改良型宇宙 Lidar エアロゾル同定アルゴリズムに基づく南シナ海におけるエアロゾルの時空間分布特性

This study addresses the challenges of low daytime signal-to-noise ratio (SNR) and insufficient aerosol layer detection in CALIPSO satellite Lidar data by proposing an improved multi-channel aerosol identification algorithm. The algorithm features an automatic optimal smoothing scale selection, substantially enhancing SNR for both daytime and nighttime observations, with the SNR increased by more than threefold for both day and night. By exploiting the 532 nm polarization channel's sensitivity to thin dust layer and the 1064 nm channel's responsiveness to thick absorbing aerosol layer, a multi-channel aerosol mask identification approach is developed. Compared with CALIPSO Level-2 products, the new algorithm identifies more comprehensive aerosol masks, with identification consistency exceeding 95% between day and night. Monthly-mean comparisons show the new algorithm's aerosol optical depth (AOD) closely matches MODIS Level-2 products (correlation coefficient R = 0.927, RMSE = 0.013), and significantly outperforms CALIPSO Level-2 data. Validation against AERONET observations at the Dongsha and Taiping sites further supports this result, showing correlation coefficients close to 0.9 and RMSE values below 0.05. Together, these findings strongly confirm the reliability of the improved retrieval algorithm. Furthermore, our results quantitatively confirm that the persistent underestimation in CALIPSO Level-2 is primarily attributable to its insufficient detection of dust and, especially, smoke aerosols. Analysis of the 2019-2023 three-dimensional distributions and transport of smoke and dust aerosols over South China Sea (SCS) and surrounding coastal-oceanic region, based on high-quality CALIPSO multi-channel remote sensing, reveals not only the expected seasonal cycles but also several novel features: highly variable winter-spring transport corridors, abrupt spatial and vertical discontinuities in anomalous years, and pronounced modulation of aerosol pathways by extreme events such as ENSO. These high-resolution spatiotemporal insights uncover complex aspects of aerosol transport, providing an advanced scientific basis for regional climate modeling, transboundary air quality management, and global environmental monitoring.