Juniper Pollen Counts

I have been gathering my own juniper pollen counts in the spring of each year since 1995. Having suffered from juniper allergies myself for some time, I wanted to get a feeling for how the presence of the pollen around my home correlated with weather and time of the year over a period of time.

Some of my colleagues at work suggested that posting the information I had gathered on the Web would be of benefit to them as well, so this is the result of their requests.

How well this data correlates to counts published in the newspapers is debatable. However, the absolute numbers seem to me to correlate well. Newspaper accounts will SOMETIMES indicate that what they report is pollen grains per cubic meter. I use a microscope to count juniper pollen grains deposited on one side of "double-sided stick cellophane tape" on a microscope slide over one square centimeter over 24 hours. I have seen claims that these two counts correlate well, but I've never seen anything to substantiate the claims except that the numbers do seem to be at least in the same order of magnitude. The slide is kept near my house in White Rock in a place out of the way from direct wind for 24 hours.

Depending on the magnitude of the count, I divide the tape in to 15 random regions, each either 1/4 square millimeter, 1 square millimeter, or 4 square millimeters. I count the number of juniper pollen grains in each of the 15 regions, and find the average and the standard deviation. I multiply each of those numbers by either 400, 100, or 25 to get the average and standard deviation for an equivalent one square centimeter.

On windy days, in particular, the standard deviations are higher. I included them to get some idea of the variation in regions on the slide at different times. In general, the standard deviation is one quarter to one third the total count.

For most juniper allergy sufferers, a count of around 100 produces noticeable symptoms. I log counts from 0 to 20,000 over the season. Since counting is a lot of effort, I tend not to start until the juniper in my local neighborhood appear to be about to release pollen AND until I begin to notice symptoms. I continue counting until the daily count begins to decline. When it has declined far enough that my energy for counting runs out, I stop.

This means I typically begin around Feb. 20, and continue until early to mid April.

In order to display this data in some form from which overall patterns can be seen and yet without much effort to display, I use xgraph under LINUX to read and display the data. Xgraph reads ascii text files for the data. I include this year's data and the geometric mean of the counts since 1995 in the ascii text file for xgraph and an xgraph display, both of which are available on these web pages. The graph is logarithmic on the count variable, since the counts cover such an enormous range. The counts are shown on the graph from 1 to 20000 and a count of 100 (the threshold for most allergy sufferers) ends up in the middle of the graph. Counts of 1 represent days when the count is 1 or zero and can also occur when I'm not counting because I'm out of town or for some other reason.

The Day Numbers are plotted along the x-axis with day zero being Jan. 31. Hence, the day numbers coincide with February.

Therefore (in non-leap years):

	Day 1  is Feb. 1
	Day 20 is Feb. 20
	Day 40 is Mar. 12
	Day 60 is Apr. 1
	Day 80 is Apr. 21
Plotting the days by month rather than day number on the x-axis was more trouble than it was worth to me. The idea was to make this exercise as easy possible to do, so I'd have enough energy to keep it up throughout a season.

Revised April 20, 2000