Abstract. The paper presents an approach to revealing the variability in aerosol type, at high spatiotemporal resolution, by combining fluorescence and Mie–Raman lidar observations. The multiwavelength Mie–Raman lidar system in operation at the ATOLL (ATmospheric Observation at liLLe) platform, Laboratoire d'Optique Atmosphérique, University of Lille, has included, since 2019, a wideband fluorescence channel allowing the derivation of the fluorescence backscattering coefficient βF. The fluorescence capacity GF, which is the ratio of βF to the aerosol backscattering coefficient, is an intensive particle property, strongly changing with aerosol type, thus providing a relevant basis for aerosol classification. In this first stage of research, only two intensive properties are used for classification, namely the particle depolarization ratio at 532 nm, δ532, and the fluorescence capacity, GF. These properties are considered because they can be derived at high spatiotemporal resolution and are quite specific to each aerosol type. In particular, in this study, we use a δ532–GF diagram to identify smoke, dust, pollen, and urban aerosol particles. We applied our new classification approach to lidar data obtained during the 2020–2021 period, which includes strong smoke, dust, and pollen episodes. The particle classification was performed with a height resolution of about 60 m and temporal resolution better than 8 min.