A rapid and robust scheme for CFOSAT SWIM nadir SWH retrieval

The Surface Waves Investigation and Monitoring (SWIM) onboard the China-France Oceanography Satellite (CFOSAT) has acquired unprecedented global ocean wave direction spectrum information. Additionally, it incorporates a nadir observation mode for significant wave height (SWH) based on principles analogous to conventional altimeters, which serves to calibrate the total energy of the wave spectrum. The retrieval of SWH from such nadir radar observations generally employs theoretical models to retrack the waveforms reflected by the sea surface, a method that demonstrates high accuracy in the open ocean environment. However, obtaining SWH using traditional retracking algorithms is computationally expensive. Furthermore, in coastal regions, waveforms are affected by factors such as the presence of land and sea ice, which means that the SWH retrieval relying on theoretical models often either fails or yields substantial errors. Leveraging the fact that SWIM does not necessitate precise ranging, we bypassed the decades-old retracking paradigm by developing dual neural network inversion models. These models perform direct SWH estimation following an initial classification of SWIM nadir waveforms. The first model targets standard open-sea waveforms, while the second is designed for coastal and sea-ice 'ocean-like' waveforms that deviate from normal patterns yet retain essential wave information. The results show that the normal-ocean model achieves a processing speed over 13 times faster than the MLE4 algorithm and nearly 22 times faster than the adaptive retracking algorithm, while maintaining comparable accuracy (RMSE = 0.31 m) to the SWIM nadir SWH product. Meanwhile, the ocean-like model enhances both the quantity and quality of usable SWH data in non-open ocean areas, leveraging the accurate mapping relationships established by the neural networks. Generally, a rapid and robust scheme was developed in this paper for retrieving nadir SWH from CFOSAT SWIM, providing an alternative methodology for retrieving nadir geophysical parameters in future nadir-looking radar systems.

Authors

Institutions

Publication Details

Journal
GIScience & Remote Sensing
Published
2026-11-17
DOI
https://doi.org/10.1080/15481603.2026.2741445
Primary Topic
Ocean Waves and Remote Sensing
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

A rapid and robust scheme for CFOSAT SWIM nadir SWH retrieval

Ying Xu, Xiuzhong Li, Shuyan Lang, Dong Zhu
GIScience & Remote Sensing
Ocean Waves and Remote Sensing
article

A rapid and robust scheme for CFOSAT SWIM nadir SWH retrieval

Ying Xu, Xiuzhong Li, Shuyan Lang, Dong Zhu
article en

Abstract

The Surface Waves Investigation and Monitoring (SWIM) onboard the China-France Oceanography Satellite (CFOSAT) has acquired unprecedented global ocean wave direction spectrum information. Additionally, it incorporates a nadir observation mode for significant wave height (SWH) based on principles analogous to conventional altimeters, which serves to calibrate the total energy of the wave spectrum. The retrieval of SWH from such nadir radar observations generally employs theoretical models to retrack the waveforms reflected by the sea surface, a method that demonstrates high accuracy in the open ocean environment. However, obtaining SWH using traditional retracking algorithms is computationally expensive. Furthermore, in coastal regions, waveforms are affected by factors such as the presence of land and sea ice, which means that the SWH retrieval relying on theoretical models often either fails or yields substantial errors. Leveraging the fact that SWIM does not necessitate precise ranging, we bypassed the decades-old retracking paradigm by developing dual neural network inversion models. These models perform direct SWH estimation following an initial classification of SWIM nadir waveforms. The first model targets standard open-sea waveforms, while the second is designed for coastal and sea-ice 'ocean-like' waveforms that deviate from normal patterns yet retain essential wave information. The results show that the normal-ocean model achieves a processing speed over 13 times faster than the MLE4 algorithm and nearly 22 times faster than the adaptive retracking algorithm, while maintaining comparable accuracy (RMSE = 0.31 m) to the SWIM nadir SWH product. Meanwhile, the ocean-like model enhances both the quantity and quality of usable SWH data in non-open ocean areas, leveraging the accurate mapping relationships established by the neural networks. Generally, a rapid and robust scheme was developed in this paper for retrieving nadir SWH from CFOSAT SWIM, providing an alternative methodology for retrieving nadir geophysical parameters in future nadir-looking radar systems.

GIScience & Remote SensingVol. 63(1)
Nanjing University of Information Science and Technology (CN), Ministry of Natural Resources (CN), National Satellite Ocean Application Service (CN)
Ocean Waves and Remote Sensing
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon