Well, here are today's results, in graph form and raw numbers:

After working on this for some time a bit of statistical analysis really needs to be applied to my numbers. Admittedly, I was a bit over excited about what I had seen on Saturday, and like many people jumped to an improper conclusion about what I was seeing. My original hypothesis was that if the window swipe count was higher, then "bingo" the smoking gun, or in this case fission products from Japan! However, the more I learn about statistics the Swipe count from Saturday was simply a statistical anomaly. This is where it gets a bit complicated. The first day I used a dirty window, so I will remove that test from the bunch. This leaves me 11 days of testing, so just using the swipe test the average it is 12.3 CPM. Now a bit of statistics needs to be applied (the Poisson Method), so adding together all the counts (remember, 5 minute count average) is 677. If I take the Square Root, 26.0192236625 and divide by 5 I get the sigma, or in other words, I need to see 5.2 counts per minute average more to have anything that is statistically significant and indicates "yes" the plume is detectable with the equipment I have.

Obviously, that puts a damper on my Saturday "discovery". So now what am I testing for? The EPA says that what they have found is a very small amount, if I detect it then I will know without a doubt that there is a greater amount of radioactive contamination, since I am not using as high tech equipment as they are. They are using continuous filtration of air and very high dollar germanium detectors (I saw one on eBay once, $10,000!). Also they need to be cooled with liquid nitrogen. So even operating one is expensive. At least now I know what I am looking for. However, even if I do get the expected count, that doesn't necessarily mean that the radiation is coming from contamination from Japan.

Still Monitoring in Yukon Oklahoma, PM1703m Pocket Scintillator 7 microRoentgen per hour; PM1208 Geiger Counter Wristwatch .05 microSieverts per hour