A comparison of seismic and radar methods to establish the thickness and density of glacier snow cover

We show that geophysical methods offer an effective means of quantifying snow thickness and density. Opportunistic (efficient but non-optimized) seismic refraction and ground-penetrating radar (GPR) surveys were performed on Storglaciaren, Sweden, co-located with a snow pit that shows the snowpack to be 1.73 m thick, with density increasing from similar to 120 to similar to 500 kg m(-3) (with a +50 kg m(-3) anomaly between 0.73 and 0.83 m depth). Depths estimated for two detectable GPR reflectors, 0.76 +/- 0.02 and 1.71 +/- 0.03 m, correlate extremely well with ground-truth observations. Refraction seismic predicts an interface at 1.90 +/- 0.31 m depth, with a refraction velocity (3730 +/- 190 m s(-1)) indicative of underlying glacier ice. For density estimates, several standard velocity-density relationships are trialled. In the best case, GPR delivers an excellent density estimate for the upper snow layer (observed = 321 +/- 74 kg m(-3), estimated = 319 +/- 10 kg m(-3)) but overestimates the density of the lower layer by 20%. Refraction seismic delivers a bulk density of 404 +/- 22 kg m(-3) compared with a ground-truth average of 356 +/- 22 kg m(-3). We suggest that geophysical surveys are an effective complement to mass-balance measurements (particularly for controlling estimates of snow thickness between pits) but should always be validated against ground-truth observations.