The Generative AOD Adversarial Neural Network (GANN) is a neural network model used for satellite aerosol optical depth (AOD) retrieval

Citation Author(s):
Yulong
Fan
Shandong University of Science and Technology
Submitted by:
yulong fan
Last updated:
Sat, 01/20/2024 - 08:34
DOI:
10.21227/3d26-m978
License:
54 Views
Categories:
Keywords:
0
0 ratings - Please login to submit your rating.

Abstract 

The current quantitative retrieval of Aerosol Optical Depth (AOD) typically uses Top-of-atmosphere (TOA) reflectance data obtained by radiometric calibration. Errors can be introduced during the conversion of DN values to TOA reflectance, affecting the retrieval of AOD. Especially when the surface reflectance is relatively high, the conversion error will bring significant errors to the AOD retrieval, as in such cases, the contribution of aerosols to the radiation received by satellite sensors is relatively small. The deep learning (DL) technology enables the use of DN values for AOD retrieval, bypassing the conversion errors associated with TOA reflectance, since it relies on a sample learning approach instead of radiative transfer calculations. Hence, we propose a novel AOD retrieval algorithm using DN values and a GANN model that we specifically designed for quantitative parameters retrieval. To ensure the number and representativeness of training samples, a total of 37,103 samples from different underlying surfaces from 2014 to 2018 were obtained using a space-time matching strategy. K-cross validation demonstrates that employing DN values for AOD retrieval yields higher accuracy compared to using TOA reflectance. Independent 2019 data were used to assess GANN, MCD19A2 and MOD04_3K aerosol products. Our model showed superior performance compared to the other products in terms of a higher correlation (R=0.8184) with AERONET AOD and obtaining more reliable retrievals (7601). The sensitivity experiments revealed that GANN exhibits limitations in areas characterized by high aerosol loads and heterogeneity. 

Instructions: 

College of Geodesy and Geomatics, Shandong University of Science and Technology, Shandong Qingdao 266590, China

Dataset Files

    Files have not been uploaded for this dataset