11:45 AM - 12:00 PM
[AOS18-05] Advancement of Wave Observation Technology Using Ocean Radar
Keywords:High-frequency Oceanographic Radar, Wave observation, S/N ratio, Doppler spectrum
1. Introduction
Ocean radar is a land-based observation instrument capable of continuously monitoring wide-area information (such as waves and currents) from the coast to offshore. In recent years, expectations for its application in wave observation have been increasing. However, unlike conventional observation instruments that directly measure wave heights (e.g., offshore GPS wave gauges), ocean radar utilizes radio waves for wave observation. Due to this characteristic, addressing external noise is essential for improving wave observation accuracy. In this study, we developed a novel analytical method for wave observation using ocean radar and conducted a comparative validation with GPS wave gauges. As a result, we successfully achieved a significant improvement in wave observation accuracy, even under high-wave conditions, which were previously difficult to observe accurately.
2. Development of an Observation Method Considering Ocean Radar Characteristics
(1) Spatial Interpolation Considering the Signal-to-Noise Ratio (SNR)
We evaluated "external noise," which significantly impacts the wave observation accuracy of ocean radar, in terms of the "signal-to-noise ratio (SNR)." In the study area, external noise was relatively high, and various noise reduction measures were tested using different SNR conditions. Consequently, we developed a new method that explores a concentric range around the observation point, extracts only the observation results with high SNR within that range, and adopts the median of those data.
(2) Determination of Bimodal Peak Occurrence Areas
Wave analysis using ocean radar calculates wave height and period based on the ratio of energy between the primary scattering region (main region) and the secondary scattering region (sub-region) in the observed data (Doppler spectrum). Therefore, clearly distinguishing these regions is a prerequisite for accurate wave observation. However, in complex flow fields, such as the Kuroshio Current area, the peak of the primary scattering region sometimes splits into a bimodal shape, known as the "bimodal peak" phenomenon. In such cases, it becomes difficult to correctly separate the regions, leading to decreased observation accuracy. To address this issue, we devised an automatic bimodal peak detection method and developed a new approach that performs analysis only in regions where bimodal peaks are not present.
3. Improvement in Wave Observation Accuracy Using the Developed Method
(1) Spatial Interpolation Considering SNR
A significant improvement effect was observed, particularly in wave height estimation. In conventional methods, when the SNR at the observation point was low, observation accuracy decreased (as it relied solely on the SNR of a single observation point). However, the newly developed method interpolates data spatially while considering SNR, enabling high-accuracy observations even under high-wave conditions. This approach successfully reduced outliers and yielded results more consistent with traditional wave gauges.
(2) Determination of Bimodal Peak Occurrence Areas
A significant improvement effect was observed, particularly in wave period estimation. Conventional methods tended to underestimate wave periods during the arrival of swell-type waves. However, the newly developed method achieved wave period estimations comparable to those obtained from validation data.
4. Conclusion
This study demonstrated that applying the proposed method to conventional ocean radar enables more stable data acquisition. In future work, we aim to explore the potential of ocean radar as an alternative observation instrument that could serve as a backup or replacement for these conventional wave gauges.
Ocean radar is a land-based observation instrument capable of continuously monitoring wide-area information (such as waves and currents) from the coast to offshore. In recent years, expectations for its application in wave observation have been increasing. However, unlike conventional observation instruments that directly measure wave heights (e.g., offshore GPS wave gauges), ocean radar utilizes radio waves for wave observation. Due to this characteristic, addressing external noise is essential for improving wave observation accuracy. In this study, we developed a novel analytical method for wave observation using ocean radar and conducted a comparative validation with GPS wave gauges. As a result, we successfully achieved a significant improvement in wave observation accuracy, even under high-wave conditions, which were previously difficult to observe accurately.
2. Development of an Observation Method Considering Ocean Radar Characteristics
(1) Spatial Interpolation Considering the Signal-to-Noise Ratio (SNR)
We evaluated "external noise," which significantly impacts the wave observation accuracy of ocean radar, in terms of the "signal-to-noise ratio (SNR)." In the study area, external noise was relatively high, and various noise reduction measures were tested using different SNR conditions. Consequently, we developed a new method that explores a concentric range around the observation point, extracts only the observation results with high SNR within that range, and adopts the median of those data.
(2) Determination of Bimodal Peak Occurrence Areas
Wave analysis using ocean radar calculates wave height and period based on the ratio of energy between the primary scattering region (main region) and the secondary scattering region (sub-region) in the observed data (Doppler spectrum). Therefore, clearly distinguishing these regions is a prerequisite for accurate wave observation. However, in complex flow fields, such as the Kuroshio Current area, the peak of the primary scattering region sometimes splits into a bimodal shape, known as the "bimodal peak" phenomenon. In such cases, it becomes difficult to correctly separate the regions, leading to decreased observation accuracy. To address this issue, we devised an automatic bimodal peak detection method and developed a new approach that performs analysis only in regions where bimodal peaks are not present.
3. Improvement in Wave Observation Accuracy Using the Developed Method
(1) Spatial Interpolation Considering SNR
A significant improvement effect was observed, particularly in wave height estimation. In conventional methods, when the SNR at the observation point was low, observation accuracy decreased (as it relied solely on the SNR of a single observation point). However, the newly developed method interpolates data spatially while considering SNR, enabling high-accuracy observations even under high-wave conditions. This approach successfully reduced outliers and yielded results more consistent with traditional wave gauges.
(2) Determination of Bimodal Peak Occurrence Areas
A significant improvement effect was observed, particularly in wave period estimation. Conventional methods tended to underestimate wave periods during the arrival of swell-type waves. However, the newly developed method achieved wave period estimations comparable to those obtained from validation data.
4. Conclusion
This study demonstrated that applying the proposed method to conventional ocean radar enables more stable data acquisition. In future work, we aim to explore the potential of ocean radar as an alternative observation instrument that could serve as a backup or replacement for these conventional wave gauges.