بازيابي ارتفاعي عمق نوري ريزگرد با استفاده از داده سنجنده SEVIRI: منطقه مطالعاتي قاره اروپا

Vertical Retrieva‎l of AOD using SEVIRI data, Case Study: European Continent

مهندسي نقشه برداري

Multi-layer Aerosol Optical Depth (AOD) estimation with sufficient spatial an‎d temporal resolution is crucial for effective aerosol monitoring, given the significant variations over time an‎d space. While ground-based observations provide detailed vertical profiles, satellite data are essential for addressing the spatial an‎d temporal gaps. This study utilizes profiles from the Cloud-Aerosol Lidar with Orthogonal Polarization (CALIOP) an‎d data from the Spinning Enhanced Visible an‎d Infrared Imager (SEVIRI) to estimate vertical AOD values at 1.5, 3, 5, an‎d 10 km layers. These estimations are achieved with spatial an‎d temporal resolutions of 3 km × 3 km an‎d 15 minutes, respectively, over Europe troposphere. We employed machine learning models—XGBoost (XGB) an‎d Ran‎dom Forest (RF)—trained on SEVIRI data from 2017 an‎d 2018 for the estimations. Validation using CALIOP AOD retrieva‎ls in 2019 confirmed the reliability of our findings, emphasizing the importance of wind speed (Ws) an‎d wind direction (Wd) in improving AOD estimation accuracy. A comparison between seasonal an‎d annual models revealed slight variations in accuracy, leading to the selec‎tion of annual models as the preferred approach for estimating SEVIRI multi-layer AOD values. Among the annual models, the XGB model demonstrated superior performance over the RF model at all four layers, yielding more reliable AOD estimations with R² values 0.99, 0.97, 0.98, an‎d 0.98 for the four layers from low to high altitude layers. Further validation using data from EARLINET stations across Europe in 2020 indicated that the XGB model still achieved better agreement with EARLINET AOD profiles, with R² values of 0.86, 0.80, 0.75, an‎d 0.59, an‎d RMSE values of 0.022, 0.012, 0.015, an‎d 0.005, respectively. We performed a qualitative validation of multi-layer AOD estimations by comparing spatial trends with CALIOP AOD retrieva‎ls for SEVIRI pixels on four dates in 2019, showing strong agreement across varying AOD levels. Additionally, the model successfully estimated AOD at 15-minute intervals for two real events—a Saharan dust plume an‎d the Mount Etna eruption—revealing consistent physical characteristics, including long-range transport in the upper layers an‎d a gradual increase in AOD from lower to higher tropospheric layers during volcanic events. The results demonstrate that the proposed method facilitates comprehensive monitoring of AOD behaviour throughout the four vertical layers of the troposphere, offering important insights into the dynamics of aerosol occurrence.

5