Summer and early-fall sea-ice concentration in the Ross Sea: Comparison of in situ ASPeCt observations and satellite passive microwave estimates

Abstract

Sea-ice conditions were observed using the ASPeCt observation protocol on three cruises in the Ross Sea spanning the Antarctic summer season (APIS, December 1999-February 2000; AnSlope 1, March-April 2003; AnSlope 2, February-April 2004). An additional dataset was analyzed from helicopter video surveys taken during the APIS cruise. The helicopter video was analyzed using two techniques: first, as an ASPeCt dataset where it was sampled visually for ice concentration, floe sizes and ice type on a point basis at 11 km intervals; second, computerized image processing on a subset of nine helicopter flights to obtain ice concentration on a continuous basis (1 s intervals) for the entire flight. This continuous sampling was used to validate the point-sampling methods to characterize the ice cover; the 'ASPeCt sampling' on the helicopter video and the use of the ASPeCt protocol on the ship surveys. The estimates for average ice concentration agreed within 5% for the continuous digitized data and point sampling at 11 km intervals in this comparison. The ship and video in situ datasets were then compared with ice concentrations from SSM/I passive microwave satellite data derived using the Bootstrap and NASA-Team algorithms. Less than 50% of the variance in summer ice concentration observed in situ was explainable by satellite microwave data. The satellite data were also inconsistent in measurement, both underestimating and overestimating the concentration for summer conditions, but improved in the fall period when conditions were colder. This improvement was in the explainable variance of >70%, although in situ concentration was underestimated (albeit consistently) by the satellite imagery in fall.