Abstract: The theoretical and in situ investigations in recent years indicated that the dynamic change information of crustal stress can be obtained through the observation of bedrock temperature. In particular, the magnitude and spatial distribution characteristics of the co-seismic stress variation obtained based on the co-seismic temperature response of the Kangding M_S6.3 earthquake were consistent with the results obtained by seismology, which confirms the validity of the field analysis of the co-seismic stress variation by temperature observation. In the future, with the further improvement of temperature measurement technology, this method for detecting dynamic change in crustal stress through bedrock temperature is expected to bring new opportunities for earthquake science.
However, the stress changes caused by earthquakes are related to the distance from the measurement point to the source and decay rapidly with the increase of the distance, and they are also related to magnitude and decrease exponentially with magnitude. For example, the co-seismic temperature response observed in the Kangding M_S6.3 earthquake did not appear in the subsequent M_S5.8 earthquake. One key reason is that the detection of co-seismic stress changes in M_S5.8 earthquake requires a precision of temperature measurement system to be up to 0.01mK. But, the precision of the instruments used at that time was about 0.2mK, which could not detect temperature changes in the order of magnitude of 0.01mK. So, in order to make the above-mentioned method play a greater role in earthquake science, especially in detecting the stress variation information before a strong earthquake, it is urgent to develop a more accurate temperature measurement system.
In this paper, a new version of high-precision temperature measurement system is developed successfully, after considering a series of technical improvements such as constant current commutation driving and multiple Kalman digital filtering, based on the low temperature drift fixed value resistor and the temperature measuring resistor with a balanced bridge of four-wire temperature sensors. The new system has a designed temperature resolution of 0.003mK. According to in situ tests, the precision of temperature measurement reaches 0.03mK. In addition, field observations have confirmed the feasibility of the temperature measurement system. Based on the relation between stress change and temperature response, a dynamic stress change of the magnitude of 0.03MPa can be obtained, which implies that the observable range of thermal stress can be greatly improved with comparison to the previous version of 0.2MPa. It should be emphasized for earthquake research that co-seismic Coulomb stress is an importance parameter, whose magnitude focuses on a range of about 0.01~0.1MPa. Up to now, co-seismic Coulomb stress change cannot be measured directly(at least partially)because of limitation of detecting method. Since the ability of measuring dynamic stress by our system developed in this paper has reached to 0.03MPa, reaching the magnitude of the co-seismic Coulomb stress change, the potential application of the system is to explore the change of the co-seismic Coulomb stress. This is of great benefit to promote the development of ways of observing the dynamic changes in crustal stress through bedrock temperature.

Key words: stress measurements, temperature measurement system, high precision, constant-current source

摘要： 热测应力为观测地壳应力的动态变化提供了一个新视角, 但它的深入应用强烈依赖于高精度温度测量技术的进步。 文中基于低温漂定值电阻与测温电阻组成的平衡桥式四线制温度传感器, 考虑了恒流换向驱动和深度Kalman数字滤波等一系列技术改进后, 研发了新一版高精度温度测量系统, 设计的温度分辨率为0.003mK。 经过实际检验: 温度测量精度已达0.03mK, 且通过野外观测证实了其测温方案具有可行性。 据此, 可较大幅度提高热测应力的可观测范围, 技术上可获得0.01MPa量级的动态应力变化, 达到了同震库仑应力变化测量的量级。 这意味着新的观测系统可明显地拓展热测应力的应用范围。

关键词: 热测应力, 温度测量, 高分辨率, 恒流源