Chromaticity Angle‑Assisted Remote Sensing Inversion Method for Shallow Water Depth

Jinhui Shi; Bingyi Liu; Peizhi Zhu; Qi Zhou; Qiuyue Xu; Yan He

doi:10.3788/AOS241738

Acta Optica Sinica, Volume. 45, Issue 12, 1201009(2025)

Chromaticity Angle‑Assisted Remote Sensing Inversion Method for Shallow Water Depth

Jinhui Shi¹, Bingyi Liu^1、*, Peizhi Zhu¹, Qi Zhou¹, Qiuyue Xu¹, and Yan He^2,3,4

¹College of Marine Technology, Faculty of Information Science and Engineering, Ocean University of China, Qingdao 266100, Shandong , China

²Aerospace Laser Technology and System Department, Shanghai Institute of Optics and Fine Mechanics, Chinese Academy of Sciences, Shanghai 201800, China

³Wangzhijiang Innovation Center for Laser, Shanghai 201800, China

⁴Center of Materials Science and Optoelectronics Engineering, University of Chinese Academy of Sciences, Beijing 100049, China

show less

Abstract Get PDF(in Chinese)

Figures & Tables(13)

Fig. 1. CIE 1931 chromaticity diagram

Download full size

Fig. 2. Remote sensing image of study area

Download full size

Fig. 3. Distributions of water depth samples. (a) Distribution range of water depth; (b) distribution range of standard deviation

Download full size

Fig. 4. Monthly cloud coverage of study area

Download full size

Fig. 5. Scatter plots of models. (a) RF; (b) CARF; (c) XGB; (d) CAXGB; (e) SVR; (f) CASVR; (g) Lyzenga; (h) Stumpf

Download full size

Fig. 6. Residual distributions of models. (a) RF; (b) CARF; (c) XGB; (d) CAXGB; (e) SVR; (f) CASVR; (g) Lyzenga; (h) Stumpf

Download full size

Fig. 7. Proportion of error distributions for models

Download full size

Fig. 8. Bathymetric map based on CAXGB model

Download full size

Fig. 9. Depth profile analysis. (a) Location map of profile line; (b) comparison between inverted profile and the measured profile; (c) accuracy evaluation plot

Download full size

Fig. 10. SHAP response diagram for CARF model

Download full size

Table 1. Parameterization of models

View table

Table 1. Parameterization of models

Model	Parameter setting
RF	n_estimators is 165， max_depth is 15， max_features is ’sqrt’， min_samples_leaf is 2， min_samples_split is 3， bootstrap is True
XGB	n_estimators is 181， learning_rate is 0.067， max_depth is 11， gamma is 0.098， subsample is 0.7
SVR	kernel is’rbf’， penalty parameter C is 15， gamma is 0.025， epsilon is 0.15
Lyzenga	a₀=1.233， a₁=10.681， a₂=-11.804， a₃=-0.105
Stumpf	m₀=-51.942， m₁=57.117

Table 2. Overall accuracy comparisons of models
View table
Table 2. Overall accuracy comparisons of models
Model RMSE /m MAE /m MRE /%
RF 1.30 0.98 13.93
CARF 1.13 0.84 11.79
XGB 1.19 0.87 11.93
CAXGB 1.11 0.81 11.05
SVR 1.32 0.98 15.26
CASVR 1.22 0.91 14.15
Lyzenga 1.42 1.04 37.20
Stumpf 1.49 1.13 38.63

Table 3. Comparisons of model accuracy in different water depth ranges

View table

Table 3. Comparisons of model accuracy in different water depth ranges

Model	（0， 5］ m		（5， 10］ m		（10， 15］ m		（15， 25］ m
Model	RMSE /m	MAE /m	RMSE /m	MAE /m	RMSE /m	MAE /m	RMSE /m	MAE /m
RF	0.93	0.66	1.15	0.89	1.26	0.95	1.71	1.38
CARF	0.77	0.54	0.98	0.74	1.14	0.87	1.47	1.16
XGB	0.67	0.47	1.01	0.74	1.23	0.94	1.59	1.22
CAXGB	0.67	0.46	0.94	0.69	1.15	0.88	1.45	1.12
SVR	0.93	0.64	1.11	0.85	1.28	0.99	1.80	1.38
CASVR	0.89	0.62	1.05	0.80	1.23	0.95	1.56	1.22
Lyzenga	1.34	0.83	1.07	0.81	1.31	1.01	1.91	1.51
Stumpf	1.28	0.92	1.06	0.83	1.41	1.11	2.06	1.64

Tools

Get Citation

Copy Citation Text

Jinhui Shi, Bingyi Liu, Peizhi Zhu, Qi Zhou, Qiuyue Xu, Yan He. Chromaticity Angle‑Assisted Remote Sensing Inversion Method for Shallow Water Depth[J]. Acta Optica Sinica, 2025, 45(12): 1201009

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites