Related Experiment Video
Updated: Sep 20, 2025
07:07
Author Spotlight: A High-Resolution, Single-Grain, In Vivo Pollen Hydration Bioassay for Arabidopsis thaliana
Published on: June 30, 2023
2.8K
Field Evaluation of an Automated Pollen Sensor
Chenyang Jiang1, Wenhao Wang2, Linlin Du2
1Department of Biostatistics and Bioinformatics, Rollins School of Public Health, Emory University, Atlanta, GA 30322, USA.
Summary
A new automated pollen sensor shows strong correlation with manual counts during peak seasons, offering real-time data. While effective for tree pollen, further improvements are needed for weed and grass identification.
Area of Science:
- Environmental Science
- Aerobiology
- Allergy Research
Background:
- Seasonal pollen triggers allergic respiratory diseases.
- Current US pollen monitoring relies on labor-intensive manual counting with significant delays.
- Automated, real-time pollen monitoring offers a potential solution.
Purpose of the Study:
- To field-test a new automated, real-time pollen imaging sensor (APS-300).
- To compare sensor performance against traditional manual counting methods.
- To assess the sensor's internal consistency and spatial heterogeneity of pollen concentrations.
Main Methods:
- Collocated an APS-300 sensor with a Rotorod M40 sampler in Atlanta, GA (2020).
- Conducted internal consistency assessment with two collocated APS-300 sensors (2021).
- Investigated spatial and temporal heterogeneity of pollen concentrations.
Main Results:
- Strong correlation (r=0.85) between APS-300 and Rotorod M40 during peak pollen season.
- APS-300 showed slight underestimation of total pollen counts due to fewer identified tree taxa.
- High correlation (r=0.93-0.99) between two collocated APS-300 sensors; substantial spatial/temporal heterogeneity observed.
Conclusions:
- The APS-300 provides internally consistent, real-time pollen data strongly correlated with gold-standard methods during peak seasons.
- Automated sensors offer mobility and real-time data advantages over manual counting.
- Further algorithm refinement is necessary for accurate weed and grass pollen identification.
