New edit of the April 30 Westmoreland, KS EF-3 tornado - tried to bring out the background sky a bit more. From my view 5 miles to the southeast, it had a very different look/feel compared to the violent near-ground motion in the up-close videos I've seen.

I both love and hate that I still have excitement for the Pirates at this point in a disaster of a season.

This is the worst deal in a Pirates trade deadline filled with bad deals. And this has been the worst deadline in a franchise filled with bad deadlines. There is no punishment in sports worse than being a Pirates fan.

RT @jjcoop36: In 3 batters, the NL All-Stars have given Paul Skenes more run support than he has received in 7 of his 20 starts this year.

RT @tornado_talk: Campbelltown, PA F3 #Tornado – July 14, 2004! A powerful F3 tornado carved a 7.5-mile path through Lebanon County, PA, st….

3 MLB games in the past month, 15 homers. 1 Oneil Cruz 458 foot moonshot. (And 0 Pirates wins.)

Some poor kid in the outfield facing a 120 mph line drive.

RT @DHSgov: The mainstream media is deliberately lying about the events leading up to the catastrophic flooding in Texas. The National We….

RT @PGHClothingCo: Aggregate score of the weekend series:. Pirates 30.Mets 4.

By definition you can have only one "tornado of the year". But I feel like the Northern Plains have had a dozen TOTY contenders in the last two weeks alone.

RT @AaronRigsbyOSC: Clear Lake SD again!!!

Bullseye 🎯

Lastly, I see pros/cons in replacing the NAM (& others) with an RRFS ensemble. Could prove valuable to focus on just a few key modeling systems, in addition to saving $/resources. But I also get the value in a diverse suite of guidance, including long running models like the NAM.

As an aside, please don't use simulated reflectivity to make a forecast. It's not worthless (I'll use it in forecast briefings as a quick overview), but digging in to the synoptic and mesoscale environment is so much more valuable than the cheap/easy look at sim reflectivity.

For severe convection, I'll use both but honestly don't love either the 3km NAM or the RRFS. RRFS just doesn't seem to handle convective initiation all that well, while the NAM seems to produce too much in the way of cold pools and upscale growth.

The NAM certainly has its biases (too cool/moist in the boundary layer to name one big one), but knowing those biases is half the battle. And sometimes those biases help make for more accurate boundary position forecasts, especially drylines and CAD.

I do think the NAM still outperforms it in shallow, cold air situations. Whether that's arctic cold fronts or cold pools and outflow boundaries. Similar to many other models, the RRFS seems to erode those too fast.

My subjective (and thus perhaps not super scientific) perception of using the RRFS for a couple years, in comparison to the NAM:. At longer ranges (48-84 hours) I feel like it handles the synoptic details better than the NAM, particularly surface low tracks and warm/cold fronts.

RT @okcthunder: FOR THE FIRST TIME IN OKC HISTORY, THE THUNDER ARE NBA CHAMPIONS 🏆.

Severe THUNDERstorm warning for OKC ⛈️ ⛈️⛈️.