Graphix Based Weather Warnings?

[Bill Tabor]

Hey as Earthwatch is now gone, are there any other sites that show a visual county display of warnings such as severe, flood, tornado? I know that SPC shows these if you click on their watch boxes, but otherwise if no watch in effect not sure how to find these. Perhaps NWS sites have there own as well? Or perhaps something better?

Thanks

 

[mikegeukes]

http://wwwa.accuweather.com/severe/index-s...dc=0&traveler=0
http://www.nws.noaa.gov/
http://www.weather.com/maps/maptype/severe...ndex_large.html
http://weather.cod.edu/mcnexrad/Southcentral.ww.gif

Mike

 

[Bill Tabor]

Thanks Mike. Certainly a good start. Still I'll miss the 'click down' functionality of Stormwatch - even though it wasn't always up to date. Some of these are pretty good though.

 

[Bill Tabor]

Actually that second one (NWS link) may be the best one as you can click down on it pretty well.

 

[Greg Stumpf]

http://wdssii.nssl.noaa.gov/geotiff/

 

[Damon Poole]

Check these out -

1. http://www.weather.gov/view/largemap.php - Updates every 5 minutes with national coverage, and links to individual state pages. This site was put up when IWIN was first proposed for termination.

2. http://www.weather.gov/largemap.php - Same as above, except it has links to individual local WFO websites.

and,

3. http://www.srh.noaa.gov/data - Warning Map for the Southern Region WFO's with links to individual WFO websites and current SR Warning Status.
Also updated every few minutes.


Damon Poole

 

[rdale]

GRLevelX software adds the lat/long boxes to the display (speaking of which, apparently many NWS offices have yet to jump onboard because they still look an awful lot like county outlines in many cases!)

 

[Bill Tabor]

Nice links guys. I'm not sure exactly what to think of the lat / long warnings. Sure it will make the SAME radios more accurate so they won't be going off if nothing directly affects that part of the county out of the area of concern. Maybe that is good for public. However myself I usually like to know if there is anything tornadic anywhere near. I have mine set up to warn me on all counties surrounding me so I have reasonable warning time. Otherwise people on the edge of a county may not get much warning. The other issue is I think it will be hard for people to relate lat/long coordinates with what they hear over the radio or tv without some indication of what part of their county it is. Some people don't even know what there county looks like or what towns or locations are in different parts of their county. Plus for fast moving systems small lat/long areas may quickly become too small to cover the threat. In other words, specific is good unless it cuts into our warning buffer lead time, reduces the clarity of the message to the public.

 

[rdale]

"Otherwise people on the edge of a county may not get much warning."

Why would the warning not be issued in advance? Not sure I get that one...

"The other issue is I think it will be hard for people to relate lat/long coordinates with what they hear over the radio or tv without some indication of what part of their county it is."

Nobody really uses those except weather-weenies, a large majority get the storm track from their local TV station and we draw the tracks ourselves so have no need for lat/long warnings either.

 

[Bill Tabor]

"Otherwise people on the edge of a county may not get much warning."

Why would the warning not be issued in advance? Not sure I get that one...

[/b]

I guess I am assuming that the lat/long warning would still be confined to county boundaries for people in county in question. If they live right on the county line (say western line) and the storm is approaching from the west then when the warning is issued it seems like they wouldn't have near as much time as people further east. Their lead time will be based on warning lead time, and how quickly that warning is disseminated to them.

So, I'm not saying the warning wouldn't be issued in advance - just saying people on the edge of the county (on closest side of approach) may not get as much warning as those in other parts.

True that currently would be a problem with the old county specific warning scenario as well.

I guess what I am also saying is, the smaller the warned area, the fewer people nearby and lying outside that area will be aware of the situation. Using slightly larger county areas even though not as specifically accurate might have been beneficial in waking some people up and getting them prepared in the event that the severe threat then turned in their direction (which will often be the case for thos just downstream from a current lat/long warning).

Also we may technically have the ability to specifically warn on a particular area with high accuracy and detail but there may be a point of diminishing returns for making those warned areas smaller for reasons I just mentioned above. I know this is all motivated by reducing the FAR rate for larger areas, but the inverse problem is at some point areal size may be small to the point of being ineffective at relating the overall threat level to the public at large. That said, I am not saying the current lat/long implementation is to that point, but this should be kept in mind and studied in the future.

 

[rdale]

Weather radios still (and will for quite a long time) utilize counties for the alert tones - not the lat/long. The lat/long boxes are made so companies can utilize that info for automated cellphone / pager alerts, etc. versus just taking the automatic track of the storm.

But again many NWS offices still zig-zag their warning boxes to match the county lines, so it's not much help.

 

[Greg Stumpf]

But again many NWS offices still zig-zag their warning boxes to match the county lines, so it's not much help.[/b]

Actually, mostly that's a consequence of a feature in WARNGEN that allows a warning forecaster to forgo the polygon, and instead revert to the old method and click a county on or off. The resulting polygon is automatically created by the AWIPS system. Of course, those automated polygons don't match the county boundaries very well and could stand to use some major improvement (IMO).

 

[rdale]

Why would a forecaster use the old method? Does it make verification scores better? Otherwise seems to be a step in the wrong direction...

 

[Joe Nield]

GRLevelX software adds the lat/long boxes to the display (speaking of which, apparently many NWS offices have yet to jump onboard because they still look an awful lot like county outlines in many cases!)
[/b]

This is likely due to the fact that the polygon warning philosophy is still in testing. However, it will likely be national policy by convective season 2007.

 

[Greg Stumpf]

Why would a forecaster use the old method? Does it make verification scores better? Otherwise seems to be a step in the wrong direction...[/b]

WFOs are "graded" on their warning performance. If warnings are verified by county (polygon), then the warning forecaster will issue county- (polygon-) based warnings to maximize their scores. This is much easier to describe using figures, but think about it and it makes sense. Consider oddly-shaped counties, or postions of counties covered by one polygon.

This is likely due to the fact that the polygon warning philosophy is still in testing. However, it will likely be national policy by convective season 2007.[/b]

Actually, polygon warning capability has been around since the early days of AWIPS, with some WFOs using them as early as 1996ish. But as I said - because it is still policy to verify warnings based on counties, we see a lot of warning polygons "re-shaped" to county outlines. This will hopefully end soon as warnings become verified by polygon.