Geological causes and geomorphological precursors of the Tsaoling landslide triggered by the 1999 Chi-Chi earthquake, Taiwan

The Tsaoling landslide, one of the largest landslide areas in Taiwan, has been affected by catastrophic events triggered by rain or earthquakes six times since 1862. These landslides, including that caused by the 1999 earthquake, have essentially not been reactivated old slides, but were sequential new ones that developed upslope, retrogressively. The landslide area is underlain by Pliocene sandstone and shale to form a dip slope with a bedding plane, dipping uniformly at 14°. The slip surface of the 1999 landslide was smooth and planar, parallel to the bedding plane with a slightly stepped profile; it formed within thinly alternated beds of fine sandstone and shale with ripple lamination or in a shale bed. The shale is weathered by slaking and probably by sulfuric acid, which is inferred to be one of the major causes of the intermittent retrogressive development of the landslides. The weathering was likely accelerated by the removal of overlying beds during earlier landslides in 1941 and 1942. The top margin of the 1999 landslide, in plan view, coincided with a V-shaped scarplet, which can be clearly recognized on aerial photographs taken before the landslide. This geomorphological feature indicates that this landslide had already moved slightly before its 1999 occurrence, providing precursory evidences.     

Significant resistivity changes in the fault zone associated with the 1999 Chi-Chi earthquake, west-central Taiwan

The purpose of this paper is to demonstrate the possible influence of an earthquake on the resistivity distribution in a fault zone. We collected resistivity image profiles across a proposed fault trace prior to the 7.3 magnitude 1999 shallow earthquake in the Chi-Chi area of Taiwan. Significant positive resistivity anomalies were observed in the hanging wall after the earthquake. However, there were negligible resistivity changes in the footwall. From an examination of geoelectric phenomena and surface rupture in the fault zone, it is believed that geoelectric anomalies are associated with abrupt displacement along the active Chelungpu fault. This result indicates a potential for resistivity methods to provide a basis for the monitoring of an active fault.     

Seismic reversal pattern for the 1999 Chi-Chi, Taiwan, MW 7.6 earthquake

We in this study have calculated the standard normal deviate Z-value to investigate the variations in seismicity patterns in the Taiwan region before and after the Chi-Chi earthquake. We have found that the areas with relatively high seismicity in the eastern Taiwan became abnormally quiet before the Chi-Chi earthquake while the area in the central Taiwan with relatively low seismicity showed unusually active. Such a spatially changing pattern in seismicity strikingly demonstrates the phenomenon of “seismic reversal,” and we here thus present a complete, representative cycle of “seismic reversal” embedding in the changes of seismicity patterns before and after the Chi-Chi earthquake.     

Seismic behavior in central Taiwan: Response to stress evolution following the 1999 Mw 7.6 Chi-Chi earthquake

Following the 1999 M_w 7.6 Chi-Chi earthquake, a large amount of seismicity occurred in the Nantou region of central Taiwan. Among the seismic activities, eight M_w ⩾ 5.8 earthquakes took place following the Chi-Chi earthquake, whereas only four earthquakes with comparable magnitudes took place from 1900 to 1998. Since the seismicity rate during the Chi-Chi postseismic period has never returned to the background level, such seismicity activation cannot simply be attributed to modified Omori’s Law decay. In this work, we attempted to associate seismic activities with stress evolution. Based on our work, it appears that the spatial distribution of the consequent seismicity can be associated with increasing coseismic stress. On the contrary, the stress changes imparted by the afterslip; lower crust–upper mantle viscoelastic relaxation; and sequent events resulted in a stress drop in most of the study region. Understanding seismogenic mechanisms in terms of stress evolution would be beneficial to seismic hazard mitigation.     

Characterization of slip zone associated with the 1999 Taiwan Chi-Chi earthquake: X-ray CT image analyses and microstructural observations of the Taiwan Chelungpu fault

To characterize the fault-related rocks within the Chelungpu fault, we performed X-ray computed tomography (CT) image analyses and microstructural observations of Hole B core samples from the Taiwan Chelungpu-fault Drilling Project. We identified the slip zone associated with the 1999 Chi-Chi earthquake, within the black gouge zone in the shallowest major fault zone, by comparison with previous reports. The slip zone was characterized by low CT number, cataclastic (or ultracataclastic) texture, and high possibility to have experienced a mechanically fluidized state. Taking these characteristics and previous reports of frictional heating in the slip zone into consideration, we suggested that thermal pressurization was the most likely dynamic weakening mechanism during the earthquake.     

