A COMPARISON OF CONVENTIONAL AND PASSIVE MICROWAVE SEA-ICE DATASETS FOR HUDSON-BAY

A sea-ice dataset derived from passive microwave data acquired by the Scanning Multichannel Microwave Radiometer (SMMR) is compared with a conventional sea-ice dataset from Ice Branch, Atmospheric Environment Service, Canada. The conventional set uses data from several sources including ships of opportunity, reconnaissance aircraft, satellite photographs and climate stations. The comparison was made for a 3-year period from 1979 to 1981 over Hudson Bay, an area that is covered with first-year ice only. For 8 of the 12 months of the year, monthly area-average ice concentrations are within one tenth; larger differences are evident during periods of melting and freeze-up. Extensive ponding on first-year ice during the melt season is interpreted as open water by the SMMR algorithm, leading to significant ''errors'' in the passive microwave dataset. By comparing the two datasets, we were able to show the extent of ponding on the ice for a complete seasonal melt cycle. During freeze-up the algorithm, under some conditions, provides a better estimate of the amount of ice than a conventional dataset owing to the difficulty of observing the presence of new ice.