Predicting sea surface wave and wind parameters from satellite radar images using machine learning
| dc.contributor.author | Borg, Filip | |
| dc.contributor.author | Brobeck, Axel | |
| dc.contributor.department | Chalmers tekniska högskola / Institutionen för rymd-, geo- och miljövetenskap | sv |
| dc.contributor.department | Chalmers University of Technology / Department of Space, Earth and Environment | en |
| dc.contributor.examiner | Eriksson, Leif | |
| dc.contributor.supervisor | Amell, Adrià | |
| dc.contributor.supervisor | Elyouncha, Anis | |
| dc.date.accessioned | 2023-11-10T07:40:32Z | |
| dc.date.available | 2023-11-10T07:40:32Z | |
| dc.date.issued | 2023 | |
| dc.date.submitted | 2023 | |
| dc.description.abstract | Accurate predictions of wave and wind parameters over oceans are crucial for various marine operations. Although buoys provide accurate measurements, their deployment is limited, which necessitates the exploration of alternative data sources. Sentinel-1, a satellite mission capturing Synthetic Aperture Radar (SAR) images with high coverage, presents a promising opportunity. However, establishing the relationship between SAR images and wave/wind parameters is not straightforward. This project aims to develop a machine learning model that can effectively extract this relationship. To accomplish this, data from all available buoys measuring significant wave height and wind speed in the year 2021 were utilized. The corresponding SAR images were located, and 2 km×2 km sub-images were extracted around each buoy. From each sub-image, a set of features were extracted. These sub-images and features served as input to train machine learning models capable of predicting buoy measurements, supplemented with model data as necessary. The project presents two final deep learning models: one utilizing only the extracted features and another employing both the sub-images and features. These multi-class regression models simultaneously predict significant wave height and wind speed. The model using only features achieved a Root Mean Square Error (RMSE) of 0.553 m for significant wave height and 1.573 m/s for wind speed. The model incorporating both sub-images and features achieved an RMSE of 0.459 m for significant wave height and 1.658 m/s for wind speed. The code for the project can be found on https://github.com/SEE-GEO/sarssw. | |
| dc.identifier.coursecode | seex30 | |
| dc.identifier.uri | http://hdl.handle.net/20.500.12380/307351 | |
| dc.language.iso | eng | |
| dc.setspec.uppsok | LifeEarthScience | |
| dc.subject | Machine Learning, Computer Vision, Synthetic Aperture Radar, Sig nificant Wave Height, Wind Speed, Radar, Master Thesis, Chalmers University of Technology | |
| dc.title | Predicting sea surface wave and wind parameters from satellite radar images using machine learning | |
| dc.type.degree | Examensarbete för masterexamen | sv |
| dc.type.degree | Master's Thesis | en |
| dc.type.uppsok | H | |
| local.programme | Data science and AI (MPDSC), MSc |