Why does the co-seismic slip of the 1999 Chi-Chi (Taiwan) earthquake increase progressively northwestward on the plane of rupture?

The Chi-Chi 1999 earthquake ruptured the out-of-sequence Chelungpu Thrust Fault (CTF) in the fold-and-thrust belt in Western Central Taiwan. An important feature of this rupture is that the calculated slip increases approximately linearly in the SE–NW convergence plate direction from very little at its deeper edge to a maximum near the surface. We propose here a new explanation for this co-seismic slip distribution based on the study of both stress and displacement over the long-term as well as over a seismic cycle. Over the last 0.5 My, the convergence rate in the mountain front belt is accommodated by the frontal Changhua Fault (Ch.F), the CTF and the Shuangtung Fault (Sh.F). Based on previously published balanced cross sections, we estimate that the long-term slip of the Ch.F and of the CTF accommodate 5–30% and 30–55% of the convergence rate, respectively. This long-term partitioning of the convergence rate and the modeling of inter-seismic and post-seismic displacements suggest that the peculiar linear co-seismic slip distribution is accounted for by a combination of the effect of the obliquity of the CTF to the direction of inter-seismic loading, and of increasing aseismic creep on the deeper part of the Ch.F and CTF. Many previous interpretations of this slip distribution have been done including the effects of material properties, lubrication, site effect, fault geometry and dynamic waves. The importance of these processes with respect to the effects proposed here is still unknown. Taking into account the dip angle of the CTF, asperity dynamic models have been proposed to explain the general features of co-seismic slip distribution. In particular, recent works show the importance of heterogeneous spatial distribution of stress prior to the Chi-Chi earthquake. Our analysis of seismicity shows that previous large historic earthquakes cannot explain the amplitude of this heterogeneity. Based on our approach, we rather think that the high stress in the northern part of the CTF proposed by Oglesby and Day [Oglesby, D.D., Day, S.M., 2001. Fault geometry and the dynamics of the 1999 Chi-Chi (Taiwan) earthquake. Bull. Seismol. Soc. Am. 91, 1099–1111] reflects the latitudinal variation of inter-seismic coupling due to the obliquity of the CTF.     

Electrical structures of the source area of the 1999 Chi-Chi, Taiwan, earthquake: Spatial correlation between crustal conductors and aftershocks

Wide-band magnetotelluric (MT) data were collected on an east–west profile, approximately perpendicular to the local strike of the Chelungpu thrust, through the hypocentral area of the Chi-Chi earthquake for imaging the seismogenic structure. MT data were then inverted for two-dimensional resistivity models plus best-fitting static shift parameters using a nonlinear conjugate gradient algorithm that minimizes the sum of the normalized data misfits and the smoothness of the model. As shown in the inverted 2D resistivity models presented in this paper, an electrical conductor beside the hypocenter of the Chi-Chi earthquake indicates that deep-crustal fluids may participate in the rupture process of the Chi-Chi earthquake. A striking spatial correlation between the crustal conductor and occurrence of aftershocks beneath the Chelungpu fault suggests a postseismic pore pressure adjustment ongoing after the mainshock. Additionally, the hypocenter exhibits an electrical resistive zone, consistent very well with a predicted compact zone from a crustal deformation and transient fluid flow modeling.     

Co-seismic displacements, folding and shortening structures along the Chelungpu surface rupture zone occurred during the 1999 Chi-Chi (Taiwan) earthquake

A nearly 100-km-long surface rupture zone, called Chelungpu surface rupture zone, occurred mostly along the pre-existing Chelungpu fault on the northwestern side of Taiwan, accompanying the 1999 Chi-Chi Ms 7.6 earthquake. The Chelungpu surface rupture zone can be divided into four segments based on the characteristics of co-seismic displacements, geometry of the surface ruptures and geological structures. These segments generally show a right-step en echelon form and strike NE–SW to N–S, and dip to the east with angles ranging from 50 to 85°. The co-seismic flexural-slip folding structures commonly occurred in or near the surface rupture zone from a few meters to a few hundreds of meters in width, which have an orientation in fold axes parallel or oblique to the surface rupture zone. The displacements measured in the southern three segments are approximately 1.0–3.0 m horizontally and 2.0–4.0 m vertically. The largest displacements were measured in the northern segment, 11.1 m horizontally and 7.5 m vertically, respectively. The amount of co-seismic horizontal shortening caused by flexural-slip folding and reverse faulting in the surface rupture zone is generally less than 3 m. It is evident that the co-seismic displacements of the surface rupture zone are a quantitative surface indicator of the faulting process in the earthquake source fault. The relations between the geometry and geomorphology of the surface rupture zone, dips of the co-seismic faulting planes and the striations on the main fault planes generated during the co-seismic displacement, show that the Chelungpu surface rupture zone is a reverse fault zone with a large left-lateral component.     

