Towards an Automatic Pollen Detection System in Ambient Air Using Scattering Functions in the Visible Domain

Abstract

Pollen grains strongly affect human health by inducing allergies. Although the monitoring of airborne pollens particles is of major importance, the current measurement methods are manually conducted and are expensive, limiting the number of monitoring stations. Thus, there is a need for relatively low-cost instruments that can work automatically. The possible detection of pollen in urban ambient air (Paris, France) has been reported using the LOAC optical aerosol counter. These measurements indicate that the pollen grains and their nature could be determined using their scattering properties. For this purpose, the scattering functions (intensity and linear polarization) of 21 different airborne pollens were established in the laboratory using a PROGRA2 instrument. The linear polarization curves were close together, with a maximum polarization lower than 10% in the red domain and 5% in the green domain. The variability from one sample to another was partly due to the different sizes of the grains. An instrument with an absolute accuracy of about ±1% for polarization measurements should then be needed, coupled with a counting instrument to take into account the effects of size. On the other hand, the scattering curves for intensity presented with different shapes and strong differences up to a factor of 20 at some scattering angles, due to the size, shape, surface texture, and composition of the grains. Thus, we propose a proof of concept for new automated sensors that can be used in dense networks to count and identify pollen grains by analyzing the light they scatter at some specific angles.

Spatial Variation of Airborne Pollen Concentrations Locally around Brussels City, Belgium, during a Field Campaign in 2022–2023, Using the Automatic Sensor Beenose

“…Using a 360 • inlet, the ambient air flow is pumped through an optical chamber at a flow of about 15 L•min −1 , and the particles individually cross a laser beam. The scattered light is collected by optical detectors at 4 different scattering angles, 15 • , 60 • , 125 • , 160 • , which allows to partly describe the scattering curves and to retrieve some physical properties of the particles [25]. The measurements at the first angle are almost insensitive to the refractive index of the particles when they are not perfect spheres, thus a direct relation between particle size and scattering intensity can be established during laboratory calibration [34].…”

“…A new automated and real-time pollen sensor named Beenose, developed by the Lify-Air company, could cope with these constraints, and could be deployed in dense networks. Beenose is based on light scattered by particles crossing a laser beam in the visible domain, which has been reported to help in distinguishing some pollen families from other ones in several studies [ 24 , 25 , 26 , 27 , 28 , 29 , 30 ]. The same principle has also been used to develop optical counters with typology determination capabilities, to distinguish between anthropogenic pollution particles, natural mineral particles, and pollens [ 31 , 32 ].…”

A Laboratory Evaluation of the New Automated Pollen Sensor Beenose: Pollen Discrimination Using Machine Learning Techniques

“…In our previous study [ 37 ], we showed that pollens exhibit specific intensity scattering curves allowing their detection among other types of particles and even to distinguish between their various families. In this paper we introduce Beenose, a new small and relatively low-cost pollen sensor that has been developed based on our previous results.…”