dallen7908

Looks like we may say goodbye to the 50s/60s beginning on the 6th with perhaps the highest 15-day snowfall mean from the EPS this season (3" with 2 HEC-like amount for immediate dC area) PNA negative throughout but some hope for a negative EPS too NAO positive for the next 10-14 days or so but could be negative late in the period Hopefully, this improving pattern has legs; however the 3-4 week outlook from CPC issued yesterday calls for above normal temperatures and precipitation.

I would be more discouraged if we wasted a -AO period during February. Sitting at 0.3" entering the New Year seems like the new normal inside the Beltway. Is it true that BWI is sitting at 0.1" or did they score a few tenths on another storm. Yes the extended range still looks to feature a ++AO, +NAO, neutral EPO, -PNA. Over days 6-10, 3 of 21 GEFS members and 3 of 51 EPS members give us 1" or more of snow. One EPS member gives us 7" during the January 2/3 time frame. Our 3 category chance of above normal temperatures is 40-50% for week 2 a far cry from the 70--80% hell we've seen during deep dark abysmal periods (DDAPS).

As last Friday's 3-4 week outlook said. "Intraseasonal variability in the tropics remains weak as there are still no strong MJO or other equatorial wave signals. The Indian Ocean Dipole remains strongly positive, although its amplitude has decreased by about 50% since the beginning of December. ENSO neutral conditions also persist; the SST anomalies in the Nino 1+2, 3, and 4 regions haven't changed much over the past month. Since there are no strong intraseasonal signals to use as Week 3/4 predictors, this forecast is based entirely on dynamical model guidance." Unfortunately, the EPS and GEFS (the dynamical model guidance) show a +NAO, ++EPO, +AO, and a slightly negative PNA during the first week of January. Bleak but does it really matter given this is not a window when it makes sense to make a week 3 forecast.

GEFS yes maybe, EPS no "Ambiguous" sum it up. Too far out there to matter.

Not according to the indices +NAO, +EPO (barely), +AO, -PNA for the early January period. Weak signal for New Year's eve storm

WRT to the cold CFS ... The following is part of the CPC 3-4 week outlook. Did the forecaster Kyle MacRitchie mean to say "over-forecasts cold in the Northeast" or am I missing something. Anyways they "forecast" a 50% (50%) chance of above (below) normal temperatures during week 3/4. They do forecast warmer than normal temperatures and precipitation over the southeast. The weenie in me says we could be on the cold end of something substantial. The ECMWF and JMA models, as well as a number of SubX models, forecast weaker troughing over Canada and suggest a warmer pattern over the eastern CONUS than the CFS does. Our calibration routine also lowers the cold probabilities in the CFS so that they appear more in-line with the ECMWF and JMA, which suggests that the CFS historically over-forecasts warmth in the Northeast.

To mix metaphors, we're punting the 2nd inning. No reason to be too upset, but its been a while since we've had a good tracking period over the Holidays. Not sure it matters but the ever-reliable GEFS and EPS ensembles predict that the AO and NAO will become positive shortly after the 1st - if true we could have uncooperative indices in both the Atlantic and Pacific. Our odds of getting rain during the next 10 days aren't that much higher than our odds of getting snow; however, the tail end of December and early January could be wet as the ridge relaxes - maybe a mixed event in there for us?

Not sure but I believe he used Brier scores

... am at the AGU meeting this week and I "snuck out" of my session to hear a talk on the Unified Forecasting System: Some highlights: Version 15.1 of the GFS officially became operational on June 12th of this year. It has a 13 km horizontal resolution and 64 layers. Data are assimilated at a 25 km resolution. Version 14 was retired on September 30. A minor change was made to the GFS in early November. It is now assimilating additional data sets including information from GOES-17. Version 12 of the GEFS is scheduled to become operational in August 2020. It will have the same dynamical core as version 15 of the GFS and its resolution will increase from 40- to 25-km and will include 31 members as opposed to the current 21. The system will be run 4 times per day with the 00 UT run being 35 days in length and the other runs 16 days in length. Apparently, the ensemble system has skill out to 10.4 days versus 9.8 for the current system with 2-3 hours of the additional skill due to the additional ensemble members. The next version of the GFS (version 16) is scheduled to become operational in January 2021. It is in the pre-operational stage now but has been frozen. It should have 127 levels compared to the current 64 and has bias fixes including (hopefully) one for the lower tropospheric cold -bias. Not sure what the fix is. Version 13 of the GEFS is scheduled to become operational in FY2023. It should be a "fully" coupled system that includes an ocean model, a wave model, an aerosol model etc.

