As I mentioned in my last blog, the study of aeroallergens depends highly on having data. In order to have data, we need to obtain samples of outdoor air, and analyze them to identify and quantify each type of pollen and fungal spore particle collected. We need to do this on a continuous basis, at set intervals, at multiple locations, for many years, to obtain a data set large enough to be able to study aeroallergen seasonal behaviour.
Unlike weather and air quality monitoring, there is no government-sponsored or government-operated monitoring network for aeroallergen levels anywhere in North America. In Canada, Aerobiology Research Labs was created to fill this void, with the primary operational goal of collecting and building a data warehouse of stable, consistent and comparable daily aeroallergen counts; a goal not unlike that of Professor Keeling’s Carbon Dioxide Record. Since 1994, Aerobiology Research has operated a monitoring network that has grown to 30 sites across Canada, with most sites having data going back more than 15 years.
The methods we use to obtain this data are as simple and efficient as possible. Each collection site operates a “rotation impaction” sampler, that operates continuously from early spring until mid October. The operator at each site changes the sample every day, at the same time, in the early morning. Samples are then shipped to our lab for analysis, which is done by hand (actually by eye) using optical microscopy and our lab’s highly trained and highly experienced staff.
The “rotation impaction” technology used to collect our samples is simple, robust, and effective. Throughout the day, the sampler spins a pair of thin, sticky sampling rods at 2400 revolutions per minute – effectively pushing them through the air at 72km/h. Particles in the air impact the leading edge of these rods and become embedded. The speed of rotations is constant, so the number of rotations controls the total amount of air sampled – as the rods spin in a circle, wind effects are negated since every direction is sampled at the same time, and any increase in airflow in one direction means a decrease in airflow in the opposite direction. For data comparison purposes, the number of rotations are kept as constant as possible by ensuring the samples are always changed at 24 hour intervals, at the same time each day.
Operating a sampler is a very simple task – anyone that can use of a pair of forceps can do it – and it takes only a few minutes to change a sample. However, the commitment of changing the samples at the same time every day from March to October, without exception, is both significant and appreciable – each collection season takes the dedication of many people to make this system work.
To convert these samples into data points, each sample is analyzed using the most efficient and comprehensive classifier ever created – the human brain. Sample rods are mounted under a microscope, and field-by-field each particle collected is identified and tallied, using various physical characteristics that are unique to each type of pollen or fungal spore. Once complete, these tallies are then converted to represent the expected number of particles found in a cubic meter of air during that day.
The individuals responsible for this analysis are highly trained – they spend months completing an intensive training course, including hundreds of practice counts, in order to be qualified to do this work. They are also engaged continuously in refresher training and quality control comparisons to ensure that the identifications being performed are consistent.
The cost of manual analysis in a centralized location is not insignificant, but it is the most efficient and cost-effective means available to date. Although it may seem obvious that computers could do a better or faster job, effective technology has not yet been developed. The problem lies with the diversity of the 90+ types of particles that are identified – the largest pollen grains can be upwards of 120μm in diameter, and the identifying artifacts on fungal spores can be smaller than 0.5μm. Custom technology is required for both imaging and processing, and because only a small handful of aeroallergen monitoring networks exist world-wide, the market simply isn’t there to support its development.
And therein lies the problem – because data collection has not been automated, and requires such highly specialized manual labour to be effective, it is expensive to do. This cost is a barrier to creating or expanding proper, comprehensive and standards-controlled monitoring. However, without having such monitoring in place for a number of years, this consistent and reliable data won’t exist, and so the business case to support monitoring often fails. In Europe, governments and health industries have pooled resources to support large aeroallergen monitoring projects, but in North America this is not the case. To my knowledge, Mexico has no aeroallergen monitoring at all. In the United States there is a sparse group of allergists, universities and research institutions that independently collect and analyze their own local samples and submit their records voluntarily to a national database, but the network lacks the controls needed to ensure data consistency. Neither country has sources of funding readily available to support an operation on the scale or quality of what exists in Canada.
There is no question that the data collected by aeroallergen monitoring is important. Unfortunately, importance is not enough to sustain this type of operation on its own – not in this economy. It takes dedication, commitment, and creativity to ensure aeroallergen monitoring continues in Canada for many years to come. Every season that the data set grows bigger makes knowledge available that could not have been known before. Without the data, you just can’t know – the aeroallergen equivalent to the Keeling Curve may be just around the corner.
 “50 years on: The Keeling Curve legacy” http://news.bbc.co.uk/2/hi/science/nature/7120770.stm