DTP

And unfortunately this is somewhat of a zero-sum game given the wide range of geography in this subforum Hope everyone gets something....nice to see some flakes (kid in me), but tbh this kinda f's up my program development-wise bc of construction delays.....can't rough grade for survey pins, or prep for a slab pour/underslab plumbing (or pour the slab) with snow on the ground, or the subsequent mud after it melts.....plus pushes sub crews out with the backlog.....I would gladly donate whatever we get around CLT to some of the Raleigh/Southside VA peeps- it can snow all it wants to after I get the slabs down and 3 full afternoons above 70F into the cure lol

SC DNR has a really good write-up for a close analog to this type of setup that occurred on 1.2-1.3.2002 PDF file is too big to post, but worth the read if you do a quick Google search....it shows the entire upper air synoptic progression/summary in good detail throughout the storm

Lol- weird triple post thing- it didn't show the first one or the second go through

Deform band, if it stalls/trains, is where you get 2'+ totals @ 3-4"+ an hr Wouldn't believe it here if I hadn't seen it happen in the Piedmont- 04' storm in CLT- you could literally hear the faint ticking of the snow it was falling so hard....Mom got 24" by the airport in Rock Hill I was up on North Face at Mt Crested Butte when there was nearly 10" in one hour- craziest thing I ever saw- chair ride up is ~20 minutes to the top- chair was clear at the bottom- ~5" deep by the time I got off

There will be a sharp cutoff on the western edge from this type of storm- eastern side of these type of setups outperform

Best analog i can think off the top of my head is Jan 01'...we had thundersnow in the Upstate under the ULL...I lived in Rock Hill at the time, and we got 4-6"....temp profile here is 5-7 degrees colder on this storm if my memory serves correctly With the location of the trough to our south, it is a good to great setup for the E Upstate as long as you have a negative tilt....I am quite a bit more confident after seeing the models when metadata from the sampling last night was ingested I feel yout pain- it is like night and day living down here v the other side of 85- I lived on the N side of CLT/Harrisburg for 15 years before moving to TC- we would typically have at least 2 good storms a year up there

W edge has to tilt about 20-30 degrees eastward at the base for it to be optimal, all other other things being equal....the southward dig and overall location of the trough base is a good sign, however

We will have better luck on this storm....you are in a pretty good spot for this type of setup

Yessir- was just looking at the same thing....verbatim textbook setup for snow in E Upstate/CLT and pts N and E....I am more confident after the storm was sampled

Delay glitch on the graphics edit software- idk why graphical software providers haven't managed to fix this over the last 15 years....I can almost 100% guarantee that whoever created the graphic saw the corrected and spell-checked version before posting.....unfortunately the graphics program doesn't always register the corrected version that the end user sees

Definitely on the analytics end, but yes and no on the manual end....field construction is dangerous, dirty and atmospherically unpleasant in any season down here. Materials don't interface well with conventional manufacturing robotics outside of panelized housing systems and steel rollups....the 3d printed home is essentially the same as a block house, but both still have to wood-framed on the interior....drone imagery for grading large sites is probably the biggest jump i have seen- and the data can def be plugged into AI

The craziest thing I saw statistically with AI is the ability to sort valid predictive data from non- in massive data sets.... you can see the program actually latch on to the correct data pts while discarding the useless garbage.....and at ever increasing speeds The groupings went from looking like buckshot when the program first started learning to damn near completely linear (and incredibly tight relative to SD from the mean) within a couple weeks of monitoring...it was impressive and somewhat disconcerting at the same time

Yes- I believe this to be true- it just takes a bit of time....also why I am no longer not creating statistical models in finance for prime dealers....once AI begins learning on correct data/assumptions, there was no way I could compete moving forward- their models were too granular and incredibly robust....plus what would normally take me several weeks/months to create- AI can now do in a matter of.... hours I saw it early on, and fortunately I have no desire to ever work in finance moving forward...I prefer being on a construction site far more than being behind a laptop tbh

Yup- this is the point where I start paying attention lol.....

