Abstract
Marine operations in ice-infested waters require reliable and timely information about the sea ice conditions. The Canadian Ice Service produces and distributes the ice information to mariners operating in Canadian waters, mainly in the form of daily ice charts. Unfortunately, due to the time difference between the production and the use of the ice charts, the ice information is always out of date, which endangers the safety of marine operations. To efficiently overcome this problem, a reliable model for predicting the sea ice concentration over time is developed. Examining the ice charts of the Gulf of St. Lawrence during the period 1987 to 1998 showed that the sea ice conditions change according to a regular pattern to some extent. Therefore, a neural network function approximation system could model, and hence predict, these changes efficiently when trained, using multiple-year ice concentration readings. Initially, the training was done in the batch mode. However, this was found inefficient when abrupt changes in the values of the ice concentration were encountered. Therefore, a sequential model, which uses the modular neural network structure, was developed. In addition to overcoming the drawbacks of the batch method, the sequential model is more suitable for real-time applications.
Original language | English |
---|---|
Pages (from-to) | 87-92 |
Number of pages | 6 |
Journal | International Hydrographic Review |
Volume | 4 |
Issue number | 2 |
Publication status | Published - 2003 |
ASJC Scopus subject areas
- Oceanography
- Environmental Science (miscellaneous)