OceanStWx

Latest NAM and GFS are definitely teetering on blue bomb around here (big totals could be mitigated somewhat by crappy snow growth at times though). HRRR torchy. Euro somewhere in the middle. Looking at the low level cross sections it's mostly below freezing at 950 mb at GYX, would struggle at times for PWM. On its face though, the NAM and GFS are pretty classic crosshair signature for PWM. Hard to argue that 6+ isn't at least being discussed at the table if not being served.

What we wouldn't give for that mid level on July 4th at Scooter's lake house.

Checking the creeks running near bankfull on the way into work today all I could picture was Ginxy nodding along like Jeremiah Johnson.

In January. So it is winter adjacent.

This is him now.

I wish Josh did winter QPF, he's definitely more of a collaborator. But he's just the snow desk.

The biggest problem I see with it, is when the super ensemble spits out QPF that is unreasonable. Can we get WPC to make the requisite edits to their national forecast to fix what happens downstream in our local area. Like when the dry slot moves in and models habitually spit out too much QPF. Now in order to get the forecast right locally I have to do something unscientific with snow/ice ratios in certain areas to make sure I don't get wild storm totals. Or when QPF goes wild in WAA and I have give Dendrite 40:1 ratios to capture where I think deformation will actually drop a bunch of snow. We are going to produce snow character maps now, so I don't need people thinking they can leaf blow their driveway when it will be more standard ratio stuff but a machine spit out the super ensemble.

We're already seeing the QPF being taken away from us (which obviously makes getting the snow even harder when you don't control the primary input source). I think the biggest issue I have is that WPC's resources are small compared the size of the area they need to forecast for. The end result is leaning heavily on the super ensemble blend while sacrificing local details/knowledge. We've seen that even using an ensembled approach that the ensembles can make large swings. We cannot yo-yo that much in our public forecasts and have any kind of credibility. There isn't much that drives me more up the wall than "the model says this, so I'm adjusting like that". And that is largely what the super ensemble does, it's just a bit more robust.

At least Scooter waits to melt until late January.

It's kind of a high stakes game of chicken for the weenies. Who breaks down and melts and who trusts the pattern.

It's on the main page https://www.cpc.ncep.noaa.gov/ under 8-14 day hazards, right next to where you get the temp outlooks. I know this is a CPC product, but it feels like WPC especially is going to start moving towards a more SPC style on winter weather. We'll have risks and eventually I think some would like them to take over watches. I personally think there are some fundamental differences there when it comes to how quickly severe evolves vs. winter weather, but it's also well above my pay grade. This WSSI product is going to start getting a lot more press as well. This is generated from our local gridded forecast and adjusted by climatology locally to a degree. It incorporates a bunch of things beyond just temps and snowfall amounts. We've done a lot of work to make sure that a 12 inch storm around here doesn't spit out "prepare to be without critical services for weeks" and other statements that make more sense for ATL than here.

You'll be peeling yourself off the turf outside the coop.

I've seen it displayed various ways, but it all pretty much points to about two weeks delayed from astro winter, like we always say.

Morch on, brother!

Getting into position out on runway 29.

The good news is...

Models are "smart" enough these days to recognize snowpack, the problem is that they will lay it down where they think it should go and as you say the effects cascade downstream. So for instance if we think we get a correction vector colder, we probably don't want to assume the storm that follows is a good bed where the GFS says it is.

In some respect it's pointless to sweat the details of model runs this time of year. A good pattern will put snow down somewhere, and that will ultimately change the baroclinicity that affects future events.

That's kind of how I feel about the resolution upgrades to the models. Man what I would give for a nice smoothed 80 km fronotogensis grid sometimes.

Q2. But it's an ensemble upgrade. EPS going to 9 km and the extended is now going be a separate run rather than an extension of the medium range model.

Parallel universe? But we're actually upgrading to 16.3 today. Looks like mainly some snow depth changes and increasing data assimilation (which could be important).

Somebody was testing products on the system 9/14/2009 but forgot to actually put things into test mode. https://mesonet.agron.iastate.edu/wx/afos/p.php?pil=MWWOKX&e=200909141234

Last time Mount Tolland blew its top?