Three persons died as a result of a mid-air collision over Wisconsin during a photo flight. It is a sad event, but a very interesting flight track:
http://flightaware.com/live/flight/N3735W
A brief online narrative is found at
http://aviation-safety.net/database/record.php?id=20060205-0
The aviation-safety.net site referenced above send out e-mail notifications of plane crashes on a world-wide basis. The format of that e-mail provides a bit more information on one page than does the web site.
Subscribe to the ASN Accident digest mailinglist: http://aviation-safety.net/digest/subscribeform.php
Are you sure that flight track is related to the accident? Looks like something that happened a while ago…
No, I can’t be sure. To the contrary, if the track ends back at KMKE, it wouldn’t have been the track related to the accident. Still, it sure is interesting! 
I tried to find the track of the second plane that was involved. However, that plane has never filed an IFR flight plan, so it has no history page.
The collison was reported to have occurred near Watertown, WI and the other aircraft landed at Juneau-Dodge County airport(KUNU). This is inland and Nothwest of Milwaukee. The track is probably from the Jan. 18th flight and appears to have circled in the Sheyboygan, WI area. Maybe they were not IFR for the accident flight.
Maybe they were not IFR for the accident flight.
Granted that the last flight track shown is not the crash track. However, an aircraft need not be IFR for there to be a track. All that is needed is to have “flight following” to generate a flight track – provided that there was some previous IFR flight, which initiates a flight tracking page.
Recognizing these asymmetries, the flightaware mapping team has begun creating what I call “synthetic flights.” An example is where we start spontaneously receiving positions on an aircraft for which no flight plan has been filed and/or no departure message has been received. The asymmetry is that we will have sets of tracks (flights) independent of flights in the database as well as, in some cases, multiple tracks that “look like” different flights that reside within one departure and arrival – for instance, the case where a flight is out of radar coverage for a long period of time.
We are now storing tracks in the database and have collected several million. We expect to allow searching and display of historical tracks in the fairly near future.
We are also close to being able to “replay” all of the data received since we started collecting it, representing almost an entire year, and hence will be able to make all the tracks available, even ones that precede our storing them in the database and to fill any gaps where the software wasn’t working properly, etc.
Tons of work has been going on behind the scenes that hasn’t quite percolated into the view of the flightaware userbase, but you will be seeing some amazing new stuff in the coming weeks and months. Promise.
Karl