Lots of interest in SPC categories today. Specifically Moderate being level 4 of 5, and not being "in the middle". A couple comments. (1/17)

SPC originally only issued categorical outlooks: Slight, Moderate, and High. Moderate was the middle category, where many now say it belongs. (2/17)

Starting in the late 90s and continuing into the early 2000s, SPC worked with scientists at NSSL to determine what the appropriate probabilities should be to achieve the different categorical thresholds. (3/17)

Over the years these probabilities were refined and have been consistent going back to late 00s. SPC categorical outlooks are driven by the probabilities. We have lookup tables you can use to see the mappings: https://www.spc.noaa.gov/misc/SPC_probotlk_info.html (4/17)

So how do we arrive at a Moderate Risk. For simplicity, let's focus solely on tornado probabilities for the rest of this thread. (5/17)

The current tornado probabilistic thresholds are 2, 5, 10, 15, 30, 45, and 60. For the sake of simplicity, let us ignore the 10% significant tornado "hatched area". (6/17)

In the 3-tiered categorical outlook system, the probabilities were essentially 2% == SEE TEXT, 5-14% == Slight, 15-29% == Moderate; 30%+ == High. (7/17)

One problem (of many) with splitting the tornado probabilities up like this is that an overwhelming majority of the time we're in the Slight Risk tornado probabilities. (8/17)

Feedback from EMs was often along the lines of "is today a regular slight or an enhanced slight" because the character of a 5% tornado day versus a 10% tornado day is different. (The tornado threat had doubled.) (9/17)

To address this feedback SPC added Marginal (to explicitly denote SEE TEXT areas) & Enhanced (to denote the enhanced slight area) in 2014. (10/17)

No new forecast was actually made; we just changed the probabilstic-to-categorical mapping to better represent the underlying probability areas. (11/17)

Why does all this background information matter? Because if you remember that the probabilities are driving the categories, the middle tornado probability (15%) never changed and still maps to Moderate. (12/17)

Moderate (15%) still equates to the midpoint between No/Limited tornado threat (<2%) and High tornado threat (30%+) thresholds. (13/17)

To say it another way, the underpinnings of the categorical labels do not increase linearly. (14/17)

(For tornadoes) The jump from Slight to Enhanced is a doubling of the threat; Enhanced to Moderate is a 50% increase; Moderate to High is a doubling again. (15/17)

So the Moderate categorical label *is* the middle of the tornado-threat probabilistic thresholds, even though it is level 4. The problem is there are 4 labels to go from <2-14; and then only two to go from 15-30+. (16/17)

And, to be perfectly clear, YES, this is confusing. NO, I don't think everyone will understand this nuance. YES, I think we should revisit the categorical labels. YES, I believe I'm going to regret posting this thread. (17/17)