Abstract Heat as a groundwater tracer has potential application in identifying earthquake-induced changes in fluid flow. Since convective heat transport modifies the phase and amplitude of periodic oscillations under pure heat conduction condition, temporal variation of groundwater flux can thus be derived from multi-depth periodic temperature signals. Here we present groundwater flux changes after two large earthquakes nearby, determined from in situ bedrock temperature measurements in the Xianshuihe fault zone, at the eastern margin of the Tibetan Plateau. Five-year temperature time series at six different locations are recorded by high sensitivity temperature sensors (10−4 K) installed at various depths ranging from 0 to 20 m. Groundwater fluxes are estimated through time-varying amplitudes and phases of the annual oscillations, extracted from the Dynamic Harmonic Regression technique. Results show different hydrological responses to these two earthquakes. After the 2013 Ms 7.0 Lushan earthquake, sustained rise in upward flux was revealed at all locations in the intermediate-field, contradicting to the modeled coseismic static dilatational strains. However, groundwater flux patterns after the 2014 Ms 6.3 Kangding earthquake are consistent with the coseismic volumetric strain distribution of a quadrantal pattern of this event. We propose that enhanced hydraulic permeability by dynamic stress might cause the increased upward fluid flux after the Lushan earthquake in the intermediate-field, while coseismic static volumetric strain may account for the hydrological changes after the Kangding earthquake in the near-field. This is the first time that temperature–time series are used to quantify continuous earthquake-related changes in groundwater fluxes. In situ temperature monitoring can provide insights into earthquake-driven hydrologic responses, and also offer additional information to identify temporal changes in crustal strain or hydraulic properties.

中文翻译：

---

使用长期基岩温度时间序列检测响应于两次大地震的地下水通量变化

摘要 作为地下水示踪剂的热量在识别地震引起的流体流动变化方面具有潜在的应用价值。由于对流热传输在纯热传导条件下改变了周期性振荡的相位和幅度，因此可以从多深度周期性温度信号导出地下水通量的时间变化。在这里，我们展示了附近两次大地震后地下水通量的变化，这是根据青藏高原东缘鲜水河断裂带的原位基岩温度测量确定的。六个不同位置的五年温度时间序列由安装在 0 到 20 m 范围内的不同深度的高灵敏度温度传感器 (10-4 K) 记录。地下水通量是通过年度振荡的时变振幅和相位来估计的，从动态谐波回归技术中提取。结果显示了对这两次地震的不同水文响应。在 2013 年 7.0 级庐山地震之后，中场所有位置都显示出向上通量的持续上升，这与模拟的同震静态膨胀应变相矛盾。然而，2014 年 6.3 级康定地震后的地下水通量模式与该事件象限模式的同震体积应变分布一致。我们认为，动应力增强的透水率可能导致中场芦山地震后向上流​​体通量的增加，而同震静态体积应变可能解释了近场康定地震后的水文变化。这是第一次使用温度-时间序列来量化地下水通量中与地震相关的连续变化。原位温度监测可以深入了解地震驱动的水文响应，还可以提供额外的信息来识别地壳应变或水力特性的时间变化。