It is believed that the TOR product can be improved by simply conveying more about what is known (or not known) regarding any specific tornado situation according to the current level of science and the present state of technology. This is especially true for Florida due to the diversity in tornado environments (cool season / high shear, warm season / low shear, tropical cyclone) and the common behavior of tornadoes within these environments. The initial issuance of the TOR product remains the first and best chance to invoke the desired public response. Therefore, the aim is directed at increasing the value of the initial warning through a situational approach. Although a thorough threat assessment is based on many elements, it is the "degree of confidence of tornado existence" along with the "known character of tornadoes within the particular environment" that carry the most weight in the warning decision process. However, this information is rarely reflected within the actual warning text which causes most TOR products to become generically the same. For instance, a TOR issued for a landfalling waterspout may not look much different than a TOR issued for a strong/violent tornado. It is the ability to properly convey this difference, without compromising safety, where public service can be greatly improved. This is accomplished by consistently using a library of terms and phrases within the warning product according to prescribed ECFL Tornado Warning Situations.
Warning generation software (WarnGen - developed by the Forecast Systems Laboratory), has been installed at all NWS offices as part of the Advanced Weather Interactive Processing System (AWIPS). The tornado warning situation scheme described in this paper has been locally configured into the AWIPS/Warngen software at NWS Melbourne to ensure the consistency in terminology through automation. A limited version of this concept has been tested in east Oklahoma by the NWS Tulsa office yielding positive feedback. Uniquely, situation specific tornado call-to-action statements were also included. Many came from direct findings of the disaster surveys from the outbreaks of 22-23 February 1998 in central Florida and 3 May 1999 in Oklahoma. Situational terms, phrases, and calls-to-action are presented in the paper, along with a demonstration of the modified warning software.
![](images/flag.png)
![](images/blue_ribbon.png)