scholarly journals Deep Neural Networks for Aerosol Optical Depth Retrieval

Atmosphere ◽  
2022 ◽  
Vol 13 (1) ◽  
pp. 101
Author(s):  
Renee Zbizika ◽  
Paulina Pakszys ◽  
Tymon Zielinski

Aerosol Optical Depth (AOD) is a measure of the extinction of solar radiation by aerosols in the atmosphere. Understanding the variations of global AOD is necessary for precisely determining the role of aerosols. Arctic warming is partially caused by aerosols transported from vast distances, including those released during biomass burning events (BBEs). However, measuring AODs is challenging, typically requiring active LIDAR systems or passive sun photometers. Both are limited to cloud-free conditions; sun photometers provide only point measurements, thus requiring more spatial coverage. A more viable method to obtain accurate AOD may be found through machine learning. This study uses DNNs to estimate Svalbard’s AODs using a minimal set of meteorological parameters (temperature, air mass, water vapor, wind speed, latitude, longitude, and time of year). The mean absolute error (MAE) between predicted and true data was 0.00401 for the entire set and 0.0079 for the validation set. It was then shown that the inclusion of BBE data improves predictions by 42.167%. It was demonstrated that AODs may be accurately estimated without the use of expensive instrumentation, using machine learning and minimal data. Similar models may be developed for other regions, allowing immediate improvement of current meteorological models.

2021 ◽  
Vol 13 (15) ◽  
pp. 3027
Author(s):  
Saleem Ibrahim ◽  
Martin Landa ◽  
Ondřej Pešek ◽  
Karel Pavelka ◽  
Lena Halounova

The recent COVID-19 pandemic affected various aspects of life. Several studies established the consequences of pandemic lockdown on air quality using satellite remote sensing. However, such studies have limitations, including low spatial resolution or incomplete spatial coverage. Therefore, in this paper, we propose a machine learning-based scheme to solve the pre-mentioned limitations by training an optimized space-time extra trees model for each year of the study period. The results have shown that our trained models reach a prediction accuracy up to 95% when predicting the missing values in the MODIS MCD19A2 Aerosol Optical Depth (AOD) product. The outcome of the mentioned scheme was a geo-harmonized atmospheric dataset for aerosol optical depth at 550 nm with 1 km spatial resolution and full coverage over Europe. As an application, we used the proposed machine learning based prediction approach in AOD levels analysis. We compared the mean AOD levels between the lockdown period from March to June in 2020 and the mean AOD values of the same period for the past 5 years. We found that AOD levels dropped over most European countries in 2020 but increased in several eastern and western countries. The Netherlands had the most significant average decrease in AOD levels (19%), while Spain had the highest average increase (10%). Moreover, we analyzed the relationship between the relative percentage difference of AOD and four meteorological variables. We found a positive correlation between AOD and relative humidity and a negative correlation between AOD and wind speed. The value of the proposed prediction scheme is further emphasized by taking into consideration that the reconstructed dataset can be used for future air quality studies concerning Europe.


2005 ◽  
Vol 23 (4) ◽  
pp. 1093-1101 ◽  
Author(s):  
S. N. Tripathi ◽  
Sagnik Dey ◽  
A. Chandel ◽  
S. Srivastava ◽  
Ramesh P. Singh ◽  
...  

Abstract. The Moderate Resolution Imaging Spectroradiometer (MODIS) onboard EOS Terra measures global aerosol optical depth and optical properties since 2000. MODIS aerosol products are freely available and are being used for numerous studies. In this paper, we present a comparison of aerosol optical depth (AOD) retrieved from MODIS with Aerosol Robotic Network (AERONET) data for the year 2004 over Kanpur, an industrial city lying in the Ganga Basin in the northern part of India. AOD retrieved from MODIS (τaMODIS) at 0.55µm wavelength has been compared with the AERONET derived AOD (τaAERONET), within an optimum space-time window. Although the correlation between τaMODIS and τaAERONET during the post-monsoon and winter seasons (R2~0.71) is almost equal to that during the pre-monsoon and monsoon seasons (R2~0.72), MODIS is found to overestimate AOD during the pre-monsoon and monsoon period (characterized by severe dust loading) and underestimate during the post-monsoon and winter seasons. The absolute difference between τaMODIS and τaAERONET is found to be low (0.12±0.11) during the non-dust loading season and much higher (0.4±0.2) during dust-loading seasons. The absolute error in τaMODIS is found to be about ~25% of the absolute values of τaMODIS. Our comparison shows the importance of modifying the existing MODIS algorithm during the dust-loading seasons, especially in the Ganga Basin in northern part of India.


2013 ◽  
Vol 122 ◽  
pp. 298-309 ◽  
Author(s):  
Fernando Castro Videla ◽  
Francesca Barnaba ◽  
Federico Angelini ◽  
Pablo Cremades ◽  
Gian Paolo Gobbi

2021 ◽  
Vol 13 (3) ◽  
pp. 415
Author(s):  
Yangyang Jin ◽  
Zengzhou Hao ◽  
Jian Chen ◽  
Dong He ◽  
Qingjiu Tian ◽  
...  

