Meteorology in Tunnel - Effect of barometric pressure changes on extensometer observations

日本长野松代地震台100 m石英管引伸计观测噪声的主要来源之一是大气温度变化对石英管的影响。温度变化又是由于大气压力变化引起的内部空气的绝热压缩/膨胀引起的。在本研究中，我们对温度、湿度和压力进行了连续的观测，以准确地了解隧道内的情况。有以下主要結果: 1.气压和温度有很好的相关性。大多数观测到的温度变化是由压力变化引起的。温度变化与压力变化近似成比例，比例系数随频率系统性变化。该系数在短于2小时的时间段内几乎是常数，而随着时间段变长，该系数变小。隧道内的相对湿度始终高于90%，且全年无变化。我们无法证实温度的变化是否伴随着湿度的变化。在较短的周期内，湿空气的比例系数（0.05K/hPa）比干空气（0.085K/hPa）更接近理论值（0.042K/hPa）。考虑了空气的绝热变化和空气与围岩之间的热传递的理论模型很好地解释了压力效应的频率依赖性，这些结果得出结论，隧道内的温度变化可以由压力记录预测，精确的压力和温度观测可以消除伸长计记录中的观测噪声。然而，仍然存在一些无法解释的事实，包括所观察到的比例性有利于湿空气模式而不是干空气模式。

One of the main sources of observational noise for 100 meter quartz tube extensometers in Matsushiro Seismological Observatory, Matsushiro, Nagano, Japan is the effect of air temperature changes on the quartz tubes. The temperature changes, in turn, are caused by adiabatic compression/expansion of the inside air due to atmospheric pressure changes. In this study, we performed continuous observation of temperature, humidity and pressure in order to understand exactly what are going on inside the tunnel. The main results are :1. Air pressure and temperature are in very good correlation. Most of the observed temperature changes are caused by pressure changes.2. Temperature changes are approximately proportional to pressure changes, and the coefficient of proportionality varies systematically with frequency. The coefficient is almost constant at periods shorter than 2 hours, whereas it becomes smaller as the period gets longer.3. Relative humidity inside the tunnel is always higher than 90%, and does not change throughout a year. We could not verify whether temperature changes accompany humidity changes or not.4. The coefficient of the proportionality (0.05K/hPa) at the shorter periods is closer to the theoretical value for humid air (0.042K/hPa), rather than that for dry air (0.085K/hPa).5. The frequency dependence of the pressure effect is very well explained by a theoretical model which takes into account adiabatic changes of the air and heat transfer between air and surrounding rocks.These results lead to a conclusion that temperature changes inside the tunnel are predictable from pressure records and that precise observations of pressure as well as temperature enable us to remove observational noise from extensometer records. However, there still remains unexplained facts, including that the observed proportionality favors a humid air model rather than a dry air model.