Landslide hazards triggered by the 2008 Wenchuan earthquake, Sichuan, China

The 2008 Wenchuan earthquake (M _s = 8.0; epicenter located at 31.0° N, 103.4° E), with a focal depth of 19.0 km was triggered by the reactivation of the Longmenshan fault in Wenchuan County, Sichuan Province, China on 12 May 2008. This earthquake directly caused more than 15,000 geohazards in the form of landslides, rockfalls, and debris flows which resulted in about 20,000 deaths. It also caused more than 10,000 potential geohazard sites, especially for rockfalls, reflecting the susceptibility of high and steep slopes in mountainous areas affected by the earthquake. Landslide occurrence on mountain ridges and peaks indicated that seismic shaking was amplified by mountainous topography. Thirty-three of the high-risk landslide lakes with landslide dam heights greater than 10 m were classified into four levels: extremely high risk, high risk, medium risk, and low risk. The levels were created by comprehensively analyzing the capacity of landslide lakes, the height of landslide dams, and the composition and structure of materials that blocked rivers. In the epicenter area which was 300 km long and 10 km wide along the main seismic fault, there were lots of landslides triggered by the earthquake, and these landslides have a common characteristic of a discontinuous but flat sliding surface. The failure surfaces can be classified into the following three types based on their overall shape: concave, convex, and terraced. Field evidences illustrated that the vertical component of ground shaking had a significant effect on both building collapse and landslide generation. The ground motion records show that the vertical acceleration is greater than the horizontal, and the acceleration must be larger than 1.0 g in some parts along the main seismic fault. Two landslides are discussed as high speed and long runout cases. One is the Chengxi landslide in Beichuan County, and the other is the Donghekou landslide in Qingchuan County. In each case, the runout process and its impact on people and property were analyzed. The Chengxi landslide killed 1,600 people and destroyed numerous houses. The Donghekou landslide is a complex landslide–debris flow with a long runout. The debris flow scoured the bank of the Qingjiang River for a length of 2,400 m and subsequently formed a landslide dam. This landslide buried seven villages and killed more than 400 people.     

Statistical approach to earthquake-induced landslide susceptibility

Susceptibility analysis for predicting earthquake-induced landslides has most frequently been done using deterministic methods; multivariate statistical methods have not previously been applied. In this study, however, we introduce a statistical methodology that uses the intensity of earthquake shaking as a landslide triggering factor. This methodology is applied in a study of shallow earthquake-induced landslides in central western Taiwan. The results show that we can accurately interpret landslide distribution in the study area and predict the occurrence of landslides in neighboring regions. This susceptibility model is capable of predicting shallow landslides induced during an earthquake scenario with similar range of ground shaking, without requiring the use of geotechnical, groundwater or failure depth data.     

Topographical changes revealed by high-resolution airborne LiDAR data: The 1999 Tsaoling landslide induced by the Chi–Chi earthquake

The 1999 Chi–Chi earthquake triggered the catastrophic Tsaoling landslide in central Taiwan. We mapped the landslide area and estimated the landslide volume, using a high-resolution digital elevation model from airborne LiDAR (Light Detection And Ranging), aerial photographs and topographic maps. The comparison between scar and deposit volumes, about 0.126 km³ and 0.150 km³ respectively, suggests a coseismic volume increase of 19% due to decompaction during landsliding. In July 2003, the scar and deposit volumes were about 0.125 km³ and 0.110 km³ respectively. These estimates suggest that 4 years after the event, the volume of landslide debris removed by river erosion was nearly 0.040 km³. These determinations are confirmed by direct comparison between the most accurate topographic models of the post-landslide period, indicating a very high erosion rate at the local scale (0.01 km³/year) for the deposit area of the landslide. Such a large value highlights the importance of landslide processes for erosion and long-term denudation in the Taiwan mountain belt.     

