- Title
- Development of new retracking methods for mapping sea levels over the shelf areas from satellite alimetry data
- Creator
- Idris, Nurul Hazrina
- Relation
- University of Newcastle Research Higher Degree Thesis
- Resource Type
- thesis
- Date
- 2014
- Description
- Research Doctorate - Doctor of Philosophy (PhD)
- Description
- Through research carried out in the last few years, sea level anomalies (SLAs) from altimeter range measurements have been improved in the near coastal zone between 50-10 km from the coastline using waveform retracking techniques. However, closer than about 10 km from the coastline, the improvement of altimetry data accuracy is still challenging due to the complex nature of the coastal topography and rougher coastal sea states. Although there is a healthy diversity of waveform retracking algorithms that have been beneficial to the coastal community, there is a lack of clear recommendations and guidelines on which retracker should be used under the various conditions. This dissertation presents a waveform retracking system that improves the accuracy of coastal altimetry data through the optimal selection and seamless switching of retrackers. The principles of the system are twofold. The first is to reprocess altimeter waveforms using the optimal retracker, which is sought, based on the analysis from a fuzzy expert system. The second is to minimise the relative offset in the retrieved SLAs caused by switching from one retracker to another, using a neural network. With the retracking system, the risk of assigning the waveform to an inappropriate retracker is minimised by including information about the waveform shapes and statistical features of the retracking results in the fuzzy expert system. The system also reduces inconsistency in the retracked SLAs when switching retrackers by employing the neural network to handle the nonlinear relationship between the retracker and the scattering surface, thus providing seamless transition from the open ocean to coast, or vice versa. The retracking system has been demonstrated to 20 Hz waveforms of Jason-1 and Jason-2/OSTM missions from 2009 to 2011. It has been applied to areas of the Great Barrier Reef in Australia and the Prince William Sound in Alaska. The regional investigations have demonstrated that the retracking system can effectively improve the quality of the altimeter derived SLAs in coastal regions. It reduces the standard deviation of the unretracked sea levels by up to 500 cm for Jason-1 and 300 cm for Jason-2. It extends the SLA profiles further (1-7 km) to the coastline and recovers up to 70% more data than the existing retrackers from the Sensor Geophysical Data Records(SGDR). Comparison with the SLAs from the tide gauges indicates that the SLAs from the retracking system are more reliable than those of from the SGDR products, in the sense that it has a higher (>0.8) temporal correlation and smaller (<17 cm) RMS errors. The retracked SLAs from the retracking system also produce reliable geostrophic velocities as they are consistent with those of the high frequency radar velocities in the Great Barrier Reef region. Comparison with the Regional Ocean Modelling System (ROMS) at Prince William Sound shows good agreement between the SLA patterns from the retracking system and the ROMS. The results obtained in this dissertation, therefore, present a significant improvement in the accuracy and precision of the estimated SLAs and efficiently reduce the altimetry no-data gap in coastal regions. In addition, it also addresses the systematic validation protocol for validating the altimetry retracked SLAs using the HF radar in the region of the Great Barrier Reef, and using the ROMS in the region of Prince William Sound.
- Subject
- coastal altimetry; waveform retracking; fuzzy expert system; neural networks; sub-waveform retracking; tide gauge; high frequency radar; regional ocean models; satellite altimetry
- Identifier
- http://hdl.handle.net/1959.13/1051125
- Identifier
- uon:15248
- Rights
- Copyright 2014 Nurul Hazrina Idris
- Language
- eng
- Full Text
- Hits: 830
- Visitors: 1595
- Downloads: 581
| Thumbnail | File | Description | Size | Format | |||
|---|---|---|---|---|---|---|---|
| View Details Download | ATTACHMENT01 | Abstract | 628 KB | Adobe Acrobat PDF | View Details Download | ||
| View Details Download | ATTACHMENT02 | Thesis | 7 MB | Adobe Acrobat PDF | View Details Download |