Retrieval and validation of improved coastal altimetry datasets from a new retracking strategy

The measurements of sea surface height (SSH), significant wave height (SWH) and wind speed derived from satellite altimetry are important data sources for coastal applications. However, altimetry data within 10 km of the coast for most altimeters are often discarded because the altimetric waveforms are contaminated by bright targets (e.g. land and sheltered waters). In addition, the accuracy of sea state bias (SSB) correction, which is crucial to improving the precision of high-rate sea levels, is affected by the degradation of SWH and wind speed measurements. Therefore, this thesis aims to retrieve more reliable altimeter data within the 0–10 km coastal strip by reprocessing coastal waveforms. This research developed a new Brown-peaky (BP) retracker to process coastal altimetric waveforms. The BP retracker adopts two algorithms, the three-parameter weighted least squares (WLS3) estimator and the adaptive leading edge subwaveform (ALES) retracker, to handle coastal waveforms. The WLS3 estimator uses the full waveform and reduces contamination by assigning downsized weights to contaminated waveform gates (Peng and Deng, 2018a). The ALES retracker uses only a portion of the waveform to avoid contamination in the waveform trailing edge (Passaro et al., 2014a). The combination of the WLS3 estimator and ALES retracker enhances the ability to retrieve reliable estimates from contaminated waveforms. Three parameters are resolved from the BP retracker: waveform amplitude, epoch and rise time of the leading edge. These parameters can then be converted into wind speed, SSH and SWH. The BP-derived SSHs, SWHs and wind speeds have been validated in different coastal regions for three altimetry missions: Jason-1, Jason-2 and Jason-3. The validation is conducted by internal evaluation of data availability and precision, and external comparisons with in situ measurements—that is, tide gauges, buoys and anemometers. The results show that the BP retracker can achieve better performance than official products, extending high-quality 20 Hz estimates from the open ocean to 5 km off the coast on average. To enhance the capability of the BP retracker, the modified Brown-peaky (MBP) retracker was developed. The MBP-retracked sea level anomalies (SLAs), which are SSHs relative to the mean sea surface (MSS), are then used to validate synthetic aperture radar (SAR) mode SLAs from the Sentinel-3A altimeter. The results show that the Sentinel-3A achieved superior performance to the MBP-retracked Jason-3, both over open oceans and in coastal zones. The precision of 1 Hz MBP-derived SLAs (~1.7 cm) is comparable to 2 Hz Sentinel-3A SLAs, indicating that the Sentinel-3A can provide precise SLAs at finer spatial scales. However, the SAR mode data can also be seriously degraded when the ground track encounters complex coastal topography. Therefore, a coastal-dedicated retracker such as SAMOSA+ (SAR Altimetry Studies and Applications over Ocean+) is still needed to recover more reliable data for oceanographic applications.