Interpretation of landslide distribution triggered by the 2005 Northern Pakistan earthquake using SPOT 5 imagery

The 2005 northern Pakistan earthquake (magnitude 7.6) of 8 October 2005 occurred in the northwestern part of the Himalayas. We interpreted landslides triggered by the earthquake using black-and-white 2.5-m-resolution System Pour l’Observation de la Terre 5 (SPOT 5) stereo images. As a result, the counts of 2,424 landslides were identified in the study area of 55 by 51 km. About 79% or 1,925 of the landslides were small (less than 0.5 ha in area), whereas 207 of the landslides (about 9%) were large (1 ha and more in area). Judging from our field survey, most of the small landslides are shallow rock falls and slides. However, the resolution and whitish image in the photos prevented interpreting the movement type and geomorphologic features of the landslide sites in detail. It is known that this earthquake took place along preexisting active reverse faults. The landslide distribution was mapped and superimposed on the crustal deformation detected by the environmental satellite/synthetic aperture radar (SAR) data, active faults map, geological map, and shuttle radar topography mission data. The landslide distribution showed the following characteristics: (1) Most of the landslides occurred on the hanging-wall side of the Balakot–Garhi fault; (2) greater than one third of the landslides occurred within 1 km from the active fault; (3) the greatest number of landslides (1,147 counts), landslide density (3.2 counts/km²), and landslide area ratio (2.3 ha/km²) was found within Miocene sandstone and siltstone, Precambrian schist and quartzite, and Eocene and Paleocene limestone and shale, respectively; (4) there was a slight trend that large landslides occurred on vertically convex slopes rather than on concave slopes; furthermore, large landslides occurred on steeper (30° and more) slopes than on gentler slopes; (5) many large landslides occurred on slopes facing S and SW directions, which is consistent with SAR-detected horizontal dominant direction of crustal deformation on the hanging wall.     

汶川地震触发崩滑地质灾害空间分布及影响因素

通过遥感解译和实地考察,获取了2008年汶川地震触发崩滑的空间分布,利用GIS空间分析和Lo-gistic回归,分析崩滑的空间分布特征及其影响因素,建立了地震触发崩滑与其影响因素之间的回归方程。结果表明,(1)研究区共有5 154个崩滑群,覆盖总面积1 139 km2;(2)崩滑沿北川—映秀发震断层的两侧(断层上盘区占90%),呈北东向宽度不一的条带状分布;(3)Ⅺ和Ⅹ烈度区崩滑面积占区域面积的73.2%,Ⅷ度及以下烈度区崩滑面积比例较小;(4)崩滑发育及空间分布不仅受控于发震断层的活动,断层上下盘效应、地形放大效应等也是其重要影响因素。崩滑与其影响因子的回归方程表明:(1)到北川—映秀发震断层距离因子和到震中距离因子的偏回归系数远大于其他因子的偏回归系数,北川—映秀断层发震活动是控制崩滑空间分布的主导因子;(2)岩性软硬程度对崩滑空间分布的影响不显著;(3)地形坡度、高程、坡度变率、多年累积降雨、人工修路及植被覆盖对崩滑的发育产生影响。地形高程因子对崩滑空间分布的影响大于坡度、坡度变率因子的影响。人工道路、多年降雨及植被覆盖对地震崩滑的影响程度依次降低。     

Response changes of some wells in the mainland subsurface fluid monitoring network of China, due to the September 21, 1999, Ms7.6 Chi-Chi Earthquake

