Low-to-medium temperature geothermal fluids in the granite regions of southeastern China are an important renewable energy resource, but they are also a source of contamination containing highly toxic elements such as fluoride and arsenic. This study analyzed the origin of the geothermal fluids in a regional-scale hydrogeological unit in the city of Xiamen, China, based on isotope and hydrochemical analyses. The Br/Cl ratios suggested that the inland geothermal fluid is merely recharged by rainwater from the mountain edge, while the coastal geothermal fluid is originally recharged by the seawater and later mixed with rain-derived groundwater. The geothermal water featured high SiO2 and detectable Zn concentrations. The former reflects the significant water–granite interaction along the flow path, and the latter indicates the active hydraulic connection between surface waters, shallow aquifers and deep geothermal fluids. High radon content was detected near the deep conductive fault adjacent to a geothermal well, demonstrating that the fault damage zone acts as a major conduit for upward transport of the deep geothermal fluid. As a result, the fault damage zones developed in the granite are necessary for the formation of geothermal water, which leads to the uneven distribution of geothermal water in the subsurface. High-temperature geothermal water can be found in those regions with fairly sparse fault damage zones. In contrast, in the region with high-density fault activities, the active communication between shallow cool water and deep geothermal fluids can decrease the water temperature.
The enumerating algorithm has been introduced into the fitting procedure of the ASR model. Based on the detailed study of 21 large earthquakes with M≥6. 8 in the Chinese Mainland,the statistical features of seismic strain release before large earthquakes have been summarized. In the mass,the strain release models can be divided into five types. The first is the DA model,in which the strain release accelerates in broader areas and decelerates in small areas around the epicenter. Approximately 38% of earthquake samples are of this type. The second is the AD model,in which the strain release decelerates in broader areas and accelerates in smaller areas around the epicenter with an occupying ratioof approximately 19%. The third is ASR,in which only accelerating strain release can be observed. Cases of this model amount to about 14%. The fourth is DSR,in which only decelerating strain release can be checked,amounting to about 24%. There is only one earthquake sample of the fifth type (LSR),which shows a linear strain release. There is a 3~6 years difference in the duration of pre-shock sequences between the accelerating and decelerating models. This means that seismic quiescence against a background of increased seismicity of small earthquakes before large earthquakes are a typical feature in general. For the DA model,the average size of critical regions for steady accelerating and decelerating strain release is about 260km to 400km and 100km to 200km,respectively,3 to 5 times and 1 to 2 times the rupture size of an earthquake of magnitude 7. 0. The AD model is the opposite of the DA model. The model parameter,m value,has good stability. The ratio of ASR is about the same for accelerating seismic strain release phenomena,no matter what the strain release models are,or how large the strain release quantity is. With regard to decelerating seismic strain release phenomena, the DA model has the most distinctive decelerating strain release characteristic and is the typical feature of seismic strain release,i. e. "decelerating in-accelerating out seismic strain model". 相似文献