Introduction

As I write this in the dark in the van on the way to dinner, we bounce across the Chelungpu fault just south of Dali. We studied this site in some detail yesterday, Monday, measuring the height of a compound scarp formed by the Chi Chi earthquake at about 3.3 meters. This recent rupture was superposed upon a preexisting scarp of about the same height. As we ended the work day yesterday we reflected on how, in this case, it would have been possible to reduce earthquake losses significantly by avoiding building structures directly on the fault.

On Tuesday we surveyed the southern portion of the September Chelungpu fault rupture, carefully examined a spectacular building failure in Tai Chung, and then drove in to the epicentral area, including the towns of Chi Chi and Suli, which lie about 25 kilometers to the east of the Chelungpu fault trace. There we found the highest proportion of damaged structures, suggesting that either "basin resonance" or seismic wave "focusing" affected the ground motions experienced here. We saw no evidence of recent slip on the fault that lies immediately to the east, as we had read in some early reports, however we did see that buildings have been constructed very near where the fault is likely one day to rupture. We also saw huge numbers of landslides all over the steep mountainous terrain west of the Chelungpu fault trace. In this dispatch we'll discuss the TaiChung building failure, southern exposures of the Chelungpu fault, The spectacular MinZhu bridge failure, damage in Puli, and landslides.

Collapse at Puli

TaiChung Building Failure

The partially collapsed 14-story structure about 10 km west of the Chelungpu fault trace, taught us that the failure of a building may have a number of different contributing causes. Each by itself may not kill the building, but the combined effect is catastrophic. Even in hindsight, sorting through the rubble, it is extremely difficult to be able to identify which, if any, of the problems, if remedied, would have made the building strong enough to at least have saved the lives of those that perished.

The lowest floor of the eastern section of the building is currently collapsed, the western section intact, and the connecting middle section tilted. The building did not reach this state immediately, but was severely damaged by the mainshock, and then successive large aftershocks finally brought it to its current sad state. Many people died in this collapse, and apparently some bodies are not yet able to have been retrieved. At the site we saw evidence of poor construction, revealed by reinforcing iron rods not being wired together, and by the presence of space-filling (and expense saving) large cans within the concrete. We saw evidence of subsidence around the base of the building, which is consistent with local accounts that the building occupies partially reclaimed wetland.