Deep Neural Networks for Aerosol Optical Depth Retrieval

Zbizika, Renee; Pakszys, Paulina; Zieliński, Tymon

doi:10.3390/atmos13010101

Cited by 10 publications

(3 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Various alternative techniques for AOD retrieval reported in the literature fall into four primary categories: (1) backward solving radiative transfer (RT) or clear-sky models using solar radiation measurements [24][25][26][27], (2) methodologies based on sunshine duration (SD) measurements [28][29][30], (3) image processing techniques using sky radiances from all-sky imagers [31][32][33][34], and (4) machine learning (ML) and deep learning algorithms employing various independent parameters as input features [35][36][37][38][39].…”

Section: Introductionmentioning

confidence: 99%

A Machine Learning Approach to Retrieving Aerosol Optical Depth Using Solar Radiation Measurements

Logothetis,

Salamalikis,

Kazantzidis

2024

Remote Sensing

View full text Add to dashboard Cite

Aerosol optical depth (AOD) constitutes a key parameter of aerosols, providing vital information for quantifying the aerosol burden and air quality at global and regional levels. This study demonstrates a machine learning strategy for retrieving AOD under cloud-free conditions based on the synergy of machine learning algorithms (MLAs) and ground-based solar irradiance data. The performance of the proposed methodology was investigated by applying different components of solar irradiance. In particular, the use of direct instead of global irradiance as a model feature led to better performance. The MLA-based AODs were compared to reference AERONET retrievals, which encompassed RMSE values between 0.01 and 0.15, regardless of the underlying climate and aerosol environments. Among the MLAs, artificial neural networks outperformed the other algorithms in terms of RMSE at 54% of the measurement sites. The overall performance of MLA-based AODs against AERONET revealed a high coefficient of determination (R2 = 0.97), MAE of 0.01, and RMSE of 0.02. Compared to satellite (MODIS) and reanalysis (MERRA-2 and CAMSRA) data, the MLA-AOD retrievals revealed the highest accuracy at all stations. The ML-AOD retrievals have the potential to expand and complement the AOD information in non-existing timeframes when solar irradiances are available.

show abstract

Section: Introductionmentioning

confidence: 99%

A Machine Learning Approach to Retrieving Aerosol Optical Depth Using Solar Radiation Measurements

Logothetis,

Salamalikis,

Kazantzidis

2024

Remote Sensing

View full text Add to dashboard Cite

show abstract

“…The accuracy and reliability of FCNN-retrieval parameters were all better than that of the official aerosol product. Zbizika et al [52] used deep neural network (DNN) to estimate Svalbard's AODs utilizing auxiliary data (temperature, air mass, water vapor, and wind speed) and obtained results close to the ground measurements. Lu et al [53] used global mid-low latitude ground AOD measurements to train the DNN model, which achieved high-precision AOD retrieval and the results had a high correlation with ground measurements.…”

Section: Introductionmentioning

confidence: 99%

Satellite Aerosol Optical Depth Retrieval Based on Fully Connected Neural Network (FCNN) and a Combine Algorithm of Simplified Aerosol Retrieval Algorithm and Simplified and Robust Surface Reflectance Estimation (SREMARA)

Fan

Sun

2023

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

View full text Add to dashboard Cite

Aerosol satellite retrieval can provide detailed aerosol information on a large scale, which becomes one of the main ways of global aerosol research. However, rapid and accrue aerosol retrieval by satellite is challenging, typically requiring radiation transfer models (RTMs) and surface reflectance (SR). An aerosol retrieval algorithm (SEMARA) combining the simplified aerosol retrieval algorithm and simplified and robust surface reflectance estimation can obtain local high-precision aerosol optical depth (AOD) without RTMs and SR datasets, while the method cannot perform large-scale and long-term aerosol retrieval. Hereby, a machine learning (ML) method based on the fully connected neural network (FCNN) and SEMARA was proposed. The new method optimizes the traditional sample construction of the ML and can achieve aerosol retrieval at a larger spatial and temporal scale. Moderate resolution imaging spectroradiometer data were applied to AOD retrieval on four typical regions globally. The AOD retrievals were validated using aerosol robotic network measurements in comparison to MOD04_3K AOD and the SEMARA. The accuracy validation indicators of the new method, in which the root-mean-square error (RMSE) was 0.109, mean absolute error (MAE) was 0.072, Pearson correlation coefficient (R) was 0.8983, and approximately 79.69% of the retrievals fell within the expected error (EE), performed better than MOD04_3K (RMSE = 0.1972 MAE = 0.1403, R = 0.7692 and Within EE = 55.24%) and the SEMARA method (RMSE = 0.2465 MAE = 0.1106, R = 0.0.5968 and Within EE = 72.85%) in all study regions, and the AOD retrievals can better reflect the spatial variation of AOD with better spatial continuity and coverage. Index Terms-Aerosol optical depth (AOD), fully connected neural network (FCNN), simplified aerosol retrieval algorithm (SARA), simplified and robust surface reflectance estimation (SREM).

show abstract

“…Therefore, an experimental quantification of BC pollution (and more generally, of total LAA) and associated HR together with its distribution in the Arctic area are main scientific targets to unravel the future of the Arctic from a climate change perspective [3,44]. Thus, specific studies of these phenomena are required [45].…”

Section: Introductionmentioning

confidence: 99%

Anthropic Settlements’ Impact on the Light-Absorbing Aerosol Concentrations and Heating Rate in the Arctic

Losi,

Markuszewski,

Rigler

et al. 2023

Atmosphere

Self Cite

View full text Add to dashboard Cite

Light-absorbing aerosols (LAA) impact the atmosphere by heating it. Their effect in the Arctic was investigated during two summer Arctic oceanographic campaigns (2018 and 2019) around the Svalbard Archipelago in order to unravel the differences between the Arctic background and the local anthropic settlements. Therefore, the LAA heating rate (HR) was experimentally determined. Both the chemical composition and high-resolution measurements highlighted substantial differences between the Arctic Ocean background (average eBC concentration of 11.7 ± 0.1 ng/m3) and the human settlements, among which the most impacting appeared to be Tromsø and Isfjorden (mean eBC of 99.4 ± 3.1 ng/m3). Consequently, the HR in Isfjorden (8.2 × 10−3 ± 0.3 × 10−3 K/day) was one order of magnitude higher than in the pristine background conditions (0.8 × 10−3 ± 0.9 × 10−5 K/day). Therefore, we conclude that the direct climate impact of local LAA sources on the Arctic atmosphere is not negligible and may rise in the future due to ice retreat and enhanced marine traffic.

show abstract

Deep Neural Networks for Aerosol Optical Depth Retrieval

Cited by 10 publications

References 14 publications

A Machine Learning Approach to Retrieving Aerosol Optical Depth Using Solar Radiation Measurements

A Machine Learning Approach to Retrieving Aerosol Optical Depth Using Solar Radiation Measurements

Satellite Aerosol Optical Depth Retrieval Based on Fully Connected Neural Network (FCNN) and a Combine Algorithm of Simplified Aerosol Retrieval Algorithm and Simplified and Robust Surface Reflectance Estimation (SREMARA)

Anthropic Settlements’ Impact on the Light-Absorbing Aerosol Concentrations and Heating Rate in the Arctic

Contact Info

Product

Resources

About