Analyzing the Presence and Concentration of Airborne Microplastics Using Python and Photo Analysis Techniques

There is little scientific knowledge surrounding the study of airborne microplastics, the effect they have on human health, or even how many are present in our environment. This study will provide data on airborne microplastics through the use of a device that was designed to collect and identify microplastics in the air in real time. This portable system is simpler than current methods, which involve complicated and time-consuming chemical processes that must be performed in the lab after obtaining long-term air samples. A Raspberry Pi controlled a small camera with a high-powered microscope attachment. A hand-held, battery-powered device used small fans to drive air through a filter that could collect airborne particles that could include microplastics. This device was deployed numerous times on the sea ice in Utqiagvik, Alaska, three hundred and thirty miles north of the Arctic circle. This device obtained a number of images that showed the presence of airborne microplastics. Now, a machine learning algorithm will be developed using TensorFlow to process the numerical data found with ImageJ photo analysis software to quantify the amount of airborne microplastics found above the Arctic sea ice, as well as in other test locations in southwestern Virginia. This code, in combination with the handheld microscope camera, will be capable of collecting numerical data on the prevalence of airborne microplastics in the field in real time.