A Method for Hyperspectral Reflectance Reconstruction from Automatic Observation with Multispectral Radiometer

Zhihong Ma; Lingling Ma; Yaokai Liu; Yongguang Zhao; Ning Wang; Chuanrong Li; Lingli Tang

doi:10.3788/AOS201939.0728004

Acta Optica Sinica, Volume. 39, Issue 7, 0728004(2019)

A Method for Hyperspectral Reflectance Reconstruction from Automatic Observation with Multispectral Radiometer

Zhihong Ma^1,2, Lingling Ma^1、*, Yaokai Liu¹, Yongguang Zhao¹, Ning Wang¹, Chuanrong Li¹, and Lingli Tang¹

¹ Key Laboratory of Quantitative Remote Sensing Information Technology, Academy of Opto-Electronics, Chinese Academy of Sciences, Beijing 100094, China

² School of Electronic, Electrical and Communication Engineering, University of Chinese Academy of Sciences, Beijing 100049, China

show less

Abstract Get PDF(in Chinese)

Figures & Tables(12)

Fig. 1. Flow of spectral reconstruction for multispectral reflectance

Download full size

Fig. 2. Automatic observation system of the Baotou calibration site. (a) Overview of artificial targets in calibration site; (b) multispectral radiometer; (c) sunphotometer CE318

Download full size

Fig. 3. Spectral response function of multispectral radiometer

Download full size

Fig. 5. Difference between reconstructed hyperspectral reflectance and measured reflectance of SVC

Download full size

Fig. 6. Acquired Sentinel-2 images. (a) Sentinel-2B (2018-09-21, RAOT@550 nm=0.20, RWVC=0.39); (b) Sentinel-2A (2018-10-29, RAOT@550 nm=0.07, RWVC=0.56)

Download full size

Fig. 7. Spectral response function of Sentinel-2. (a) Sentinel-2A; (b) Sentinel-2B

Download full size

Table 1. Filter parameters of multispectral radiometer

View table

Table 1. Filter parameters of multispectral radiometer

Channel No.	Central wavelengthλ /nm	FWHM ${B_{λ}}_{}$ /nm	Transmission ${τ_{λ}}_{}$ /%
1	439.72	10.48	≥37
2	499.79	10.27	≥50
3	679.77	10.13	≥50
4	869.60	10.96	≥50
5	1015.50	9.44	≥40
6	1298.63	11.05	≥40
7	1549.25	10.87	≥50
8	1649.23	12.52	≥50

Table 2. Accuracy of hyperspectral reflectance determination (October 30th)
View table
Table 2. Accuracy of hyperspectral reflectance determination (October 30th)
Time Average relative error /% RMSE
11:00 -0.989 0.005
12:00 -0.755 0.005
13:00 -0.911 0.005
14:00 -0.889 0.003

Table 3. Comparison of onboard observed and predicted at-sensor radiance of Sentinel-2 based on the reconstructed hyperspectral reflectance

View table

Table 3. Comparison of onboard observed and predicted at-sensor radiance of Sentinel-2 based on the reconstructed hyperspectral reflectance

Image	Band	Observed radiance /(W·m^-2·sr^-1·μm^-1)	Spectral reconstruction results
			Simulated radiance /(W·m^-2·sr^-1·μm^-1)					Relative error /%
			Considering BRDF		Without considering BRDF			Considering BRDF		Without considering BRDF
			Sentinel-2B2018-09-21	Blue	82.51		85.17		86.03		-3.12	-4.09
Green	78.12			77.64		78.47			0.62	-0.45
Red	70.87			67.13		68.12			5.56	4.04
NIR	50.95			46.66		47.06			9.19	8.27
Sentinel-2A2018-10-29	Blue	60.52		62.18		62.94			-2.66	-3.84
	Green	56.37		56.08		56.71			0.52	-0.60
	Red	52.19		49.65		50.25			5.12	3.86
	NIR	36.89		33.55		33.83			9.94	9.05

Table 4. Uncertainty factors and uncertainty of ground spectral extension results

View table

Table 4. Uncertainty factors and uncertainty of ground spectral extension results

Influencing factor	Uncertainty /%
Calibration error of multispectral radiometer	2.00
Downward irradiance due to AOT error	0.24
Downward irradiance due to WVC error	0.09
Solar irradiance model	1.00
MODTRAN model	2.00
BRDF model	1.42
Reconstruction coefficient	0.23

Table 5. Uncertainty analysis of Sentinel-2 radiometric calibration based on the reconstructed hyperspectral reflectance

View table

Table 5. Uncertainty analysis of Sentinel-2 radiometric calibration based on the reconstructed hyperspectral reflectance

Uncertainty contributor	Uncertainty /%	Uncertainty of simulated TOA radiance /%
Uncertainty contributor	Uncertainty /%	Blue	Green	Red	NIR
Reconstructed hyperspectral reflectance of target	3.34	2.12	2.57	2.90	3.09
Surface homogeneity	1.39	0.90	1.09	1.23	1.31
Surface BRDF characteristic	1.50	0.95	1.16	1.30	1.39
AOT	5.50	0.17	0.32	0.39	0.37
WVC	10.00	0.01	0.01	0.03	0.23
MODTRAN model	2.00	2.00	2.00	2.00	2.00
Solar irradiance model	1.00	1.00	1.00	1.00	1.00
Overall uncertainty		3.35	3.77	4.10	4.29

Tools

Get Citation

Copy Citation Text

Zhihong Ma, Lingling Ma, Yaokai Liu, Yongguang Zhao, Ning Wang, Chuanrong Li, Lingli Tang. A Method for Hyperspectral Reflectance Reconstruction from Automatic Observation with Multispectral Radiometer[J]. Acta Optica Sinica, 2019, 39(7): 0728004

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites

Paper Information

Category: Remote Sensing and Sensors

Received: Feb. 1, 2019

Accepted: Mar. 21, 2019

Published Online: Jul. 16, 2019

The Author Email: Ma Lingling (llma@aoe.ac.cn)

DOI:10.3788/AOS201939.0728004

Topics

laser devices and laser physics

Lasers and Laser Optics

Laser physics

laser manufacturing

Instrumentation, Measurement and Metrology

Table 1. Filter parameters of multispectral radiometer

Table 1. Filter parameters of multispectral radiometer

Table 2. Accuracy of hyperspectral reflectance determination (October 30th)

Table 2. Accuracy of hyperspectral reflectance determination (October 30th)

Table 3. Comparison of onboard observed and predicted at-sensor radiance of Sentinel-2 based on the reconstructed hyperspectral reflectance

Table 3. Comparison of onboard observed and predicted at-sensor radiance of Sentinel-2 based on the reconstructed hyperspectral reflectance

Table 4. Uncertainty factors and uncertainty of ground spectral extension results

Table 4. Uncertainty factors and uncertainty of ground spectral extension results

Table 5. Uncertainty analysis of Sentinel-2 radiometric calibration based on the reconstructed hyperspectral reflectance

Table 5. Uncertainty analysis of Sentinel-2 radiometric calibration based on the reconstructed hyperspectral reflectance