The earth’s magnetic field is not constant. It is prone to the pushes and pulls of the interplanetary magnetic field. While studying this interaction, a few years ago, A Abraham, G Renuka and Ligi Cherian at the University of Kerala formulated a new index of disturbance in the local magnetic field: the vertical variance index, the root sum square of variations in the horizontal component of the local magnetic field .
They tested the utility of this index using data from the near-equator ground stations at Ascension Island in the Atlantic Ocean and Huancayo, Peru. They could show that the vertical variance, a measure of the fluctuations in the time series data, correlated well with changes in the interplanetary magnetic field.
During their analysis, they noticed a curious phenomenon. The vertical variance of the two stations did not agree when they examined minute-to-minute variations in both stations. Now, in a recent paper, they argue that the differences between the stations can be leveraged to predict earthquakes .
The researchers analysed the correlation of the ratio of geomagnetic-interplanetary field vertical variance between pairs of magnetic field stations at Alibaug in India, PhuThuy in Vietnam, Huancayo in Peru, and Qsaybeh in Lebanon for December 2004. They examined the Advanced National Seismic System data of earthquakes of Richter scale 5 or more for the same month and noticed that the correlations in the vertical variance ratio between two stations start reducing before earthquakes, if the epicentres are within 2000 kilometres of these stations.
“Earthquakes of moment magnitude above 5 can be identified by the fall in the vertical variance ratio correlations”, explains G Renuka, University of Kerala.
The team repeated their observations again with the data of earthquakes in April 2012. This time, they analysed the vertical variance ratios between the field stations at Hyderabad, Alibaug and Addis Ababa in Ethiopia. Again the sharp dip in the correlations in vertical variance ratio showed up just before the earthquake. The correlation loss is completely absent if there are no earthquakes near any of the two magnetic field stations.
“The dip in the ratio also correlates well with ultra-long frequency signals from the ionosphere, another signature of earthquakes. So we feel that our results will be useful to the scientific community”, says A Abraham, Kerala University.
There is thus a need to start monitoring the index at the magnetic field stations for predicting earthquakes. India presently has four magnetic field stations – Choutuppal, Hyderabad, Alibaug and Jaipur – linked to the international network of magnetic observatories. A large-scale experiment needs to be undertaken to compute and compare the vertical variance with the magnetic field stations in nearby countries to test the potential of the index for forecasting earthquakes in India.
 Journal of Geophysical Research Atmospheres, 115: A01207 (2010);
 Indian Journal of Physics, 94 (8): 1147-1151;
Udham P K
Freelance Writer, Pune
STEAMindiaReports so that earthshaking discoveries don’t rattle people