About 60 hydrologic changes in response to the Chi-Chi earthquake with Ms7.6 on September 21, 1999, occurred in 52 wells, including groundwater level, temperature, discharge rate, well pressure and radon, etc., in the subsurface fluid monitoring network. These response changes were mainly co-seismic, but some pre- and post-earthquake changes occurred mainly within 5 days before and after the Chi-Chi earthquake. The response changes of different wells clustering in different tectonic areas showed different features. These changes are distributed in five areas named as A, B, C, D and E. The response changes in A area with short hypo-central distance (less than 550 km) were mainly pre-earthquake changes occurring more than 5 days before the event. Those in area B (in Huanan tectonic block) and C (in Huabei tectonic block) were mainly co-seismic changes. The hypo-central distance is about 1100–1280 and 800–1160 km, respectively. These changes were high-frequency water-level oscillations induced by seismic waves and accompanied by prominent and permanent water-level jumps and drops. There are also some post-seismic changes including discharge rate and water radon and well pressure changes in area C. Those in area D in the Yanshan tectonic block were mainly co-seismic and post-seismic changes including water level, water temperature, and water radon concentration, etc., showing prominent and permanent water-level jumps and drops and rising concentrations of water radon. The hypo-central distance is about 1750–2060 km. Those in Area E were mainly co-seismic changes showing prominent and permanent water-level jump. The hypo-central distance is about 1810–2120 km. Three moderate earthquakes occurred in area D and one strong earthquake occurred in area E 4 months after the Chi-Chi earthquake. The different features of the response changes might be caused by the changes of local hydrologic conditions (like permeability) induced by seismic waves. On the other hand, these response changes might indicate the near-critical conditions in the area where the response changes clustered. Such changes might be understood by the crustal buckling hypothesis. It is thought that the response changes might be a kind of precursor that implies elevated earthquake risk in the region.     

A study on the identification of liquefaction-induced failures on ground surface based on the data from the 1999 Kocaeli and Chi-Chi earthquakes

One of the major causes of earthquake damage is liquefaction. However, it doesn't result in severe harm unless it leads to ground surface damage or ground failure. Therefore, prediction of potential for ground surface damage due to liquefaction is one of the important issues in microzonation studies for liquefaction-induced damage in areas with high seismicity. In 1985, based on a database compiled from Chinese and Japanese earthquakes, Ishihara considered the influence of the non-liquefied cap soil on the occurrence or non-occurrence of ground failure (mainly sand boiling), and proposed an empirical approach to predict the potential for ground surface damage at sites susceptible to liquefaction. However, some investigators indicated that this approach is not generally valid for sites susceptible to lateral spread or ground oscillation. In this study, a contribution to improve the approach by Ishihara is made. For the purpose, an index called liquefaction severity index (LSI) and data from two devastating earthquakes, which occurred in Turkey and Taiwan in 1999, were employed. The data from liquefied and non-liquefied sites were grouped and then analysed. Based on the observations reported by reconnaissance teams who visited both earthquake sites and the results of the liquefaction potential analyses using the filed-performance data, a chart to assess the potential for ground surface disruption at liquefaction-prone areas was produced. The analyses suggest that the procedure proposed by Ishihara is quite effective particularly for the occurrence of sand boils, while the bounds suggested in this method generally may not be valid for the prediction of liquefaction-induced ground surface disruption at sites susceptible to lateral spreading. The chart proposed in this study shows an improvement over the Ishihara's approach for predicting the liquefaction-induced ground surface damage. The microzonation maps comparing the liquefaction sites observed along the southern shore of Izmit Bay and in Yuanlin, and the surface damage and non-damage zones predicted from the proposed chart can identify accurately the liquefaction (sand boiling and lateral spreading) and no-liquefaction sites.     

Analysis of an anti-dip landslide triggered by the 2008 Wenchuan earthquake in China

The Guantan landslide, with a total displaced mass of about 468 × 10⁴ m³, was triggered by the 2008 Wenchuan earthquake and succeeding rainfall in Jushui Town, Sichuan Province, China. The landslide occurred on an anti-dip hard rock slope with a weak rock founding stratum of 200 m in thickness. To investigate the failure mechanism of the Guantan landslide, dynamic behaviors of hard and soft rock slopes were investigated by means of large scale shaking table tests. The laboratory models attempted to simulate the field geological conditions of the Guantan landslide. Sinusoidal waves and actual seismic waves measured from the Wenchuan Earthquake were applied on the slope models under 37 loading configurations. The experimental results indicated that deformation mainly developed at a shallow depth in the upper part of the hard rock slope and in the upper (near the crest) and lower (near the toe) parts of the soft rock slope. An equation for predicting the depth of sliding plane was proposed based on the location of the maximum horizontal acceleration. Finally, it was concluded that the failure process of the Guantan landslide occurred in three stages: (1) toppling failure caused by compression of the underlying soft rock strata, (2) formation of crushed hard rock and sliding surface in soft rock as the result of seismic shocks, particularly in the horizontal direction, and (3) aftershock rainfall accelerates the process of mass movement along the sliding plane.     

