[Xastir-Dev] Re: [Xastir] May 4th Tornado Outbreak

[Gerry Creager N5JXS]

A good WFO will do exactly this: If the duty forecaster feels the need 
is warranted, he's going to issue the warning.  If the need for a 
watchbox diminishes, he can retract it.  Good practice: They try not to 
cry 'wolf' too often, but keep the calls realistic.  The 'kill/cancel' 
alerts are good data to keep around.

If you follow our Forecast Discussions, you find that the logic 
associated with forecasts, watches, and changes are discussed in 
readable form.  Not all NWS regions do it, but ours (Lance's and mine, 
Southern Region) encourages this strongly.

gerry

J. Lance Cotton wrote:
> On Monday 05 May 2003 17:50, Curt Mills, WE7U wrote:
> 
>>On Mon, 5 May 2003, J. Lance Cotton wrote:
>>
>>>Not sure at the moment. I'm in a GMT -5 timezone. Would that cause a late
>>>or early expiration? I wouldn't notice it anyway, I think, since wxserver
>>>sends out the kill packet when an alert expires anyway, right?
>>
>>No.  We currently ignore kill or cancel packets.
>>
>>The alert packets can contain a start date/time and an expiration
>>date/time.  We use the expiration timestamp to delete it.  If the
>>code to do that is wrong, that could explain a few things.
> 
> 
> Gotcha. In my area, the NWS frequently will issue a severe thunderstorm watch 
> or warning and then cancel it an hour later because it (the weather) didn't 
> behave as expected. I guess eventually we should handle kill and cancel 
> packets.
> 

-- 
Gerry Creager -- gerry.creager at tamu.edu
Network Engineering -- AATLT, Texas A&M University	
Cell: 979.229.5301 Office: 979.458.4020 FAX: 979.847.8578
Page: 979.228.0173
Office: 903A Eller Bldg, TAMU, College Station, TX 77843