What can we learn from the spot database?

2 posts / 0 new
Last post
KF6KYI
Offline
Last seen: 8 years 11 months ago
Joined: 2008/08/31 - 23:55
What can we learn from the spot database?
AttachmentSize
Image icon spots.png5.94 KB

I've been beaconing for about a month, and decided to take the opportunity to see just how many spots and where they occurred. I discovered that I'd been spotted about 4285 times by 96 uniq stations. I plotted the time in UTC against the number of spots, and ended up with the curve shown in the attatched plot.

I'm 7 hours off from UTC, so that indicates that the peak probably occurs right around 7:00pm local time, and then drops off. Of course, since I am on the west coast of the United States, the peak may not be indicative of actual propagation, but could be because there are relatively few stations to the west (you have to go out to hawaii, japan, australia and new zealand to get spots). So, the curve we see isn't reflective of propagation, but propagation merged with the density of spotting stations on the ground.

So, I guess I'm interested in just what kind of analysis like this has been performed. In theory, we run these MEPT transmitters to learn something about propagation, but we don't see a lot of systematic attempts to learn something about propagation using this data. WSPRnet.org has 1.5M spots in it, spread out over months. It seems like we could do a lot, like:

  • Figure out the geographic distribution of stations (are there really no south american, african and eastern european stations?)
  • I haven't had any European spots from my location. Is that me (are there nearby hams who get reliable european spots) or is it actually more difficult than other paths which might be longer.
  • Can we normalize for the distribution of stations?

And probably a lot more questions...

It would be awesome if we could get more direct access to the spots data. It's currently showing about 8200 updates in the last 24 hours, which means about 250k records in a month, which I'd estimate as being roughly 20Mb for a month's worth (if my rough estimate is right). What are the odds that we might be able to get a .CSV dump of a months worth of data?

Just some bored musings on a Saturday...