Estimation of Soil Organic Carbon Content Based on Spectral Indices and Continuous Wavelet Transform

Mingchong Gong; Hong Wang; Lei Zhang; Jiujun Xiao; Jing Liu; Yandong Chen

doi:10.3788/LOP241216

Journals >Laser & Optoelectronics Progress >Volume 62 >Issue 3 >Page 0330005 > Article

Laser & Optoelectronics Progress
Vol. 62, Issue 3, 0330005 (2025)

Estimation of Soil Organic Carbon Content Based on Spectral Indices and Continuous Wavelet Transform

Mingchong Gong^1,*, Hong Wang¹, Lei Zhang², Jiujun Xiao³..., Jing Liu¹ and Yandong Chen¹|Show fewer author(s)

Author Affiliations

¹School of Mining, Guizhou University, Guiyang 550025, Guizhou , China

²Institute of Surveying and Mapping, Guizhou Geology and Mineral Exploration Bureau, Guiyang 550018, Guizhou , China

³Guizhou Institute of Mountain Resources, Guizhou Academy of Sciences, Guiyang 550001, Guizhou , China

show less

DOI: 10.3788/LOP241216 Cite this Article Set citation alerts

Mingchong Gong, Hong Wang, Lei Zhang, Jiujun Xiao, Jing Liu, Yandong Chen. Estimation of Soil Organic Carbon Content Based on Spectral Indices and Continuous Wavelet Transform[J]. Laser & Optoelectronics Progress, 2025, 62(3): 0330005 Copy Citation Text

show less

Fig. 1. Correlation between the original spectrum and SOC content

Download full size

Fig. 2. Correlation between SIs and SOC content. (a) DI ; (b) RI; (c) RDI; (d) NDI

Download full size

Fig. 3. Correlation between wavelet energy coefficient and SOC content

Download full size

$Simulated scatter plots of SOC mass fraction for validation set by 6 models with the highest accuracy. (a) SIs-RF; (b) CWT-RF; (c) SIs+CWT-RF; (d) SIs-BPANN; (e) CWT-BPANN; (f) SIs+CWT-BPANN$

Fig. 4. Simulated scatter plots of SOC mass fraction for validation set by 6 models with the highest accuracy. (a) SIs-RF; (b) CWT-RF; (c) SIs+CWT-RF; (d) SIs-BPANN; (e) CWT-BPANN; (f) SIs+CWT-BPANN

Download full size

Sample type	Quantity	Mass fraction /（g·kg^-1）
Sample type	Quantity	Maximum	Minimum	Mean	Standard deviation
Totality	61	78.40	6.40	13.45	11.00
Calibration	43	78.40	6.40	14.26	11.47
Validation	18	42.30	7.50	14.72	9.96

Table 1. Statistical characteristics of SOC mass fraction in different samples

Variable	Number of bands or features	Model	Calibration set		Validation set
Variable	Number of bands or features	Model	R²	RMSE /（g·kg^-1）	R²	RMSE /（g·kg^-1）	RPD
OR	2001	SVM	0.74	5.74	0.58	6.25	1.59
		RF	0.78	5.34	0.69	5.40	1.84
		BPANN	0.80	5.04	0.70	5.28	1.89
SIs	20	SVM	0.75	5.65	0.62	5.97	1.67
		RF	0.85	4.42	0.79	4.48	2.23
		BPANN	0.88	3.94	0.82	4.06	2.46
CWT	20	SVM	0.78	5.35	0.63	5.89	1.70
		RF	0.87	4.16	0.80	4.33	2.30
		BPANN	0.98	1.61	0.93	2.48	4.01
SIs+CWT	40	SVM	0.81	4.99	0.65	5.75	1.73
		RF	0.92	3.21	0.88	3.41	3.02
		BPANN	0.98	1.59	0.95	2.13	4.67

Table 2. Results of SOC mass fraction estimation models for calibration and validation sets

Download Citation

Set citation alerts for the article

Tools

Set citation alerts for the article

Save the article for my favorites

Paper Information

微信扫一扫：分享

微信扫一扫：分享