Synthetic aperture radar data assimilation for sea ice analysis

Abstract

Data analysis based on a specific sensor has some limitations related to individual sensor properties and imaging geometry. Combining information from different sources diminishes these limitations and improves analysis. In this study zero dimensional variational (0D-Var) data assimilation method was used first time to combine passive microwave based analysis and synthetic aperture radar (SAR) data to reach improved sea ice analysis. For this purpose a simple geophysical model (forward model) were developed to map sea ice features into SAR feature space. Only passive microwave based analysis were utilized as background state and proper observed SAR features were employed to update background state by minimizing a cost function which depends on ice concentration as a sea ice feature. Manual image analysis charts were used for validation and improved sea ice analysis results were obtained by using the proposed method.