Fractal analysis of Ultra Low Frequency magnetic field emissions observed at Agra associated with two major earthquakes occurred in Pakistan

. Fractal analysis has been applied to the ultra-low frequency (ULF) magnetic time series data to study the critical behaviour of impending earthquake preparation zone. Burlaga Klein method has been adapted. The ULF time series data have been prepared from the routine recordings of ULF magnetic field emissions at Agra station in India with the help of 3-component search coil magnetometers for the month of September 2013 in which two large earthquakes of magnitudes M=7.4 and 6.8 occurred in Pakistan. The two important parameters namely fractal dimension D and fractal exponent β are calculated which are found to vary abnormally before the occurrence of earthquakes. From a detailed amplitude-time analysis of ULF data precursory period of anomalies in amplitude has been found of the order of 16 days. The fractal dimension and fractal exponent show complementary behaviour with each other and they are found to vary abnormally around the observed precursory period in the normal analysis of the data. This month is free from geomagnetic disturbances and hence the analysis is not contaminated from the effects of such disturbances.