# README for "lidar_aerosol_typing_manual_labelling_MLH_UGR.csv" ## Overview This dataset contains vertically resolved atmospheric aerosol properties derived from the MULHACEN multiwavelength lidar at the University of Granada (UGR) station, which belongs to EARLINET. The file represents the manual labelling and imputed version of the dataset, prepared for machine learning (ML) applications for the aerosol typing task. This dataset is part of a study currently available as a preprint in Atmospheric Chemistry and Physics Discussions (EGUsphere): https://doi.org/10.5194/egusphere-2025-269, that has been accepted for publication. The final DOI will be made available soon. ## File Name "lidar_aerosol_typing_manual_labelling_MULHACEN_UGR.csv" ## Content Description Each row in the dataset represents a unique lidar observation at a specific altitude and time. The columns include optical and physical aerosol parameters measured or derived from lidar signals at different wavelengths, as well as labels used for supervised machine learning tasks. Only the detected layers of aerosol are included in the dataset from each EARLINET profile, which are indicated by the "layer_num" column. The ending "_data" of the variables refers to the variable at high spatial resolution, i.e., height-resolved. The ending "_avg" refers to the variable averaged per aerosol layer but at high-spatial resolution. The ending “_error” refers to the error of the height-resolved data. ### Variables | Column Name | Description | |-------------------------|-------------| | altitudes | Altitude of the observation (m). | | b355_data | Backscatter coefficient at 355 nm (Mm-1sr-1). | | b355_error | Error in backscatter at 355 nm (Mm-1sr-1). | | b532_data | Backscatter coefficient at 532 nm (Mm-1sr-1). | | b532_error | Error in backscatter at 532 nm (Mm-1sr-1). | | b1064_data | Backscatter coefficient at 1064 nm (Mm-1sr-1). | | b1064_error | Error in backscatter at 1064 nm (Mm-1sr-1). | | e355_data | Extinction coefficient at 355 nm (Mm-1). | | e355_error | Error in extinction at 355 nm (Mm-1). | | e532_data | Extinction coefficient at 532 nm (Mm-1). | | e532_error | Error in extinction at 532 nm (Mm-1). | | ae_355_532_data | Ångström exponent between 355 and 532 nm. | | ci_355_532_data | Color index (backscatter ratio) between 355 and 532 nm. | | ci_532_1064_data | Color index between 532 and 1064 nm. | | cr_355_532_data | Color ratio between 355 and 532 nm. | | cr_355_532_error | Error in color ratio between 355 and 532 nm. | | cr_532_1064_data | Color ratio between 532 and 1064 nm. | | cr_532_1064_error | Error in color ratio between 532 and 1064 nm. | | lr_355_data | Lidar ratio at 355 nm (sr). | | lr_355_error | Error in lidar ratio at 355 nm (sr). | | lr_532_data | Lidar ratio at 532 nm (sr). | | lr_532_error | Error in lidar ratio at 532 nm (sr). | | ae_355_532_avg | Averaged Ångström exponent between 355 and 532 nm. | | ci_355_532_avg | Averaged Color index between 355 and 532 nm. | | ci_532_1064_avg | Averaged Color index between 532 and 1064 nm. | | cr_355_532_avg | Averaged Color ratio between 355 and 532 nm. | | cr_532_1064_avg | Averaged Color ratio between 532 and 1064 nm. | | lr_355_avg | Averaged Lidar ratio at 355 nm (sr). | | lr_532_avg | Averaged Lidar ratio at 532 nm (sr). | | dep | Particle linear depolarization (%). | | dep_error | Error in depolarization ratio (%). | | layer_num | Numerical identifier of the aerosol layer. | | date | Date of the observation (YYYYMMDD). | | dep_avg | Averaged particle depolarization (%). | | label | Final label of aerosol type manually assigned to the observation. | ## Data Processing - Missing values have been imputed using daily median values for each variable. - Labels were generated based on expert classification. ## Citation If you use this dataset, please cite the associated publication or acknowledge the data authors appropriately. ## Contact For further information, contact: Ana del Águila (anadelaguila@ugr.es) Andalusian Institute for Earth System Research Applied Physics Department University of Granada