Anybody know who to contact at NWS regarding discrepancies in the historical file?

[Josh Lietz]

Hi,

While undertaking the job of porting NWS historical tornado data into the database at Tornado History Project, I became aware of possible errors in the data as well as some blatant discrepancies between the data and other published sources.

For example, the Hallam, NE tornado of May 22, 2004 is noted as having a maximum path width of 2 1/2 miles in the damage survey here Omaha/Valley NE. However, the NWS historical tornado file lists the width at merely 4400 feet (as shown here.)

In the interest of historical accuracy, does anyone know who to contact at NWS regarding this and other discrepancies? I had some brief contact with Dan McCarthy in December (not sure of his position at NWS, but I found the historical data file on his webpage, so he is the one I contacted...) regarding some suspected errors I had found, but it never really went anywhere. His last response was this:

...I cannot make ANY changes that would differ from Storm Data. I cannot make any changes prior to 1996. This, segno would have to change from Storm Data. This database attaches all the segments together based on data provided from Storm Data through NWS Verification. So, my hands are tied....

I then asked him (since his hands were tied) who I should be talking to instead and he never got back to me. So, does anyone have any suggestions of who to contact? Or will I, no matter who I talk to, likely end up spinning my wheels?

Thanks in advance for your assistance and suggestions.

 

[rdale]

Storm Data comes from NCDC -- http://www.ncdc.noaa.gov

 

[Rick Smith]

Actually, Storm Data comes from the local NWS forecast offices. NCDC publishes and maintains the database. You should contact the Warning Coordination Meteorologist at the local NWS office.

Rick

 

[Josh Lietz]

Actually, Storm Data comes from the local NWS forecast offices. NCDC publishes and maintains the database. You should contact the Warning Coordination Meteorologist at the local NWS office.
Rick

Thanks. Do you mean to contact the Warning Coordination Meteorologist at my local NWS office, or the local office resonsible for the area in which the tornado occured?

 

[Rick Smith]

You would probably want to contact the WCM at the office responsible for the tornado(es) in question.

Rick

 

[rdale]

Thanks for the clarification - I always was under the impression NCDC was the starting point for initiating and then they took it to the local office.

 

[Mike McDonald]

The database has also been converted to shapefile format for use in a GIS. The data are available from NWS Indianapolis at
http://www.crh.noaa.gov/ind/?n=svrgis

Information there showed the data file from NCDC to be called "ONETOR." A Google search on "ncdc onetor" resulted in a Powerpoint from the SPC which has a flowchart showing the "SPC Database Derivation" as:

WFO StormDat Info -> NWS OCWWS/Verification -> SPC -> NCDC -> FTORNADO -> ONETOR

Hope this helps.

Mike

 

[Josh Lietz]

Thanks. Going off topic for a minute....That powerpoint was interesting, particularly the slide detailing the difference between the SPC and NCDC databases. If I'm reading it correctly, the SPC database is event based (meaning, in general, one record for each tornado) while the NCDC database is segment based (meaning several potential records for each tornado). The historical data file I'm using at the Tornado History Project is from SPC. I wonder if I could get my hands on the data from NCDC? It could be particularly helpful in the area of drawing long tornado paths, since each segment presumedly comes with a set of coordinates. Then, instead of one long straight line between the touchdown and lift-off points, a more accurate path could be drawn based on segments....

Back on topic...As a test run I have contacted the WCM for the Hallam, NE area (Brian Smith). If things go smoothly (i.e. If he responds to my email, can do something about it, etc...), I'll begin the process of trying to get other errors corrected through the appropriate regional offices (I have documented many other examples of suspected errors in the historical data ...)