Fractional Differential-Based Hyperspectral Inversion of Soil Organic Matter Content

Wuyao Li; Mamat Sawut; Maihemuti Balati

doi:10.3788/LOP220715

Journals >Laser & Optoelectronics Progress >Volume 60 >Issue 7 >Page 0730005 > Article

Laser & Optoelectronics Progress
Vol. 60, Issue 7, 0730005 (2023)

Fractional Differential-Based Hyperspectral Inversion of Soil Organic Matter Content

Wuyao Li¹, Mamat Sawut^{1、2、3、*}, and Maihemuti Balati^1、2、3

Author Affiliations

¹College of Geography and Remote Sensing Science, Xinjiang University, Urumqi 830046, Xinjiang, China

²Xinjiang Key Laboratory of Oasis Ecology, Urumqi 830046, Xinjiang, China

³Key Laboratory of Smart City and Environment Modelling of Higher Education Institute, Urumqi 830046, Xinjiang, China

show less

DOI: 10.3788/LOP220715 Cite this Article Set citation alerts

Wuyao Li, Mamat Sawut, Maihemuti Balati. Fractional Differential-Based Hyperspectral Inversion of Soil Organic Matter Content[J]. Laser & Optoelectronics Progress, 2023, 60(7): 0730005 Copy Citation Text

show less

Fig. 1. Geographical position of study area and distribution of sampling

Download full size

Fig. 2. Reflectance curves of soil with different organic matter contents

Download full size

Fig. 3. Correlation coefficient between SOM content and spectral reflectance transform form. (a) 1-order differential; (b) 1.2-order differential; (c) 1.4-order differential; (d) 1.6-order differential; (e) 1.8-order differential; (f) 2-order differential

Download full size

Fig. 4. Optimal mathematical transformation form of SOM results in the three models. (a) 1-order differential of SVR model; (b) 1-order differential of PLSR model; (c) 1.2-order differential of RF model

Download full size

Sample type	Quantity	Max	Min	Mean	Standard deviation	CV /%
Calibration	49	25.416	2.758	11.602	5.436	46.86
Validation	24	24.133	2.742	12.531	5.810	46.37
Totality	73	25.416	2.742	11.959	5.601	46.84

Table 1. SOM content statistical characteristics of study area samples

Spectral transformation	Characteristic band /nm	Number of wavelengths	r
0	/	0	/
0.2	/	0	/
0.4	/	0	/
0.6	/	0	/
0.8	399、435	2	0.293、0.316
1	588、597、630、648、675、792、855、1013	8	0.308、0.323、0.290、-0.286、-0.308、-0.350、-0.291、0.290
1.2	407、464、588、597、648、671、675、723、792、1013	10	0.285、-0.312、0.290、0.319、-0.310、0.280、-0.287、-0.293、-0.368、0.353
1.4	407、425、464、597、648、671、723、775、792、965、1013	11	0.316、-0.294、-0.335、0.315、-0.308、0.287、-0.307、0.282、-0.372、-0.278、0.283
1.6	387、407、425、464、597、648、671、676、723、775、792、965	12	0.296、0.343、-0.283、-0.343、0.305、-0.300、0.285、0.278、-0.311、0.298、-0.366、-0.278
1.8	387、407、464、597、648、671、676、723、775、792、797	11	0.309、0.364、-0.347、0.288、-0.291、0.278、0.289、-0.307、0.309、-0.353、0.277
2	387、407、464、648、676、723、775、792、797	9	0.303、0.377、-0.349、-0.282、0.293、-0.297、0.314、-0.336、0.278

Table 2. Characteristic bands of SOM under fractional differentiation

Model	Sample type	Model formula	R²	RMSE	RPD
SVR_—1	Calibration set	y=0.76x+2.8583	0.83	2.29	2.43
SVR_—1.2		y=0.7074x+3.5983	0.73	2.84	1.96
SVR_—1.4		y=0.6341x+4.7963	0.67	3.19	1.74
SVR_—1.6		y=0.7051x+3.9563	0.71	3.01	1.85
SVR_—1.8		y=0.7537x+2.777	0.81	2.45	2.27
SVR_—2		y=0.7488x+2.5789	0.81	2.48	4.97
SVR_—1	Validation set	y=0.777x+3.0056	0.82	1.75	3.38
SVR_—1.2		y=0.8659x+1.9667	0.93	1.13	5.21
SVR_—1.4		y=0.897x+1.4525	0.97	0.79	7.52
SVR_—1.6		y=0.9153x+1.274	0.97	0.76	7.75
SVR_—1.8		y=0.9031x+1.5441	0.96	0.86	6.83
SVR_—2		y=0.8639x+2.2584	0.93	1.19	3.73
PLSR_—1	Calibration set	y=0.5492x+5.3037	0.55	4.24	1.31
PLSR_—1.2		y=0.4062x+6.9854	0.41	4.24	1.30
PLSR_—1.4		y=0.3676x+7.4394	0.37	4.37	1.27
PLSR_—1.6		y=0.3553x+7.5838	0.36	4.42	1.26
PLSR_—1.8		y=0.3697x+7.415	0.37	4.37	1.27
PLSR_—2		y=0.3697x+7.4153	0.37	4.37	1.05
PLSR_—1	Validation set	y=0.6591x+4.2127	0.71	2.18	2.71
PLSR_—1.2		y=0.7094x+3.5908	0.71	3.12	1.86
PLSR_—1.4		y=0.6275x+4.6023	0.63	2.47	2.39
PLSR_—1.6		y=0.65x+4.3244	0.65	2.39	2.47
PLSR_—1.8		y=0.6028x+4.9078	0.60	2.55	2.32
PLSR_—2		y=0.6137x+4.7739	0.61	2.52	2.09
RF_—1	Calibration set	y=0.6285x+4.3153	0.82	2.61	2.13
RF_—1.2		y=0.5551x+5.2584	0.92	2.60	2.14
RF_—1.4		y=0.5674x+5.1131	0.91	2.58	2.16
RF_—1.6		y=0.5544x+5.2686	0.88	2.70	2.06
RF_—1.8		y=0.5864x+4.9469	0.91	2.50	2.22
RF_—2		y=0.6319x+4.7115	0.80	2.71	2.05
RF_—1	Validation set	y=0.5873x+5.0345	0.79	2.07	2.85
RF_—1.2		y=0.6418x+4.4054	0.93	1.62	3.65
RF_—1.4		y=0.604x+4.953	0.92	1.77	3.34
RF_—1.6		y=0.5824x+5.2598	0.90	1.87	3.16
RF_—1.8		y=0.5654x+5.4647	0.90	1.93	3.07
RF_—2		y=0.6048x+4.639	0.86	1.88	3.14

Table 3. SOM content modeling and validation results

Wuyao Li, Mamat Sawut, Maihemuti Balati. Fractional Differential-Based Hyperspectral Inversion of Soil Organic Matter Content[J]. Laser & Optoelectronics Progress, 2023, 60(7): 0730005

Download Citation

Set citation alerts for the article

Tools

Set citation alerts for the article

Save the article for my favorites

Paper Information