Abstract
The electrochemical seismometers based on molecular electronic transfer (MET) exhibit significantly higher sensitivity than conventional moving-coil seismometers, offering broad application prospects in seismic exploration. However, METs are extremely sensitive to environmental temperature, with their frequency response significantly altering with temperature changes. These alterations display nonlinear characteristics across different frequency bands (such as high and low frequencies), resulting in operating bandwidth drift, which severely affects data quality. Existing temperature compensation methods are only effective in certain frequency bands, and compensating for operating bandwidth drift remains challenging, currently a hot research topic. This paper proposes the FSC-DFF adaptive temperature compensation method, which utilizes Frequency Response Self-Calibration (FSC) and a Dynamic Force Balance Feedback system (DFF) to achieve temperature compensation across the entire frequency band of MET. This method effectively compensates for the working frequency band drift and supports on-site adaptive adjustment. The temperature characteristics of MET were measured within the temperature range of -20°C to 50°C, and simulations and experiments were conducted. After compensation, the sensitivity drift was reduced to 1.6%. Compared with traditional Open-Loop Post-Processing (OLPP) and Closed-Loop Thermistor Compensation (CLTC) methods, the low-frequency cutoff frequency drift decreased from 447.2% (OLPP) and 51.4% (CLTC) to 15.1%, while the high-frequency cutoff frequency drift decreased from 89.3% (OLPP) and 85.0% (CLTC) to 3.2%. This significantly expands the application prospects of electrochemical seismic detectors in areas with severe temperature changes.
Funding Information
  • National Science and Technology Major Project of China for Deep Earth (2024ZD1002503)
  • Southern Marine Science and Engineering Guangdong Laboratory (SML2022SP201)
  • Jilin Province Science and Technology Development Plan Item (20210101103JC)

This publication has 22 references indexed in Scilit: