Abstract. In this study, we explore a new approach based on machine learning (ML) for deriving aerosol extinction coefficient profiles, single-scattering albedo and asymmetry parameter at 360 nm from a single multi-axis differential optical absorption spectroscopy (MAX-DOAS) sky scan. Our method relies on a multi-output sequence-to-sequence model combining convolutional neural networks (CNNs) for feature extraction and long short-term memory networks (LSTMs) for profile prediction. The model was trained and evaluated using data simulated by Vector Linearized Discrete Ordinate Radiative Transfer (VLIDORT) v2.7, which contains 1 459 200 unique mappings. From the simulations, 75 % were randomly selected for training and the remaining 25 % for validation. The overall error of estimated aerosol properties (1) for total aerosol optical depth (AOD) is -1.4±10.1 %, (2) for the single-scattering albedo is 0.1±3.6 %, and (3) for the asymmetry factor is -0.1±2.1 %. The resulting model is capable of retrieving aerosol extinction coefficient profiles with degrading accuracy as a function of height. The uncertainty due to the randomness in ML training is also discussed.

Abstract. In this study, we explore a new approach based on machine learning (ML) for deriving aerosol extinction coefficient profiles, single scattering albedo and asymmetry parameter at 360 nm from a single MAX-DOAS sky scan. Our method relies on a multi-output sequence-to-sequence model combining Convolutional Neural Networks (CNN) for feature extraction and Long Short-Term Memory networks (LSTM) for profile prediction. The model was trained and evaluated using data simulated by VLIDORT v2.7, which contains 1 459 200 unique mappings. 75 % randomly selected simulations were used for training and the remaining 25 % for validation. The overall error of estimated aerosol properties for (1) total AOD is −1.4 ± 10.1 %, (2) for single scattering albedo is 0.1 ± 3.6 %; and (3) asymmetry factor is −0.1 ± 2.1 %. The resulting model is capable of retrieving aerosol extinction coefficient profiles with degrading accuracy as a function of height. The uncertainty due to the randomness in ML training is also discussed.