Radar detection of accreted ice over Lake Vostok, Antarctica

Airborne radar data collected over Lake Vostok, East Antarctica, show a reflection close to the Vostok ice core below the depth of the meteoric-accreted ice boundary (MAIB) and near the boundary between impurity-rich (1) and impurity-poor (2) accreted ice (A12B). The spatial pattern of detection of this reflection is aligned with ice flow over the southern half of the lake. Ice-core data from the accreted ice show large variations in the number density of inclusions, soluble chloride-ion concentration ([Cl-1]), charge-balance-derived acidity ([H+](cb)) and crystal-orientation fabric. Here we investigate the existing hypothesis that this reflection is due to the MAIB and seek possible reflection causes that explain the radar-derived reflectivity (-51 +/- 8 dB). The likely maximum reflectivity due to an insoluble mineral-inclusion contrast is -125 dB. If none of the Cl- or H+ ions are trapped within inclusions, then ice-conductivity models predict a reflectivity of -50 dB near the MAIB and -58 dB near the A12B, both primarily due to large [H+](cb) changes. If some of those ions are trapped within inclusions, then those values are smaller. Fabric contrasts at either boundary can produce significant reflectivities (-62 dB). These results indicate that the mineral-inclusion contrast near the MAIB does not cause the observed reflection. Because of uncertainty in the in situ location of the soluble impurities, and because impurity-rich accreted ice 1 is believed to only originate in the small bay in the southwestern corner of the lake yet the reflection is observed over a much wider area, we argue that a fabric contrast at the A12B is the probable reflection cause. Both predicted spatial variations in accretion rates near the upstream shoreline, or lower ice temperatures along flowbands that originate at bedrock ridges near the upstream shoreline, may explain why the reflection is not detected everywhere that accretion is predicted. Published by Elsevier B.V.