About the WSPR database reliability

As I already mentionned, the WSPR DB is far too unreliable to attempt any serious propagation study or to trigger any propagation alert (ie for multi-hop E's on 10/6m)...

Here is an example among many others
http://picasaweb.google.fr/lh/photo/rbCRPSa3mDsbUBy3EPQGIw?authkey=Gv1sR...

As human errors will always occur anyway, I think that a better filtering is required: For example when a spotted station is obviously on another band, all the spots from this spotter should be ignored... (ie case of K1JT and others on the above example).
We really need the WSPR DB to become something serious... otherwise it is useless !

Patrick