This study is about the technical feasibility and the usefulness of two methods to provide additional information from existing satellite images for ice monitoring purposes by means of a prototype implementation. The first method shows how to make radar satellite images more homogenous compensating for the changes in intensity that are due to incidence angle variations across the images. This method could reduce the risk for misinterpretation of the images and also helps compare images taken at different moments for detection of changes like ice drift.
The other method – warping of images based on predicted ice drift – could help the users assess the current ice situation especially regarding the location of the ice edge and updated position of possible cracks and openings in the ice.
The prototype implementation of the functionalities showed that technically the ideas are rather straightforward to implement. For users in the Baltic Sea, however, the interactive incidence angle correction does not give very much added value as long as the satellite images are radiometrically calibrated according to the instructions by the satellite data provider. For the advanced users, this functionality could be useful for getting the most out of the image data.
Regarding the image warping functionality, the value depends strongly on the accuracy of the drift forecasts. When included as part of an ice monitoring system for a fleet of icebreakers, the idea of applying observed ice drift vectors to the most recent satellite images, is definitely worth developing further. Using a separate component for the actual image warping, as in the pilot implementation, this nicely divides the implementation into well-defined modules that can be integrated into operational decision support systems. The prototype implementation demonstrates one possible way to do this integration.