to determine when local faults last ruptured. Over four decades of fieldwork, he has personally collected rock samples across Taiwan's mountains and valleys, joking that he has "practically touched every fault in Taiwan."

Renewed tremors in Kumamoto led many observers to ask why the same area kept shaking after a major quake there several years earlier. Chen said the confusion comes from a basic misunderstanding of how seismologists think about risk.

"People are used to thinking about earthquakes by city name, but geologists look at fault lines," Chen said. Earthquakes are tied to individual faults, not to administrative boundaries, he explained. Taiwan has more than 30 to 40 active faults, while Japan has upwards of 100. Each behaves as an independent system that accumulates its own stress over time, so when one fault ruptures and then a separate fault beneath the same city moves later, residents perceive it as repeated earthquakes in one place.

Chen also pushed back on a popular but scientifically unfounded idea that frequent minor earthquakes help prevent bigger ones. Each step up in magnitude, he noted, represents roughly a 33-fold increase in released energy, so the gap between a magnitude 5 and a magnitude 7 event is more than a thousandfold. "Many people believe that frequent magnitude 5 earthquakes release energy and help prevent a major earthquake. Scientifically, that is simply not true," he said, adding that minor quakes release only a negligible fraction of the energy behind a magnitude 7 or larger event.

Taiwan and Japan both sit along convergent plate boundaries, Chen said, but Taiwan's is shaped mainly by direct plate collision, while Japan's is driven more by subduction, including along the Nankai Trough. Both processes compress and fracture the crust into numerous faults. Kumamoto's active fault system, which is divided into several segments, stretches roughly 80 kilometers in total, he said, and a previous quake released stress along only the northern 30-kilometer segment, leaving about 50 kilometers to the south still under strain and worth continued monitoring.

Chen also addressed the geological reasoning behind Taiwan Semiconductor Manufacturing Co.'s decision to build a chip plant in Kumamoto, saying he views the site selection favorably from a geological standpoint. The plant sits near a fault segment that already ruptured in 2016 and released its accumulated stress, he said.

Semiconductor fabrication plants typically operate on five- to ten-year technology upgrade cycles, while the recurrence interval for major earthquakes on a given fault can often run hundreds of years, and sometimes more than a thousand. "Looking at the time scale of an industrial facility's life span against geological cycles, siting a plant in an area that just released its stress makes good sense for risk management," Chen said.

That logic extends to a broader point Chen considers widely overlooked: a fault that has just ruptured is, for a period, the safest place to be. He pointed to the Chelungpu Fault, which unleashed the magnitude 7.3 Chi-Chi earthquake in 1999 after building up stress for what researchers estimate was a 300- to 400-year cycle. "This means the probability of that fault rupturing again at the same magnitude within the next few decades, or even the next hundred years, is extremely low," Chen said, noting that human lifespans and building design lifespans both fall well within that safer window.

The faults that deserve the closest scrutiny, Chen said, are the ones that have gone quiet the longest without a clear sense of their cycle. Chen was referring to the Shanchiao Fault, an active normal fault running along the western edge of the Taipei Basin that Taiwan's Central Geological Survey classifies as a Category 2 hazard. The name literally means "foot of the mountain," describing where the fault meets the basin's edge. Researchers have linked it, with varying degrees of confidence, to a major earthquake or ground-subsidence event around 1694. The connection rests largely on Yu Yonghe's (郁永河) 1697 travel account *A Small Sea Journey*, which recorded local reports that the shaking and the formation of what became known as Kangxi Taipei Lake had occurred roughly three years earlier. Scientists treat the specific attribution to the Shanchiao Fault as an inference rather than established fact, leaving the fault's true recurrence interval uncertain.

The difficulty, Chen said, is the shortness of Taiwan's written record, which covers only three to four hundred years. That leaves scientists unable to say with confidence whether the fault's true recurrence interval is closer to 300 years or 500 years. "Our biggest challenge with this fault is that we simply lack enough of a historical record," he said.

Filling that gap would require paleoseismology work, such as digging trenches across the fault or drilling deep core samples to date sediment layers, Chen said. But Taipei's dense development and high-rise construction make finding a workable excavation site extremely difficult. He was careful to note that even with more data, science cannot pinpoint exactly when a major earthquake will strike. What geologists do know, he said, is that the longer a fault goes without rupturing, the higher its annual probability of doing so becomes.

Rather than fuel public anxiety, Chen argued the more productive response is investing in structural engineering and speeding up redevelopment of Taiwan's aging buildings, describing it plainly as making sure buildings are properly built. On the widely feared risk of soil liquefaction, he said the phenomenon is typically a shallow-ground issue. In areas such as sections of the Keelung River that were straightened and rezoned districts built on the resulting land, buildings whose foundations are piled deep enough to pierce weak surface soil and anchor into bedrock, the method used for Taipei 101, can remain structurally sound even if the ground around them liquefies.

The bigger vulnerability, Chen said, lies in Taipei's stock of aging walk-up apartments, many of them 30 to 40 years old or more and built to seismic codes far weaker than current standards. He said many urban renewal projects stall for years because a small number of holdout owners cannot reach consensus on a rebuild, and some reconstruction cases involving buildings already flagged as structurally hazardous have been stuck for decades. Residents of those older buildings, he stressed, carry the greatest risk when a major earthquake hits.

Chen called on the government to apply stronger policy and legal tools to accelerate the reconstruction and reinforcement of old housing stock, calling it the most realistic way to protect residents before disaster strikes. "Modern technology can already predict a typhoon's path, but when it comes to earthquakes, nobody truly knows when one will strike," he said.

Original Article in Chinese

（Related： Colombia Earthquake Death Toll Climbs to 132 as UN Joins Response ｜ Latest ）

​