High Sensitivity Magnetic Gradiometer for Earthquake Research Applications

ISRAEL 2005

Ivan HrvoicH. Ginzburg, H. Zafrir, G. Steinitz, B. Shirman, G. Hollyer

Overview

• Magnetic Earthquake Research Methods– Introduction to Past and Current Methods

– Detectability of Earthquakes by Gradiometers

– Short Base (Gradient) Measurements

• Potassium SuperGradiometer

– Installation and Data Records

• Summary

Introduction to Magnetics

• Several decades of investigation

• Based on theory of piezomagnetism

and / or electrokinetics

• Possibility of detection is related to gradual pressure build-up prior to earthquakes or “events”

Monitoring Systems (1)

• Traditional Methods– Magnetic sensors (0.1 nT sensitivity)– Long base measurements

• Some startling results• No duplication of results

• Recent Work– Induction coils

• Improved sensitivity to 25 pT• But, limited bandwidth (0.01 Hz)

Monitoring Systems (2)

• Induction Coils– Detect first derivative of magnetic

field– Detect all 3 components– Skin Effect Problems

• 50 km @ 0.01 Hz• 1.6 km @ 10 Hz

Detectability of Earthquakes by Gradiometers

• Assuming earthquakes create dipolar anomalies, can calculate detectability of earthquakes of given magnitude:

• Where o is magnetic permeability, M is magnetic moment, and α is the angle between radius vector, r, and dipole direction.

• r is distance from hypocenter to the observation point

Magnetic Moment (1)

• Assuming cos2 = 0 for simplicity, the previous equation can be solved for M, Magnetic Moment, as follows:

• Results are extended to look at datafrom the Loma Prieta earthquake of 1989 which clearly shows magnetic precursors.

Magnetic Moment (2)

• Loma Prieta (M7.1) B=2.8 nT anomaly at 17 km depth to hypocenter with 1.7 x 1011 Am2 magnetic moment

• San Juan Bautista (M5.1) 20 pT anomaly at 9.4 km depth to hypocenter with 1.7 x 108 Am2 magnetic moment

Radius Calculation (1)

• Note the 3rd potential drop off of the magnetic field and the 4th potential of the magnetic gradient with radius:

• Extreme sensitivity needed to measure (dB / dr)• Prospects for elimination of man-made

noise are better

Radius Calculation (2)

• Maximum distance at which an event can be detected by different sensors

• Skin effect not taken into account

Potassium SuperGradiometer

• 3 sensors arranged according to terrain (horizontal or vertical)

• Sensor spacing up to 140m• Long term integration is promising

140m

100m

S2

100mS3 S1

100m

Computer

SuperGrad Array

Data (1)

Data (2)

Data (3)

Installations

• Currently in use near Dead Sea, Israel

• Measures fields at 3 sensors, 20 times per second with 50 msec and 1 sec integration

• 6 channels of data to 1 fT resolution

• GPS receiver provides Universal Time

• Noise background is about 0.1 pT for 1 sec integration; giving 2 fT/m sensitivity at 50m spacing

• Since July 2002, has acquired more than 10 billion readings; likely the most ever recorded for this type of application

Initial Installation - Israel

Data - Israel

pT

8 hours

Additional Installations

• System deployed in Magnetic observatory of the Geological Survey of Canada near Ottawa

• Test of basic system configuration for 6 months

– Remote operation– Downloading of data via internet or telephone

• Installation in Mexico (Oaxaca State)

– Photos and records from Mexico

• Also seeking other jurisdictions for siting system; potentially in regions of more active tectonism

Site Location- Mexico

Site Selection - Mexico

Site Survey Data - Mexico

Gradient Survey Data - Mexico

745550 745600 745650 745700 745750

1751050

1751100

1751150

1751200

1751250

1751300

-30-25-20-15-10-5051015202530354045505560657075

nT/m

Sensor Installation - Mexico

Data Acquisition & Communication - Mexico

Data - Mexico

pT

sec

Integrated Grad / Radon

• Complementary radon emanation measurements.

• Correlation with weak earthquakes in Dead Sea rift region

• The combined Supergradiometer / Radon system is now available for application by various groups pursuing earthquake research studies.

Summary

• Based on earlier assumptions (and not considering geometrical effects), we can conclude:

– Magnetometers of 1 nT sensitivity can detect only the strongest earthquakes (M >7)

– Induction coils are good for M>6

– SuperGradiometer in Fast mode is effective for M>6

– SuperGradiometer in Slow mode is effective for M>5