The Probabilistic Outlook is the layer of the SPC Convective Outlook that gives the numeric chance of a severe tornado, hail, or wind report occurring within 25 miles of any point during the outlook period. It is the underlying, quantitative product from which the more familiar categorical labels are derived.
What the numbers mean
Each contour on a probabilistic map represents a specific probability that at least one severe report of a given hazard will occur within 25 miles of a point inside that contour. The 25-mile radius is roughly the footprint of a major metropolitan area, which is the practical unit SPC forecasts on.
On Day 1, tornado, hail, and wind each have their own probabilistic map because those hazards can be forecast separately at short range. On Day 2 and Day 3, the three hazards are combined into a single total-probability map because storm-mode confidence drops at longer lead times.
How forecasters read them
A tornado probability of 15 percent is meaningfully different from 30 percent even if both fall inside an Enhanced Risk. Chase targeting, emergency management preposition decisions, and NWS staffing all benefit from reading the probability layer rather than only the category.
The map also tells a spatial story. A tight, high-probability bullseye focuses concern on a smaller area. A broad, lower-probability region signals more diffuse risk over a wider footprint. Reading the shape of the contours helps set expectations for how coverage may look on radar later in the day.
Common misreadings
The probability is a chance of a report near a point, not the chance a storm will hit any given house. A 30 percent hail probability does not mean 30 percent of the area will see hail. It means that at any single point in the region, there is about a 30 percent chance of a severe hail report within roughly 25 miles of that point over the outlook period.
The probability also says nothing directly about intensity. A high wind probability could be dominated by many marginal-severe gust reports, or it could reflect a smaller number of very high-end events. Reading the probability alongside the significant severe hatching is what disambiguates the two.