Accuracy Improvement on Sea Surface Height Estimation Based on Waveform Retracking Analyses of Jason-2 Satellite in Java Sea

Abstract

<p><em>A waveform created by the reflected signal from altimeter satellite in offshore is generally in ideal shape (Brown-waveform) and produces an accurate sea surface height (SSH) estimation. However, over coastal waters, the waveform shape becomes complex due to a disruption by reflected signal from land, resulting inaccurate SSH estimation. The objective of this research was to improve the accuracy of SSH estimation employing waveform retracking analyses of Jason-2 altimeter satellite data in the Java Sea during the years of 2012-2014. This study used data from the Sensor Geophysical Data Record type D (SGDR-D) from Jason-2 satellite (cycle 129 - 239) and global geoid undulation data of Earth Gravitational Model 2008 (EGM08). </em><em>Waveform retracking analyses were conducted using several retracker methods</em>. <em>The performance of the all retrackers were examined using a world reference undulation geoid of EGM08. The results showed that the waveform retracking analyses were able to improve the accuracy of SSH estimation approximately 29.7% in the north coast and 56.4% in the south coast of total non-Brown-waveform in each region. Higher improvement percentage (IMP) of SSH estimation found in the southern coastal areas was due to a relatively smooth coastline formation in this region than in northern coastal region. There was no specific retracker that produce dominant IMP of SSH estimation. However, the threshold 10% retracker produced better SSH estimation than the other retrackers with dominant IMP values of 57.1% (pass 051), 48.1% (pass 064), and 25.7% (pass 127). OCOG retracker the worst retracker to estimate SSH in the Java Sea. </em></p><p><strong><em>Keywords: </em></strong><em>EGM08, </em><em>waveform retracking, SSH, Jason-2, ocean retracker, threshold retracker</em></p>

Waveform Classification and Retracking of Jason-2 and Jason-3 in Halmahera Sea

“…However, the algorithm that most often presented the highest IMP was the Threshold 10%. The finding in the Java Sea also showed that Threshold 10 was one of the best retrackers [23,25]. Conversely, in Natuna Sea, which was shallow water, Threshold 10 gave poor performance [29].…”

“…Over the Indonesian seas, previous studies (e.g., [23,25,28,29]) reported the performance of several retrackers using the geoidal height data, which highlighting the precision of retracked SSHs. While the accuracy of the retrackers is also valuable, such studies are yet to be done, thus becoming the motivation of this study.…”

“…This was because the small threshold was the recommended algorithm for waveforms not dominated by surface scattering [35] while OCOG was only based on a statistical approach by calculating the gate width and the amplitude of waveform [17]. In the other regions of Indonesia e.g., Java Sea [23,25] and Natuna Sea [29], OCOG also showed poor performance and was estimated to be less accurate.…”

“…Across Southeast Asia, several studies have been conducted to assess the retracking performance over the marginal seas [11,20,23,24,25,26,27,28,29]. The performance can be determined by comparing with geoidal height data and tide gauge data.…”