Aerosol Optical Depth Retrieval Based on Neural Network Model Using Particulate Observing Scanning Polarimeter Data

Chenyu Yang; Xiao Liu; Honglian Huang; Zhuoyi Chen; Rufang Ti; Xiaobing Sun

doi:10.3788/AOS250438

Acta Optica Sinica, Volume. 45, Issue 12, 1228011(2025)

Aerosol Optical Depth Retrieval Based on Neural Network Model Using Particulate Observing Scanning Polarimeter Data

Chenyu Yang^1,2, Xiao Liu¹, Honglian Huang^1、*, Zhuoyi Chen³, Rufang Ti¹, and Xiaobing Sun¹

¹Key Laboratory of Optical Calibration and Characterization, Anhui Institute of Optics and Fine Mechanics, Hefei Institutes of Physical Science, Chinese Academy of Sciences, Hefei 230031, Anhui , China

²University of Science and Technology of China, Hefei 230026, Anhui , China

³China Academy of Space Technology, Beijing 100094, China

show less

Abstract Get PDF(in Chinese)

Figures & Tables(13)

Fig. 1. Neural network structure for aerosol inversion

Download full size

Fig. 2. Sensitivity analysis of different parameters

Download full size

Fig. 3. Neural network training process

Download full size

Fig. 4. Comparison of AOD inversion results of POSP with AERONET products

Download full size

Fig. 5. Comparisons of AOD inversion results of POSP with AERONET products on typical regions. (a) Beijing region; (b) Baotou region; (c) Taiwan region; (d) Hong Kong region

Download full size

Fig. 6. Comparison of AOD inversion results of POSP and MODIS products in Beijing-Tianjin-Hebei region on June 10th, 2024. (a) POSP AOD; (b) MODIS AOD; (c) scatterplot comparing POSP AOD and MODIS AOD

Download full size

Fig. 7. Comparison of AOD inversion results of POSP and MODIS products in Hefei region on January 12th, 2024. (a) POSP AOD; (b) MODIS AOD; (c) scatterplot comparing POSP AOD and MODIS AOD

Download full size

Fig. 8. Comparison of AOD inversion results of POSP and MODIS products in Taiwan region on February 14th, 2024. (a) POSP AOD; (b) MODIS AOD; (c) scatterplot comparing POSP AOD and MODIS AOD

Download full size

Table 1. Specifications of the POSP sensor onboard GF5-02 satellite

View table

Table 1. Specifications of the POSP sensor onboard GF5-02 satellite

Band number	Central wavelength /nm	Spectral bandwidth/nm	SNR	Usage
1	380±3	20±3	333.04	—
2	410±3	20±3	967.61	√
3	443±3	20±3	1517.65	√
4	490±3	20±3	1599.73	√
5	670±5	20±3	2029.47	√
6	865±5	40±5	2927.03	√
7	1380±5	40±5	3585.42	—
8	1610±15	60±10	3914.96	√
9	2250±15	80±10	482.55	√

Table 2. AERONET site data for training and validation

View table

Table 2. AERONET site data for training and validation

AERONET site	Longitude	Latitude	Date range
Beijing	116.381	39.977	2021-11—2022-11
AOE_Baotou	109.629	40.852	2021-11—2022-11
Hefei	117.162	31.905	2021-11—2022-11
Chen-Kung_Univ	120.205	22.993	2021-11—2022-11
XiangHe	116.962	39.754	2021-11—2022-11
Hong_Kong_PolyU	114.180	22.303	2021-11—2022-11
TASA_Taiwan	121.001	24.784	2021-11—2022-11
Taipei_CWB	121.538	25.015	2021-11—2022-11

Table 3. Aerosol model parameters

View table

Table 3. Aerosol model parameters

Aerosol type	FMF_V	$F M F$	$r_{e f f}^{f} / μ m$	$r_{e f f}^{c} / μ m$	$v_{e f f}^{f} / μ m$	$v_{e f f}^{c} / μ m$	$m_{r}$	$m_{i}$
Urban-industrial and mixed	0.520	0.780	0.12	2.72	0.43	0.63	1.47	0.0140
Biomass burning	0.570	0.762	0.12	3.22	0.40	0.73	1.51	0.0210
Desert dust and oceanic	0.033	0.674	0.12	1.90	0.49	0.63	1.48	0.0025

Table 4. Simulation parameters

View table

Table 4. Simulation parameters

Parameter	Value	Number
SZA /（°）	0， 10， 20， 30， 40， 50， 60	7
VZA /（°）	0， 8， 16， 24， 32， 40， 48， 56	8
RAA /（°）	0， 12， 24， 36， 48， 60， 72， 84， 96， 108， 120， 132， 144， 156， 168	15
AOD 550	0， 0.1， 0.2， 0.4， 0.6， 0.8， 1.0， 1.4， 1.8， 2.2， 2.6， 3.0	12
ALH /m	1000， 2000， 3000	3
Sr	0， 0.02， 0.05， 0.10， 0.15， 0.20， 0.25， 0.30， 0.35， 0.40， 0.50， 0.60	12

Table 5. Data used for inversion validation
View table
Table 5. Data used for inversion validation
Longitude range Latitude range Date
114‒120 37‒42 2024-06-10
115‒119 29‒33 2024-01-12
120‒122 22‒25.5 2024-02-14

Tools

Get Citation

Copy Citation Text

Chenyu Yang, Xiao Liu, Honglian Huang, Zhuoyi Chen, Rufang Ti, Xiaobing Sun. Aerosol Optical Depth Retrieval Based on Neural Network Model Using Particulate Observing Scanning Polarimeter Data[J]. Acta Optica Sinica, 2025, 45(12): 1228011

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites