STORM2K

Some discussion on a few other threads indicates people feel SPC has overdone Risk areas this year. Three Highs, several Moderates and numerous Enhanced or Slights. While there have been bad storms that have caused damage and casualties, the feeling seems to be that few, if any of the higher Risk days lived up to the forecasts and that they could have stayed at lower Risk levels.

Begs the question though, is it better to over forecast and bust or to under forecast and find areas/communities unprepared?

It's not about the hype, it's about making the most accurate forecast possible and sending risk messages based on those forecasts.

Yesterday we witnessed a few big model failures due to what appears to be some bad data, plus an area where numerous radars where out of service so the warning intensity increased due to lack of accurate coverage (Georgia) (this is my opinion).

I'm not sure it's fair to evaluate this based on hype because how can one proceed with hype as criteria. The SPC is trying to make accurate predictions. My area was in a 40% criteria meaning there was a 60% chance nothing would happen.

If the local media want to hype it there's nothing that can be done, and certainly a forecast should not be withheld or reduced in impact based on how the media might behave.

The thing is it seems they based their forecast on faulty data, my opinion is they should have stayed at moderate risk and not gone high, but it's a pretty close call as there were some large tornadoes in the high risk area

Hype is not always on the fault of the forecaster entity (such as the CPC/SPC etc) they provide data and best analysis possible with that data. Hype is often the commercial/public end as interpreted, as well as social media and amateurs which viewership and readership is the purpose.

One of the biggest hypes during an outbreak like this is "20m people in the risk area" when reality is only a small fraction actually witness severe weather. On top of that a person could post "a town of "10,000 people is in the way of a tornado" when reality only a few dozen may actually see impacts from damage paths that are often only some yards across.

In short the problem lies in the interpretations rather than the source. Take opinions with caution, always heed the warnings as they partake in your area is how I take it.

In the past, we used to predict the weather by looking at the sky but with the constant involvement in IT and tech fields, now meteorologist use different tools, radar and satellites to forecast the weather changes. Sometime the weather predictions made by the meteorologist are incorrect but we should not ignore the fact that the weather predictions have now improved as compared to the past and most of the predictions turn out to be correct. These weather predictions help us to prepare for sever weather changes and it is also one of the reasons that every smart phone user use a weather app in their cell phones to stay updated about changing weather conditions.

In the past before the computer age, forcaster skill relied more on observed precursors telegraphed ahead of an event to determine the probability of it crossing a position. Now a steady flow of observations are feed into a pre-trained models whose skills is doubling every 2 years. I believe that 95% reliability is an industry standard for the 7 day forecast. The 14 day is better than the pre computer 3 day I'm sure.

That being said I do believe that climate modeling has lost an edge from the trained modeling trend in terms of sun spot activity, gravitation G and atmospheric tides, and irradiance has to be inversely related to distance. I don't believe the Milankovitch cycles are correctly being factored into the popular climate models