Aerosol is an essential parameter for assessing the atmospheric environmental quality, and accurate monitoring of the aerosol optical depth (AOD) is of great significance in climate research and environmental protection. Based on Landsat 8 Operational Land Imager (OLI) images and MODIS09A1 surface reflectance products under clear skies with limited cloud cover, we retrieved the AODs in Nanjing City from 2017 to 2018 using the combined Dark Target (DT) and Deep Blue (DB) methods. The retrieval accuracy was validated by in-situ CE-318 measurements and MOD04_3K aerosol products. Furthermore, we analyzed the spatiotemporal distribution of the AODs and discussed a case of high AOD distribution. The results showed that: (1) Validated by CE-318 and MOD04_3K data, the correlation coefficient (R), root mean square error (RMSE), and mean absolute error (MAE) of the retrieved AODs were 0.874 and 0.802, 0.134 and 0.188, and 0.099 and 0.138, respectively. Hence, the combined DT and DB algorithms used in this study exhibited a higher performance than the MOD04_3K-obtained aerosol products. (2) Under static and stable meteorological conditions, the average annual AOD in Nanjing was 0.47. At the spatial scale, the AODs showed relatively high values in the north and west, low in the south, and the lowest in the center. At the seasonal scale, the AODs were highest in the summer, followed by spring, winter, and autumn. Moreover, changes were significantly higher in the summer than in the other three seasons, with little differences among spring, autumn, and winter. (3) Based on the spatial and seasonal characteristics of the AOD distribution in Nanjing, a case of high AOD distribution caused by a large area of external pollution and local meteorological conditions was discussed, indicating that it could provide extra details of the AOD distribution to analyze air pollution sources using fine spatial resolution like in the Landsat 8 OLI.


2021 ◽  
Vol 13 (20) ◽  
pp. 4140
Author(s):  
Hao Lin ◽  
Siwei Li ◽  
Jia Xing ◽  
Jie Yang ◽  
Qingxin Wang ◽  
...  

Recent studies have shown that the high-resolution satellite Landsat-8 has the capability to retrieve aerosol optical depth (AOD) over urban areas at a 30 m spatial resolution. However, its long revisiting time and narrow swath limit the coverage and frequency of the high resolution AOD observations. With the increasing number of Earth observation satellites launched in recent years, combining the observations of multiple satellites can provide higher temporal-spatial coverage. In this study, a fusing retrieval algorithm is developed to retrieve high-resolution (30 m) aerosols over urban areas from Landsat-8 and Sentinel-2 A/B satellite measurements. The new fusing algorithm was tested and evaluated over Beijing city and its surrounding area in China. The validation results show that the retrieved AODs show a high level of agreement with the local urban ground-based Aerosol Robotic Network (AERONET) AOD measurements, with an overall high coefficient of determination (R2) of 0.905 and small root mean square error (RMSE) of 0.119. Compared with the operational AOD products processed by the Landsat-8 Surface Reflectance Code (LaSRC-AOD), Sentinel Radiative Transfer Atmospheric Correction code (SEN2COR-AOD), and MODIS Collection 6 AOD (MOD04) products, the AOD retrieved from the new fusing algorithm based on the Landsat-8 and Sentinel-2 A/B observations exhibits an overall higher accuracy and better performance in spatial continuity over the complex urban area. Moreover, the temporal resolution of the high spatial resolution AOD observations was greatly improved (from 16/10/10 days to about two to four days over globe land in theory under cloud-free conditions) and the daily spatial coverage was increased by two to three times compared to the coverage gained using a single sensor.


2012 ◽  
Vol 12 (4) ◽  
pp. 10461-10492 ◽  
Author(s):  
Y. Xue ◽  
H. Xu ◽  
L. Mei ◽  
J. Guang ◽  
J. Guo ◽  
...  

Abstract. Agricultural biomass burning (ABB) in Central and East China occurs every year from May to October and peaks in June. The biomass burning event in June 2007 was very strong. During the period from 26 May to 16 June 2007, ABB occurred mainly in Anhui, Henan, Jiangsu and Shandong provinces. A comprehensive set of aerosol optical depth (AOD) data, produced by a merger of AOD product data from the Moderate Resolution Imaging Spectroradiometer (MODIS) and the Multiangle Imaging Spectroradiometer (MIRS), is used to study the spatial and temporal distribution of agricultural biomass aerosols in Central and East China combining with ground observations from both AErosol RObotic NETwork (AERONET) and China Aerosol Remote Sensing NETwork (CARSNET) measurements. We compared merged AOD data with single-sensor single-algorithm AOD data (MODIS Dark Target AOD data, MODIS Deep Blue AOD data, SRAP-MODIS AOD data and MISR AOD data). In this comparison, we found merged AOD products can improve the quality of AOD products from single-sensor single-algorithm data sets by expanding the spatial coverage of the study area and keeping the statistical confidence in AOD parameters. There existed high correlation (0.8479) between the merged AOD data and AERONET measurements. Our merged AOD data make use of synergetic information conveyed in all of the available satellite data. The merged AOD data were used for the analysis of the biomass burning event from 26 May to 16 June 2007 together with meteorological data. The merged AOD products and the ground observations from China suggest that biomass burning in Central and East China has had great impact on AOD over China. Influenced by this ABB, the highest AOD value in Beijing on 12 June 2007 reached 5.71.


Sign in / Sign up

Export Citation Format

Share Document