Seawater motion-induced electromagnetic noise reduction in marine magnetotelluric data using current meters

Seawater motion-induced electromagnetic (EM) noise along the seafloor has a large impact on marine magnetotelluric (MT) data quality. Although the mechanical stability of ocean bottom electromagnetic receivers (OBEMs) has improved due to buoyancy optimization, completely eliminating EM noise generated by seafloor currents as a result of instrument rocking or induction from the Earth's magnetic field is still not possible. The velocity of the current represents the quantification of seafloor conditions. To mitigate this problem, we installed a current meter on an OBEM to measure the synchronous current velocity along with the OBEM data logger. For the marine EM surveys, we conducted two surveys composed of 42 marine EM data acquisition sites in the South China Sea. We observed a strong correlation between induced EM noise and current velocity when the speed was greater than 2 cm/s. Furthermore, we developed an adaptive correlation noise-canceling filter to reduce the induced EM noise, using the current meter data as a reference signal. The filter refined the coefficients using a least-mean-squares algorithm. We were able to reduce the induced EM noise by pre-filtering the raw time series data with an adaptive correlation noise-canceling filter and using current meter data from nearby sites. Since seafloor currents are clearly an issue that limits MT data quality, special efforts are necessary to reduce seawater motion-induced EM noise in marine MT surveys.