Electromagnetic signals on active volcanoes: Analysis of electrical resistivity and transfer functions at Taal volcano (Philippines) related to the 2010 seismovolcanic crisis

Electromagnetic (EM) methods are powerful for imaging lithology, geological discontinuities, and hydrothermal and plumbing systems of volcanoes. These methods can also be efficiently used for monitoring temporal changes related to the time evolution of volcanic activity. However, the magnetotelluric (MT) technique has not been yet investigated on many volcanoes. In this study, we take advantage of simultaneous magnetic and electric records made on Taal volcano in The Philippines to prospect if resistivity changes were related to the 2010-2011 seismovolcanic crisis. The seismicity began on April 19, 2010 and abruptly raised up on April 29. The crisis vanished nine months later on March 2011, and was characterized by the appearance of a source of deformation at 5 km depth which gave birth to a secondary deformation source located in the hydrothermal system at about 2.5 km depth some weeks later. Two stations, DAK and MCL, located respectively on the northern flank and inside the summit crater, were simultaneously recording the horizontal North-South and East-West components of both magnetic and electric fields from March 11 to June 13, 2010. 1D inversion modeling shows that resistivity at the depth of 5 km where dikes and deformation sources are positioned, is estimated smaller a ten of Ω m. Temporal changes of the apparent resistivity and power spectra of the components of the EM field are analyzed using both a single station approach at the two DAK and MCL stations, and a remote reference technique with a far (~ 3000 km) and a near station (~ 1.1 km). Local noises due to radio transmitters at stations prevent studying signals below 10 s of period. The day-to-day time-lapse analysis with the single station method shows several clear decreases of the apparent resistivity at the station inside the crater, and high fluctuations at the station outside of it, both in relation to the crisis for periods less than few hundred seconds . Application of the remote reference technique with far and close remote stations leads to assume that there was no clear change at the station located in the crater, while some fluctuating resistivity changes could have occurred at the second station DAK. Furthermore, the MT responses undoubtedly show that other local EM signals have taken place on the volcano before and during the crisis. In particular, electric signals linked to the reactivation of fissures and disturbances of ground fluids are present.