Iceberg size and orientation estimation using SeaWinds

SeaWinds scatterometer data from 1999 to 2009 has been used to detect and track large Antarctic icebergs on a daily basis. Here, we develop two estimation algorithms to supplement iceberg position reports with estimates of the icebergʹs major-axis length, minor-axis length, and angle of rotational orientation. The first algorithm is an image-based approach while the second is based on raw SeaWinds measurements. Maximum-likelihood objective functions that relate SeaWinds backscatter observations to model predictions are developed. The utility of both estimation approaches is analyzed in simulation and with actual data via a case study of iceberg A22a. The image-based approach is found to outperform the measurement-based technique. Results are validated using high-resolution imagery. Estimates are also compared with reports compiled by the United States National Ice Center.