Rare, but it does happen if the air mass is cold enough and the wind direction is right

It can happen, and has, both off of the Atlantic and Pimlico Sound. It would take normal ocean temp for this time of year (40's to low 50's) for it to occur, as well as a strong NNE wind coming in over a long fetch of water (several dozen to hundreds of miles). Warmer ocean temps relative to land enhance the mechanism, but air temps have to be really cold (mid 20's and lower) to have optimal conditions...hard to put both of those together in the Carolinas. It is much more common on Cape Cod than NC Lake effect snow has happened downwind from Kerr Lake, believe it or not

That would produce some serious sound- and ocean-enhanced snow given surface temps north of Hatteras up to SE VA.... not something you see very often Fetch running straight down the Chesapeake N/S Some coastal flooding to the south with the onshore, and a mega foam special at my house in OKI

Ha- remember this storm well- still one of my all-time favorites..... lived off of Harris Blvd/Davis Lake in N CLT at the time....remember waking up to 8"+ of blowing snow and a 15 degree air temp- don't see that very often down in the S Piedmont

Np- hope it helps- just hate to see everyone get so darn disappointed....the Schrodinger's cat paradox comes to mind- except with modeled storms- is it alive, or is it dead? Can't be both at once- function v reality

This...wait for more metadata to be ingested by all of the globals and mesos....take a deep breath- pretty much the only takeaway from any model run to this pt is an elevated probability of snowfall Friday and Saturday in the Carolinas, as well as some serious cold. The models are going off the tails of probability on both sides at current without much in the way of ingested metadata- the wild swings will not stop until that occurs....think of the model data in terms of three separate normal distributions that are interconnected. The middle normal distribution is the ensemble mean. Robust predictive value will rest somewhere between 50d (the center; 50= 0 SD from mean), and ~37.5d of each wing of the ND. Outside of this envelope- 25d is the middle of the wing, 12.5d and lower the tail. Now plug in time to the equation- at 96-120 hrs and little meta (actually occurrence) data, you can draw two separate normal distributions with the 50d mark for the separate distributions at 12.5d of the ensemble mean. This is what you are looking at currently- a really wide range of generated outcomes that form the middle (model) normal distribution The normal distributions on the tails get pulled inward toward the center of the ensemble mean as more time passes, moving the center of the model data sets inward toward the 50d mark of the main distribution as more metadata is ingested, therefore making the model trends more robust the closer you are to an event. The ensemble mean will also shift one direction or the other throughout as well as more data is ingested- this creates the final envelope of robust data that can really be drilled down 25d is the most susceptible point to model volatility in normal distribution (any predictive model- not just weather) - also why the models tend to "lose" a storm 2 days out (temporarily)- as the 2 semi-predictive data set means (more like a 50d line of best fit amongst solutions) that started on the tails (2 separate normal distributions) cross the 25d (25 standard deviations from the mean) area, they amplify all of the solution sets away from the mean that are not metadata derived. Just the way the math works....best way I can think of to explain it- just keep this in mind when you are looking at any potential model outcomes 4-5 days (or more) in the future Like having a piece of cardboard and shooting 1' left, and 1' right with buckshot at hr 120

I made several posts about this in the disco thread for the last storm regarding all of the models (not just the Euro), and why they have predictive difficulty more than 3 days out....I have created multiple complex models in finance involving delta hedging equity and futures options for prime dealers, so I have a pretty good handle on what I am looking at....my posts go into the "whys" of what you are asking..... Short answer/cliffs if you don't feel like searching- they aren't designed to nail down specifics this far out, just probabilities....that being said, I don't want to be a buzzkill- it is still fun to speculate based on what you are seeing now on the models in concert with upper air features Probably the lone exception is hurricane modeling; it is different and much more granular bc of the large amount of metadata infused starting early on from the tropical wave stage

The half IP front end and half ZR back end depiction of the squall line on HRRR and NAM was pretty accurate given what I just saw....

Mostly ZR in Tega Cay now- had about 0.5" of sleet or more in 30 minutes before it changed over. Probably at or maybe a little over 1.25" of sleet on the ground Temp shot up 4 degrees in 20 minutes right before the changeover from 21 to 25 F Very noticeable ice accretion in the trees, but nothing too crazy yet

Heavy sleet in Tega Cay. 21F Little bit of ZR mixed in at onset, but now all sleet

Basically opens the door to a spiral of deviation- never ends well....