Putting things into perspective - SPC categorical downgrade

Some of the weather communities I inhabit are jeering this morning at the professionals of the National Weather Service, as well as television meteorologists, hobbyists and others who do their best to figure this stuff out. We've had bust after bust this year. The long range and short term models have predicted very significant events, only to have everything fizzle seemingly moments prior to zero hour due to some unforeseen caveat. Last week, a major capping inversion put the kabash on severe weather and even general thunderstorms in our area. Before that, we had a surprise where severe weather DID manifest to a degree, causing damage throughout the western and northern portions of the state. This was offset by the fact that the SPC issued no categorical severe threat for that day until the storms started to pop up. It would seem like the SPC is batting about zero for us this year.

But here's the thing - convective outlooks are what they are. They're a broad-brush estimation of where conditions for severe weather appear as though they're going to manifest based on the data received at the given time of the outlook. If the data received is lightyears away from what actually happens in the atmosphere, then the outlook is going to be inaccurate. What can look like an incoming event days in advance can turn out to be absolutely nothing. Last year, the models - especially the NAM - were very accurate in their forecasts and I grew to trust them implicitly. The June 22nd and August 2nd events, not to mention a few other minor ones, showed up days, even weeks in advance. Based on model data alone, you could get a good idea of the forecast. This is not the case this year, and it has served to humble me a bit. Nature doesn't conform to the rules mankind sets for her. Just because our computers THINK they understand what she's going to do doesn't mean that she's going to do it. Even when top minds - those with degrees in this stuff - include their observations of satellite imagery, complex mathematical equations, surface and atmospheric data collected by weather balloons, and decades experience forecasting weather, they don't always get it right. 

I've been making my own categorical outlooks for two reasons. The first reason is to practice forecasting where thunderstorms develop and try to compare my results to what the SPC puts out. The second is because I've had my disagreements with the SPC's outlooks and want to make them more clearly, visually known. With that in mind, I simply do not have the expertise to defy the SPC this time around. Rarely do I have enough confidence to claim that they're wrong. The exceptions were on June 22nd and August 2nd of last year. This year I've been deferring more to the SPC because the models have been so hokey. I've had two policies regarding "defiance": 1) I will not issue outlooks containing "widespread likely" or "high risk" unless the SPC does as well, and 2) if the SPC issues a higher risk than what I believe is the case, I will match that risk even if my confidence is lower in the event. I do this because I do not want people to let their guard down based on my observations - I'd rather be wrong and have thunderstorms not happen than be wrong and have someone go out on a boat based on my claims that the SPC is full of crap, then get killed because they failed to take a severe weather risk I was wrong about seriously. So when the SPC issues a slight risk for severe weather, or an enhanced risk for severe weather, I'm going to match it, even if it doesn't pan out.

Where is this going? The SPC has issued a very stark categorical downgrade once again, removing the slight risk for severe weather from all but the southernmost border counties (and only a fraction of those counties are actually included in the risk). However, when you put this into perspective, you'll see that it's relative to the evolution of the system and not necessarily a big miss by the SPC. The downgrade from slight to marginal comes with lower confidence. Short range high-res models currently do not indicate storms developing for all the hours they can see (going to around 9PM). What happens beyond those hours is currently not able to be forecast by these models, so more time is necessary to figure that out. Based on those models and other data, there is still a severe threat in the marginal risk. In fact, the risk for severe storms has actually increased in the northern counties and the Thumb Area from what it was before. This includes a 2% tornado risk for quite a few areas, especially those in the eastern portion of the state and Mid Michigan. 

This one is another of those complex situations where minor details can have a huge impact on what happens in the atmosphere. We've had nothing but those this year. The models have been unable to coherently grasp the dynamics of our weather days or weeks in advance. I'm looking forward to explanations by experts as to why the models are having such a hard time this year. For now, I still think there is a severe threat today. This is based on storm system placement and just gut-instinct. I'm anxiously awaiting more data from the short range models to support it.