Ping-Yu Chang

Electrical Resistivity Variations Before and After the Pingtung Earthquake in the Wushanting Mud Volcano Area in Southwestern Taiwan

The extensive eruption of fluid from the mud volcanoes in the Yanchao area of southwestern Taiwan reveals the activities of the active Chishan fault. A series of time-lapse resistivity imaging measurements were initially conducted in the Wushanting Natural Landscape Preservation Area in Yanchao to evaluate the relationship between resistivity change and fault activity. Resistivity measurements were conducted first along seven 30‐to60‐m survey lines to build up the regional model for mud volcanoes. We then conducted consecutive hourly and daily measurements to evaluate the short-term resistivity variations. Monthly observation was initiated along two 60‐m lines in July of 2006. On December 26, 2006, two successive earthquakes with magnitudes of 6.96 and 6.99 hit the town of Hengchun in Pingtung County, about 120km southwest of the monitoring site. Before the Pingtung earthquake, the resistivity at the research site was less than 25ohm‐m. Two regions with relatively low resistivity were found at a depth gre...

Using cross borehole ground penetrating radar attenuation tomography for characterizing soil properties in the vadose zone during a two-stage infiltration test

Summary Cross Borehole Ground Penetrating Radar is currently being used to monitor the change in moisture content and to test the feasibility of using attenuation tomography in order to monitor the conductive solute transport in the vadose zone at a test site in Socorro, New Mexico. The test site is a fluvial deposit containing alternating layers of sand and clay. Data collection began in December 1998 using five wells along an 11-meter profile intersecting a three-meter by three-meter infiltrometer. The infiltrometer has been discharging water at a constant rate of 2.7 cm/day since February 1999 in order to ensure the constant flux source. After an initial phase in isolating the specific pre-inversion processing steps necessary to produce the attenuation images, as well as determine the accuracy of this imaging method, our efforts have turned to analyzing the spatial distribution of the substance attenuation properties. The results show that two clay layers with a high attenuation coefficient are identified in the XBGPR images. Comparing the water content determined by velocity tomography, the attenuation increases for about 0.3 Neper/m during the first year of water infiltration, and shows a trend associated to the development of the wetted region. Currently, sodium chloride solution is being discharged through the infiltrometer at the test site, and frequent data collection is conducted to determine changes in attenuation coefficient causing by the saline solution.

A Graphics Processing Unit Implementation and Optimization for Parallel Double‐Difference Seismic Tomography

Abstract Double‐difference seismic tomography can estimate velocity structure and event locations with high precision, but its high‐computation cost along with large memory usage prevents the use of a personal computer to process very large datasets and requires a long‐computation time. This work proposes graphics‐processing‐unit‐ (GPU)‐based acceleration schemes to run the algorithm on a personal computer for very large datasets more efficiently. Generally, the algorithm can be divided into five major steps: input, ray tracing, matrix construction, inversion, and output. This work focuses on accelerating the ray‐tracing and inversion steps, which take almost two‐thirds of the computation time. Before ray tracing, our algorithm preprocesses the data by sorting all recorded event–station paths according to their lengths. Therefore, those path estimation jobs assigned to GPU cores are suitable for the GPU architecture. Furthermore, our work also minimizes the usage of global and local memory to reduce the GPU computing time needed to handle a very large dataset. In addition to parallelizing the inversion computation, our work proposes a GPU‐based elimination method to reduce redundant computation in inversion for further acceleration. In our test, the proposed acceleration schemes can gain maximum speed‐up factors of 31.17 and 35.46 for ray tracing and inversion, respectively, in our test. Overall, the GPU‐based implementation can reach a maximum of 5.98 times faster than the central processing unit‐based implementation.

Constructing the Apparent Geological Model by Fusing Surface Resistivity Survey and Borehole Records: Constructing the Apparent Geological Model by Fusing Surface Resistivity Survey and Borehole Records

We constructed an apparent geological model with resistivity data from surface resistivity surveys. We developed a data fusion approach by integrating dense electrical resistivity measurements collected with Schlumberger arrays and wellbore logs. This approach includes an optimization algorithm and a geostatistic interpolation method. We first generated an apparent formation factor model from the surface resistivity measurements and groundwater resistivity records with an inverse distance method. We then converted the model into a geology model with the optimized judgment criteria from the algorithms relating the apparent formation factors to the borehole geology. We also employed a non-parametric bootstrap method to analyze the uncertainty of the predicted sediment types, and the predictive uncertainties of clay, gravel, and sand were less than 5%. Overall, our model is capable of capturing the spatial features of the sediment types. More importantly, this approach can be arranged in a self-updated sequence to enable adjustments to the model to accommodate newly collected core records or geophysical data. This approach yields a more detailed apparent geological model for use in future groundwater simulations, which is of benefit to multi-discipline studies.

Study of diel hydrochemical variation in a volcanic watershed using principal component analysis: Tatun Volcano Group, North Taiwan

Heavy metal contamination commonly appears in mining areas and volcanic watersheds due to the acidic drainage water. Under the hydrochemical condition, the solar photocycle would result in changes of the water temperature, in photosynthesis and in iron photoreduction, which leads to substantial hydrochemical fluctuations, especially for heavy metals. It is important to consider the daily variations in water quality when developing a hydrochemical monitoring plan for an area with highly developed agriculture, such as the Tatun Volcano Group watershed area in this study. The results show that the water chemistry is highly complicated by both solar photocycles and hydrochemical fluctuations in the upstream area. Using principal component analysis, the contributions from the two factors can be successfully separated. During the daytime, the photocycle results in the formation of aluminum hydroxide, which can remove heavy metals from water. Consequently, the content of heavy metals, including As, Cu, Ni, Co and Ba, increases after sunset and can reach a maximum before sunrise, while Fe behaves inversely due to the photoreduction. The variation of As during a diel cycle can reach 97%. However, the content of most of the heavy metals during diel cycle is incomparable with those in the earlier studies due to the formation of aluminum hydroxide instead of iron hydroxide. The other significant factor, hydrochemical fluctuation, can explain the variation of major components in water including Cl, SO4. Rare earth elements (REEs) were also analyzed and can be an excellent natural tracer in this study. The distribution of REEs shows a depletion of light REEs and an even normalized concentration of middle and heavy REEs. It is theorized that the REEs in the water in this study derive mainly from the reservoir rock of geothermal water. In a hydrochemical monitoring plan, REEs can be an indicator for identifying an anthropogenic source.