Retrieving Nocturnal Aerosol Optical Depth Using Lunar Photometry: Algorithm Improvements and Evaluation

The diurnal distribution of aerosols exhibits significant asymmetry, and nighttime aerosol characteristic data are crucial for comprehensively and profoundly understanding the aerosol diurnal cycle and the impacts of aerosol–cloud interactions in polar regions. In this study, a universal nocturnal aerosol optical depth (AOD) inversion algorithm is developed for lunar photometers using the ROLO (Robotic Lunar Observatory) Implementation for Moon photometry Observation (RIMO) lunar irradiance model. By combining observation data from the CE318-T sun–sky–lunar photometer at the Beijing-CAMS site, parameters such as the nocturnal AOD and Ångström exponent (AE) are inverted, and the inversion accuracy of the algorithm is assessed against nighttime aerosol products from the Aerosol Robotic Network (AERONET). The piecewise normalization method of the RIMO model and the temperature correction module have been integrated into the algorithm to improve the accuracy of nocturnal AOD inversion. The AOD retrieved from all channels (440, 500, 675, 870, 1020, and 1640 nm) exhibits excellent agreement with the AERONET nocturnal AOD products, with correlation coefficients (Rs) exceeding 0.99 and root mean square errors (RMSEs) below 0.01. For these channels, more than 99% of AOD biases are within ±0.02, and biases in the 675 and 870 nm channels can dip to ~±0.01. Additionally, the comparative results of the AE indicate that the R and RMSE values are 0.981 and 0.071, respectively, with both metrics exhibiting agreeable accuracy. These results demonstrate the robustness of the proposed algorithm, and high-precision nocturnal AOD products are expected to play a crucial role in regional environmental governance and global climate models.