Open Collections

Description

This dataset contains transmission electron microscopy (TEM) images and associated MATLAB files used for image processing of smoke particles collected during forest fires in Kamloops, British Columbia, Canada, in July 2021. Samples were obtained at 1259 Dalhousie Dr, Kamloops, BC V2C 5Z5 (BC Ministry of Environment) using thermophoretic sampling during both daytime (22 Jul. 2021, starting 10:41 am, ~ 1.5 hours duration) and nighttime (21 Jul. 2021, starting 18:48 pm, ~ 9 hours duration) conditions. The TEM images reveal a range of particle types, including soot (both fresh and coated), tarballs, ash-like structures, organics, and newly observed “softball” particles. Processed MATLAB data are included to reproduce the automated image analysis presented at the American Association for Aerosol Research (AAAR) 2022 Conference. Image segmentation was performed using k-means clustering, followed by morphological characterization of particle circularity, optical depth, and boundary sharpness. These metrics support differentiation of particle types and highlight differences in morphology between day and night samples. The particles were also manually classified based on visual assessment of their types. This dataset may be useful for researchers working on biomass burning emissions, aerosol particle morphology, atmospheric aging processes, and automated image analysis methods.