Snow accumulation and compaction derived from GPR data near Ross Island, Antarctica

Abstract. We present an improved method for estimating accumulation and compaction rates of dry snow in Antarctica with ground penetrating radar (GPR). Using an estimate of the emitted waveform from direct measurements, we apply deterministic deconvolution via the Fourier domain to GPR data with a nominal frequency of 500 MHz. This reveals unambiguous reflection horizons which can be observed in repeat measurements made one year apart. At two measurement sites near Scott Base, Antarctica, we extrapolate point measurements of average accumulation from snow pits and firn cores to a larger area by identifying a dateable dust layer horizon in the radargrams. Over an 800 m × 800 m area on the McMurdo Ice Shelf (77°45´ S, 167°17´ E) the average accumulation is found to be 269 ± 9 kg m⁻² a⁻¹. The accumulation over an area of 400 m × 400 m on Ross Island (77°40´ S, 167°11´ E, 350 m a.s.l.) is found to be higher (404 ± 22 kg m⁻² a⁻¹) and shows increased variability related to undulating terrain. Compaction of snow between 2 m and 13 m depth is estimated at both sites by tracking several internal reflection horizons along the radar profiles and calculating the average change in separation of horizon pairs from one year to the next. The derived compaction rates range from 7 cm m⁻¹ at a depth of 2 m, down to no measurable compaction at 13 m depth, and are similar to published values from point measurements.