I would like to thank everyone for providing feedback on the topic of thunder as it remains challenging, as it did before the NBM transition. One benefit that has come out of this transition is getting some idea of what option(s) were in use out in the field. Additionally we are getting some good feedback on the NBM thunderstorm probabilities. To summarize the feedback and issues thus far (includes those pre-NBM transition):
1. The 12 hour NBM thunderstorm probabilities are not useful because of the 6 hour PoP population.
2. There are National SmartInit issues with the NBM thunderstorm probabilities, as noted in NIC ticket 64921. NBM thunderstorm probabilities actually come in every 6 hours, representing a 12 hour period (note MOSGuide is similar to this). This could partially solve item #1 above.
3. Feedback goes both ways in terms of utilizing the previous SREF/MOSGuide blend of thunderstorm probabilities and the Experimental Combo method. Please note that the reason the SREF/MOSGuide blend was not put back into ForecastBuilder is that it is identical to the NBM.
4. The SREF, MOSGuide and NBM thunderstorm probability guidance is all calibrated to cloud to ground lightning, 0.01" of QPF, and the temporal time period of the guidance. These combine to produce low values for these data sets. The Experimental Combo method tries to rectify that and put probabilities more typical of what we are use to, while injecting some science research.
Further discussion of thunder issue will occur among regional and national teams based on the great feedback received thus far.
