A Spatio-Temporal Weighted Filling Method for Missing AOD Values

Aerosol Optical Depth (AOD) is a key parameter in defining the characteristics of atmospheric aerosols, evaluating pollution, and studying aerosol radiative climate effects. However, large amount AOD data obtained by satellite remote sensing are missing due to cloud cover other factors. To obtain with continuous distribution space, this study considers spatial temporal correlation proposes spatio-temporal weighted filling method based on sliding window supply blocks. The uses semivariogram autocorrelation function judge threshold as size, then it builds model for each fill values. We selected area full values simulation. results show that accuracy has been significantly improved compared mean method. R2 reaches 0.751, RMSE 0.021, effect smoother. Finally, was used MultiAngle Implementation Atmospheric Correction (MAIAC) Beijing–Tianjin–Hebei region 2019, AErosol RObotic NETwork (AERONET) true value testing. filled high AERONET AOD, 0.785, 0.120. A summary 13 cities shows first third quarters higher than those second fourth quarters, highest March August; among cities, Chengde Zhangjiakou lower cities.