汶川大地震诱发滑坡的震中距

本文在地震滑坡规模(体积)类型5级划分法的基础上,采用数学方法对汶川地震灾区的2个宏观震中(映秀、北川)烈度I0≥XI极震区(10个县市)的地震滑坡震中距进行了讨论。其中地震滑坡随震中距分布基本符合正弦、高斯和指数衰减规律。极震区最大滑坡震中距可达120km,最小滑坡震中距可达0.25km。结果表明,地震诱发滑坡的规模(体积)和分布范围都与震中距具有相关性。     

Interpretation of probability of landsliding triggered by rainfall

This paper is a contribution to an important aspect of the systematic and quantitative assessment of landslide hazard and risk. The focus is on site-specific and detailed assessment for rainfall-triggered landslides and, in particular, on the estimation and interpretation of the temporal probability of landsliding. Historical rainfall data over a 109-year period were analysed with particular reference to a site along the Unanderra and Moss Vale Railway Line in the State of New South Wales, Australia. It is shown that the recurrence interval of landsliding and hence annual probability of occurrence is subject to significant uncertainty and that it cannot be regarded as a constant. Accordingly landslide hazard varies spatially as well as being a function of time. For the example case study considered in this paper the annual probability of landslide occurrence was estimated to be in the range 0.026–0.172. However, the mean annual probability of landslide reactivation was estimated to be in the range 0.037–0.078. Utilisation of methods for probability assessment proposed in this paper will contribute to more realistic assessment of hazard and risk and, therefore, to more efficient risk management.     

Landslides triggered by the 1949 Khait earthquake, Tajikistan, and associated loss of life

Earthquake-triggered landslides are a major geological hazard in Central Asia. In July 1949, the M7.4 Khait earthquake triggered many hundreds of landslides in a mountainous region near the southern limit of the Tien Shan Mountains, central Tajikistan. These landslides involved widespread rock-slope failure as well as large numbers of flowslides in loess that mantles the steep slopes of the region. In the Yasman valley hundreds of loess landslides coalesced to form a massive loess flow (est. vol. 245 Mm³) that travelled up to 20 km on a slope of only 2°. In an adjacent valley, the Khait landslide involved transformation of an earthquake-triggered rockslide into a very rapid flow by the entrainment of saturated loess into its movement. It travelled 7.41 km over a vertical distance of 1421 m with an estimated average velocity of ~30 m/s. We estimate its volume as 75 Mm³, an order of magnitude less that previously published estimates. The Khait landslide was simulated using DAN. The number of casualties due to earthquake-triggered landslides in the epicentral region was considerable. Approximately 4000 people were killed in the Yasman valley loess flow as 20 villages (kishlaks) were overwhelmed. In the Khait landslide alone we estimate ca. 800 people lost their lives when the villages of Khait and Khisorak were overrun by rapidly moving debris. Our data indicates that a total of approximately 7200 people were killed by earthquake-triggered landslides in the epicentral region of the Khait earthquake and that, in terms of loss of life, the 1949 Yasman valley loess flow was one of the most destructive landslides in recent history.     

Impacts of September 21, 1999 Chi-Chi earthquake on the characteristics of gully-type debris flows in central Taiwan

Debris flows are more frequent in central Taiwan, because of its mountainous geography. For example, many debris flows were induced by Typhoon Herb in 1996. The Chi-Chi earthquake with a magnitude of 7.3, which took place in 1999 in central Taiwan, induced many landslides in this region. Some landslides turned into debris flows when Typhoon Toraji struck Taiwan in 2001. This study investigates the characteristics of the gullies where debris flows have occurred for a comparison. Aerial photos of these regions dated in 1997 (before the earthquake) and 2001 (after the earthquake) are used to identify the occurrence of gully-type debris flows. A Geographic Information System (GIS) is applied to acquire hydrological and geomorphic characteristics: stream gradient, stream length, catchment gradient, catchment area, form factor, and geology unit of these gullies. These characteristics in different study regions are presented in a statistical approach. The study of how strong ground motion affects the debris flows occurrence is conducted. The characteristics of the debris flow gullies triggered by typhoons before and after the Chi-Chi earthquake are quantitatively compared. The analysis results show that a significant transformation in the characteristics was induced by the Chi-Chi earthquake. In general, the transformation points out a lower hydrological and geomorphic threshold to trigger debris flows after the Chi-Chi earthquake. The susceptibility of rock units to strong ground motion is also examined. The analysis of debris flow density and accumulated rainfall in regions of different ground motion also reveal that the rainfall threshold decreases after the Chi-Chi earthquake.