Tracking Greenland’s melting ice with seismic waves

Greenland’s ice cap contributes nearly 20% to Earth’s mean rate of sea-level rise of 3 mm/y. To monitor the thinning ice sheet, scientists typically rely on periodic surveys taken by airborne and satellite radar, laser altimetry, and NASA’s Gravity Recovery and Climate Experiment (GRACE) mission, whose twin spacecraft orbit Earth and measure perturbations in its gravity field. MIT’s Germán Prieto, his postdoc Aurélien Mordret, and their colleagues have now adapted a seismic-wave method to do the same job. Originally developed to monitor active volcanoes and fault zones, the method exploits ambient noise signals, such as the crashing of ocean waves on the shoreline, recorded at an array of seismometers inland. (See the article by Roel Snieder and Kees Wapenaar, Physics Today, September 2010, page 44 .) Seismic waves propagate at speeds that depend on the porosity of the crust and upper mantle: The more porous the rock, the slower the waves travel. The scientists reasoned that the seasonal loading and unloading of the annual snowpack would repeatedly close and reopen pores and cracks in the rock beneath it. By cross correlating the signals from select pairs of seismometers (red inverted triangles) in western Greenland between 2012 and 2014, they retrieved a continuous record of the waves propagating between the sensors. A comparison of each daily correlation function to a reference standard yielded the relative seismic-velocity changes (dv/v) from season to season; those results, averaged over the summer of 2012, are shown on the map as connecting lines against the blue background of GRACE data. The team then developed a poroelastic model to infer the volume of ice lost or gained. Reassuringly, the model results agree with those taken by GRACE to within 91% accuracy. (A. Mordret et al., Sci. Adv. 2, e1501538, 2016, doi:10.1126/sciadv.1501538 .)