Calibration of sea ice drift forecasts using random forest algorithms
Developing accurate sea-ice drift forecasts is essential to support decision making of maritime end-users operating in the Arctic. In this study, two calibration methods have been developed for improving 10-day sea-ice drift forecasts from an operational sea-ice prediction system (TOPAZ4). The metho...
| Main Authors: | , |
|---|---|
| Format: | Text |
| Language: | English |
| Published: |
2021
|
| Subjects: | |
| Online Access: | https://doi.org/10.5194/tc-2021-24 https://tc.copernicus.org/preprints/tc-2021-24/ |
| Summary: | Developing accurate sea-ice drift forecasts is essential to support decision making of maritime end-users operating in the Arctic. In this study, two calibration methods have been developed for improving 10-day sea-ice drift forecasts from an operational sea-ice prediction system (TOPAZ4). The methods are based on random forest algorithms (supervised machine learning models) and have been trained using either drifting buoy or synthetic-aperture radar observations for the target variables. Depending on the calibration method, the mean absolute error is reduced, on average, between 5.9 % and 8.1 % for the direction, and between 7.1 % and 9.6 % for the speed of sea-ice drift. Overall, the algorithms trained with buoy observations have the best performances when the forecasts are evaluated using drifting buoys as reference. However, there is a large spatial variability in these results, and the algorithms trained with buoy observations have particularly poor performances for predicting the speed of sea-ice drift in the Canadian Archipelago, along the east coast of Greenland, and north of Svalbard. In these areas, the algorithms trained with SAR observations have better performances for predicting the speed of sea-ice drift. |
|---|