The gravitational resolving power of global seismic networks in the 0.1-10 Hz band

Among the first attempts to detect gravitational waves, the seismic approach pre-dates the digital era. Major advances in computational power, seismic instrumentation and in the knowledge of seismic noise suggest to reappraise its potential. Using the whole earth as a detector, with the thousands of digital seismometers of seismic global networks as a single phased array, more than two decades of continuous seismic noise data are available and can be readily sifted at the only cost of (a pretty gigantic) computation. Using a subset of data, we show that absolute strains h less than or similar to 10(-17) on burst gravitational pulses and h less than or similar to 10(-21) on periodic signals may be feasibly resolved in the frequency range 0.1-10 Hz, only marginally covered by current advanced LIGO and future eLISA. However, theoretical predictions for the largest cosmic gravitational emissions at these frequencies are a few orders of magnitude lower.