Bubbler86

Definitely a better/broader push of cold air in early Jan on the 18Z GFS.

He would get kicked out of half his Broadway shows for using his phone to moderate during the Aria.

HH GFS is a'rollin. Lowered Christmas day highs by a degree or 2. Still a shorts day IMO.

Definitely an improvement over 6Z though it is more pushing out a bit....highs into the 40's or near 50 the first 3 days of Jan.

4th consecutive day of highs in the 30's and lows in the 20's and teens here. One of the more wintery temp weeks in 2 years. Not the most but top 5-10.

1128 Nooners at Liberty ski resort's west facing slopes. The snow is indeed sparse. 40 degrees and the front parking lot is about 25 percent full with golfers.

I was starting to buy in yesterday when the GFS Op was veering toward a very cold Eastern half of the US but disappointed to see the last few runs bleed it away. All of this is still 300-hour stuff, but it is what the output is. It does bother me that the GFS is using the new Fv3 core, but the GEFS is not.

You are digging in deeper than I did. Ha. I have thought about the possibility of having a limit on extreme outliers but then it brings into question 1) What if the outlier is right this time? 2) Where do we draw the line on what Outlier is? 20 degrees different than the next closest member? I think a good example of skewed distribution is the snow fall member maps I posted yesterday. I think most people understand this, but it is very clear example of the issue in using ensemble snow means to predict much...the snow means that matches these members was 3-4" through much of the LSV but only 6-7 of the 30 members below show that result. The rest show no snow or more in the range of 1-2" (18 in a quick count for 1-2") but the gigantic outputs of P26, P23, P03, and P14 not only cancelled out the ones showing nothing, they double what the majority of 18 show.

Agreed, if a lot of the runs agree then you have a higher chance of reality but that requires looking at the individual members to see if a small subset is skewing the totals vs. just the means. Let's take early January for example. Pretty much all the means AND members are suggesting it is going to be colder than the week before so a pretty safe forecast you would think but when expanded to look for potential low-pressure systems, I personally think it is the same or worse than using a single Op to do the same. Since the ensembles are purposely fed different data in an effort to find an average, I cannot buy using mean low-pressure locations or really the snow fall means as being any better than the higher resolution Op run. Measuring the uncertainty of the Op run is different than trying to argue that the lower resolution ensembles are better at forecasting that the Op. It also does not help that we now know the GEFS is using a different "software base" than the Op GFS. Different does not mean better or worse but is not ideal when trying to use them to find uncertainties in the Op.

I still think much of the issue is using ensembles to try and forecast ground truth. If you have a 10-model ensemble and 9 runs show the highs will be in the 50's and one run shows the highs will be near 0, the resulting ensemble mean is going to be grossly understated despite only 10% of the suites showing such a solution. It is similar to the difference between an average salary and a median salary. With average salary figures, the large wage earners skew the numbers to make the resulting figures look between than reality. Using the ensembles to "check up on the op's" is a better use, IMO....or to look for potential.

That was why I said next BUT it has trended backwards several runs in a row. Cannot blame anyone for wanting to see the goods within 5-7 days before we make a purchase. Too far out to fret over much.

It could certainly snow if a well-timed low came calling, which it does not on this run, but the lowest high temp on the entire run is around 40....which is 10-15 higher than it was yesterday on some colder runs. The Voyager maps are showing up again in the LR. NOT A FORECAST (not directed at you) and I do not need to be told how wrong it may be, just a PBP.

Better role out some more ensembles because the Op GFS is not really getting cold enough air down into the US now. Next (model run)!

What is also impressive (segue) is what Whitetails power bill must be despite not hosting paying skiers. They have had their main slope lights on every night for weeks.

Got down to 25 here but up to 31 here as well. The real cold air held a bit east of what the rgem as forecasting. Low to mid 20's in Chester county right now.

Hopefully the start of a new run. Two OT wins on the PP in Consecutive nights. A rare feat.

I know at Whitetail a lot of trails face South because I can see them from my house on a clear day. I feel like the Liberty ones face more West or Northwest but trying to gauge though off driving up there.

To me looking at this pic, Liberty is not even close especially with rain and 50's coming.

That is pretty amazing. The low highs this week must have really helped them out. Do not see the same for Liberty at least on their web site.

A polite way to say it, nice.

Yea, wayyy too far out. I agree that early January is our first window and seeing the Op showing so much cold air in the US at that point is reassuring. Unless this falls off, it is a big plus that we are only going to have a one week can kick this winter as to potential winter weather.

It actually seemed to transfer vs head....but that is as far as I am willing to dissect 384 :-). Would have to see some more panels to see if it a model blip but regardless, we are well on our way to losing 850's on the final panel behind a SE Wind at 850 and Easterly at the surface. I just cannot come to using ensemble means for anything except patterns so just discuss the one op I saw :-)

That system on the GFS looked like it was going to be something big and then, it runs for the GL on the last panel...so that may be out next, great hope indeed!

I cannot believe Office Depot is still kickin.