UAV-based digital inline holography for real-time, autonomous aerosol diagnostics

Understanding the dispersion of aerosols from wildfire can improve our ability to model air quality and radiative forcing for various environmental applications. Currently, due to measurement challenges, there is a lack of field data of how these aerosols disperse in the atmosphere, which depends strongly on their properties (e.g., concentration, morphology, and composition) that may vary temporally and spatially over a large domain. In this work, we develop an autonomous drone system integrated with lensless digital inline holography (DIH) to measure the properties of such aerosols generated from wildfires. Combining machine learning object detection with vision-based flow diagnostics, our drone can autonomously identify a smoke plume, fly toward it, and then trace along the motion of the smoke plume, while measuring local aerosols with the DIH sensor. This tool can enable detailed in situ characterization of wildfire smoke aerosols during emission and dispersion downwind in a manner that yields real-time data for air quality and climate science. Furthermore, the capabilities of this system may be extended to applications for other airborne PM diagnostics, such as dust, pollen transport, etc. This research is part of the GAIA (Grand-scale Atmospheric Imaging Apparatus) project.