G-force measurements at Mount Nyiragongo

To evaluate the mechanical stress on the volcanic edifice that results from lava lake level variations, we deployed a self-recording, differential capacitance (MEMS Inertial Sensor STMicroelectronics LIS3LV02DQ), 3-axis X6-1A accelerometer (Gulf Coast Data Concepts, LLC) at a distance of ~100m from the center of the Nyiragongo lava lake on freshly erupted lava flows. The device range was used in high (12-bit) resolution mode, which corresponds to a sensitivity of about 1 mg. The device was set to high-sensitivity mode with four additional bits to improve resolution, yet with a much lower signal-noise ratio. Once in position, the accelerometer continuously recorded data for three-day periods in June 2010. The system was oriented so that the X- and Y-axes form a plain parallel to the lava lake. During data collection, we did not attempt to calibrate the precision of the angle because relative G-force measurements were required instead of absolute G-force measurements. To distinguish the tiny accelerations caused by temperature differentials of the atmosphere, from the forces caused by magma movements, the temperature of the X6-1A device was continuously recorded. Temperature variations were corrected for by applying a de-correlation method to the recorded signal. Data was collected at 20 Hz, regrouped into batches that cover 1 hour per observation and associated with one averaged temperature measurement. This method was reproducible because diurnal temperature variations were the main cause for heating and cooling.

为评估熔岩湖水位变化引发的火山锥体机械应力，我们在尼拉贡戈（Nyiragongo）熔岩湖中心约100米外的新喷发熔岩流上，部署了一套自记录式设备，其包含差分电容型微机电系统（Micro-Electro-Mechanical Systems，MEMS）惯性传感器（意法半导体STMicroelectronics LIS3LV02DQ）与X6-1A三轴加速度计（墨西哥湾海岸数据概念有限责任公司Gulf Coast Data Concepts, LLC）。该设备采用12位高分辨率模式工作，对应灵敏度约为1毫伽（mg）。此外，设备还被设置为可额外提升4位分辨率的高灵敏度模式，但该模式下信噪比会显著降低。部署就位后，该加速度计于2010年6月连续开展了为期3天的数据采集。系统的校准方向使得X轴与Y轴所在平面与熔岩湖平面平行。数据采集期间，由于我们仅需相对重力加速度测量结果而非绝对测量值，因此未对角度精度进行校准。为区分大气温差引发的微小加速度与岩浆运动产生的作用力，我们同步记录了X6-1A加速度计的温度数据。通过对采集信号应用去相关处理方法，完成了温度变化带来的干扰校正。数据以20赫兹（20 Hz）的采样率采集，随后被整合为每批次覆盖1小时的观测数据块，并与单次平均温度测量值关联。由于昼夜温差是设备热胀冷缩的主要诱因，该数据处理方法具备可复现性。