Estimation of Aerosol Optical Depth at 30 m Resolution Using Landsat Imagery and Machine Learning
Current remote sensing-based aerosol optical depth (AOD) products have coarse spatial resolutions, which are useful for studies at continental and global scales, but unsatisfactory for local scale applications, such as urban air pollution monitoring. In this study, we investigated the possibility of...
Published in: | Remote Sensing |
---|---|
Main Authors: | , , , , , , , |
Format: | Text |
Language: | English |
Published: |
Multidisciplinary Digital Publishing Institute
2022
|
Subjects: | |
Online Access: | https://doi.org/10.3390/rs14051053 |
id |
ftmdpi:oai:mdpi.com:/2072-4292/14/5/1053/ |
---|---|
record_format |
openpolar |
spelling |
ftmdpi:oai:mdpi.com:/2072-4292/14/5/1053/ 2023-08-20T03:59:12+02:00 Estimation of Aerosol Optical Depth at 30 m Resolution Using Landsat Imagery and Machine Learning Tianchen Liang Shunlin Liang Linqing Zou Lin Sun Bing Li Hao Lin Tao He Feng Tian agris 2022-02-22 application/pdf https://doi.org/10.3390/rs14051053 EN eng Multidisciplinary Digital Publishing Institute https://dx.doi.org/10.3390/rs14051053 https://creativecommons.org/licenses/by/4.0/ Remote Sensing; Volume 14; Issue 5; Pages: 1053 aerosol optical depth machine learning Landsat high resolution Text 2022 ftmdpi https://doi.org/10.3390/rs14051053 2023-08-01T04:14:27Z Current remote sensing-based aerosol optical depth (AOD) products have coarse spatial resolutions, which are useful for studies at continental and global scales, but unsatisfactory for local scale applications, such as urban air pollution monitoring. In this study, we investigated the possibility of using Landsat imagery to develop high-resolution AOD estimations at 30 m based on machine learning algorithms. We assessed the performance of six machine learning algorithms, including Extreme Gradient Boosting, Random Forest, Cascade Random Forest, Gradient Boosted Decision Trees, Extremely Randomized Trees, and Multiple Linear Regression. To obtain accurate AOD estimations, we used prior knowledge from multiple sources as inputs to the machine learning models, including the Global Land Surface Satellite (GLASS) albedo, the 1-km AOD product from MODIS data using the Multi-Angle Implementation of Atmospheric Correction (MAIAC) algorithm, and meteorological and surface elevation data. A total of 13,624 AOD measurements from Aerosol Robotic Network (AERONET) sites were used for model training and validation. We found that all six algorithms exhibited good performance, with R2 values ranging from 0.73 to 0.78 and AOD root-mean-square errors (RMSE) ranging from 0.089 to 0.098. The extremely randomized trees algorithm, however, demonstrated marginally superior performance as compared to the other algorithms; hence, it was used to produce AOD estimates at a 30 m resolution for one Landsat scene coving Beijing in 2013–2019. Through a comparison with overlapping AERONET observations, a high level of accuracy was achieved, with an R2 = 0.889 and an RMSE = 0.156. Our method can be potentially used to generate a global high-resolution AOD dataset based on Landsat imagery. Text Aerosol Robotic Network MDPI Open Access Publishing Remote Sensing 14 5 1053 |
institution |
Open Polar |
collection |
MDPI Open Access Publishing |
op_collection_id |
ftmdpi |
language |
English |
topic |
aerosol optical depth machine learning Landsat high resolution |
spellingShingle |
aerosol optical depth machine learning Landsat high resolution Tianchen Liang Shunlin Liang Linqing Zou Lin Sun Bing Li Hao Lin Tao He Feng Tian Estimation of Aerosol Optical Depth at 30 m Resolution Using Landsat Imagery and Machine Learning |
topic_facet |
aerosol optical depth machine learning Landsat high resolution |
description |
Current remote sensing-based aerosol optical depth (AOD) products have coarse spatial resolutions, which are useful for studies at continental and global scales, but unsatisfactory for local scale applications, such as urban air pollution monitoring. In this study, we investigated the possibility of using Landsat imagery to develop high-resolution AOD estimations at 30 m based on machine learning algorithms. We assessed the performance of six machine learning algorithms, including Extreme Gradient Boosting, Random Forest, Cascade Random Forest, Gradient Boosted Decision Trees, Extremely Randomized Trees, and Multiple Linear Regression. To obtain accurate AOD estimations, we used prior knowledge from multiple sources as inputs to the machine learning models, including the Global Land Surface Satellite (GLASS) albedo, the 1-km AOD product from MODIS data using the Multi-Angle Implementation of Atmospheric Correction (MAIAC) algorithm, and meteorological and surface elevation data. A total of 13,624 AOD measurements from Aerosol Robotic Network (AERONET) sites were used for model training and validation. We found that all six algorithms exhibited good performance, with R2 values ranging from 0.73 to 0.78 and AOD root-mean-square errors (RMSE) ranging from 0.089 to 0.098. The extremely randomized trees algorithm, however, demonstrated marginally superior performance as compared to the other algorithms; hence, it was used to produce AOD estimates at a 30 m resolution for one Landsat scene coving Beijing in 2013–2019. Through a comparison with overlapping AERONET observations, a high level of accuracy was achieved, with an R2 = 0.889 and an RMSE = 0.156. Our method can be potentially used to generate a global high-resolution AOD dataset based on Landsat imagery. |
format |
Text |
author |
Tianchen Liang Shunlin Liang Linqing Zou Lin Sun Bing Li Hao Lin Tao He Feng Tian |
author_facet |
Tianchen Liang Shunlin Liang Linqing Zou Lin Sun Bing Li Hao Lin Tao He Feng Tian |
author_sort |
Tianchen Liang |
title |
Estimation of Aerosol Optical Depth at 30 m Resolution Using Landsat Imagery and Machine Learning |
title_short |
Estimation of Aerosol Optical Depth at 30 m Resolution Using Landsat Imagery and Machine Learning |
title_full |
Estimation of Aerosol Optical Depth at 30 m Resolution Using Landsat Imagery and Machine Learning |
title_fullStr |
Estimation of Aerosol Optical Depth at 30 m Resolution Using Landsat Imagery and Machine Learning |
title_full_unstemmed |
Estimation of Aerosol Optical Depth at 30 m Resolution Using Landsat Imagery and Machine Learning |
title_sort |
estimation of aerosol optical depth at 30 m resolution using landsat imagery and machine learning |
publisher |
Multidisciplinary Digital Publishing Institute |
publishDate |
2022 |
url |
https://doi.org/10.3390/rs14051053 |
op_coverage |
agris |
genre |
Aerosol Robotic Network |
genre_facet |
Aerosol Robotic Network |
op_source |
Remote Sensing; Volume 14; Issue 5; Pages: 1053 |
op_relation |
https://dx.doi.org/10.3390/rs14051053 |
op_rights |
https://creativecommons.org/licenses/by/4.0/ |
op_doi |
https://doi.org/10.3390/rs14051053 |
container_title |
Remote Sensing |
container_volume |
14 |
container_issue |
5 |
container_start_page |
1053 |
_version_ |
1774720557254705152 |