Two days of rare September severe thunderstorms in Pennsylvania have dropped tornadoes and funnel clouds in the central and western parts of the state. I say "rare" because typically severe weather reports fall to close to zero by the middle of September (although there is sometimes a spike in wind reports in late September and early November). Here's a graph:

On Saturday, under a Marginal severe risk from the SPC, two severe thunderstorms moved across the western part of the state under tornado warnings. They were very impressive on radar, especially for normalized rotation (see map below). One tornado has been confirmed as an EF-1 in Warren County (Channel 12 has some video)

"A tornado touched down at 635 PM EDT around one half mile southeast of the intersection of Scandia and Peterson Roads in northeastern Warren County, PA. The tornado progressed to the east, crossing Peterson Rd, making damage to trees before moving across a farm field. The tornado picked up half of a large pole barn/garage and moved the debris 200 yards to the north into a stand of trees. Substantial tree damage was observed along a line of trees bordering the farm field to the east of the pole barn. Additional sporadic tree damage occurred mostly in a line for another half of a mile before an eyewitness saw two funnels lifting up over a neighbor's residence which lies on the north side of Peterson Rd roughly one and a half miles east of the intersection with Scandia Rd."

A second tornado (the icon you see on the map above in Butler County) was not confirmed by the NWS, even though the Associated Press reported it had touched down.

Sunday afternoon, the state was only in a General Risk -- not the day you'd expect to see supercell thunderstorms with hook echos on radar -- but that's exactly what we got. I chased a storm from State College, PA up 26, I-99 to Bellefonte, then north on 26 to 64 at Lamar. I can't say I've ever been in a better position with a hook echo moving over me... still amazed I didn't see any vertical rotation at my location. Proof that even when the storm presents well on reflectivity, there can be nothing there. The radar loops below show my position as the blue circle or red dot -- the first video is pre-chase, the second is during the case, the last is the second storm.

Here are the highlights from my chase video of the first and second storms (sorry for the shaky camera):

And a time-lapse of my dash cam:

Early in the first storm, before it passed the UNV airport and crossed I-99, Garrett Bastardi took these photos of a possible funnel cloud and wall cloud near the Bellefonte Airport:

J.P. Tracey shot this look at the second storm's wall cloud, which I was sitting in front of on I-99. When I was chasing that first storm, Mike Brulo was a few miles ahead of me (on the other side of the hook) and took this photo of a scud cloud:

400x266_09191958_14311258_1312967615382781_1890815701919293141_o

Later on, that storm severely damaged the town of Muncy, Pennsylvania, downing enough trees and power lines that school was closed today (Monday). The NWS in State College shared this Normalized Rotation track, showing where the storm moved. Notice the most rotation was shown early in the storm in Centre County when Garrett saw the funnel cloud, then later in the storm when it did wind damage in Muncy.