The WSPRnet Database

17 posts / 0 new
Last post
Last seen: 1 month 4 days ago
Joined: 2008/04/25 - 12:06
The WSPRnet Database

A number of people have recently asked questions about the WSPRnet database. Two recurring questions are:

1. How reliable are the data?
2. Is anybody doing serious propagation studies with it?

I posted an early response to question #1 a few days ago, but I'm not sure if it was widely read. For those interested, I now have considerably more information to share.

Anyone can download the full WSPRnet database from . Separate files are available for each month, starting in March 2008.

I've now written a very basic "scrubbing" program that reads the database files and does the following simple operations:

1. Correct the timestamp for spots posted at odd UTC minutes.

2. Remove spots for callsigns spotted only once or twice during a month. (These are nearly always the result of a false decode.)

3. When a callsign is spotted by several different reporting stations in a particular 2-minute segment, determine the correct band by "the majority rules" and then flag any wrong-band reports.

Results for March 2008 through June 2009 are illustrated in a graph posted at .
The graph shows the number of spots per month increasing from about 100,000 per month to nearly a million. The number of spotted callsigns and reporting stations per month increased steeply in 2008 and has been roughly constant at about 600 per month during 2009. The number of "bad spots" detected by the above criteria has been around one percent of all spots, or a bit less, since the beginning.

As for doing propagation studies: some interesting results can no doubt be obtained with what we already have, but a serious effort would greatly benefit from more systematic organization of frequencies in use, etc. Some have pointed out that the NCDXF/IARU International Beacon Project (see ) might be even more useful if its beacons were modified to transmit WSPR signals at least part of the time. Is anyone interested in looking into such an idea?

Comments are welcome!

-- 73, Joe, K1JT