Yes the ensembles hint that the AO may head positive again after sinking to neutral in the mid-range but that is in the hard to forecast 10-15 day the frame. Is the repeating pattern statistically significant? Please provide a reference for the new research on MJO cycling.

I guess this is a good time to remind myself that 21/21 of GEFS members showing >1" of snow for College Park does not equate with a 100% chance of >1". Fractions below depend on end point >12" 4/21 > 6" 6/21 >3" 12/21 "1" 21/22

You may cringe but I fondly remember the longest stretch with snow on the ground in this area since 1961 From the Capital Weather Gang

Wish it were this simple My_back_yard_snow = max (ensemble mean - 2", 0") Perhaps it's closer to this ... My_back_yard_snow = 0.667*ensemble_mean(d0-3) + 0.667*max(ensemble_median(d3-7)-2") + climatological_mean(d7-15)

CPC outlook for Nov 26 to Dec 2 calls for above normal temperatures and precipitation for Maryland with higher than normal uncertainty (see below). I'm curious as to why only the EPS is used in the blend. Is this common during uncertain periods because blends are only useful when the ensemble means are similar? --------------------------------------------------------- The official 8-14 day 500-hPa height blend consists of: 100% of Today's 0z European Ensemble Mean centered on Day 11 FORECAST CONFIDENCE FOR THE 8-14 DAY PERIOD: Below average, 2 out of 5, due to a relatively amplified pattern forecast over the domain, offset by significant differences between today's GEFS ensemble mean and the 0z ECMWF and Canadian ensemble means. FORECASTER: Mike C

6 Hours later - the torch is gone - Safe until 00 UT

I noticed recently that the 500 hPA anomalies from the EPS are based on a 1989-2018 mean. Aren't the GEFS anomalies based on 1980-2009? Shouldn't make much difference but perhaps shifting the 30-year period annually would make it less "crazy" to predict a colder than normal season. I also like focusing on the 7-10 day period. It seems like most of our future weather appears then although it doesn't come into focus until the 2.5 to 6.0 day period. Beyond that there are only a few background states where there is any skill at all unless a 52/48 weighted coin helps you. The EPS 15-day snow mean for the DC-area has been around 1" the last several days - all in the last week of course. What is a winter background value? 2-3"? and when do we typically reach it (last week of November)?

Disagree. 1989-1990 was one of the most memorable winters since I've been in the area. Thanksgiving eve snow storm. Spent Thanksgiving in NYC watching the Macy's parade in the snow. Early December snow. Snow loss through sublimation. Ice skating on NASA-Goddard pond. Pattern change between XMAS and New Years. On the medium range front, great to see that the number of EPS members supporting a significant snow in the DC area next weekend (yes that real nice Sunday) has risen from 1 to 3 over the last 24 hours.

Partly cloudy with a few flurries in College Park too

EPS College Park thru Nov 17 0" 12/51 <1" 16/51 1" 7/51 2" 10/51 3" 3/51 4" 1/51 5" 2/51

... Isn't there a way the NWS could pinpoint their warnings a bit better? During the last 3-4 weeks, I've been under 5 or 6 severe thunderstorm warnings and have not received a drop of rain from 4 of them

We'll have to wait and see here; a week out the Laurentian Mountains are certainly in the game

For the most part I was impressed with how well the models did in the 2.5 - 6 day time frame. The models were mostly useless in the 8-14 day time frame; however, this is a time frame when models only provide useful information during certain windows when teleconnections indicate that a longer range forecast should have skill (certain combinations of the QBO MJO ENSO post-stratospheric warming).

Pretty sure you're not serious but hope you're right; obviously, April 1st is very close to the time of year when only the day-10 model climate is cold enough for accumulating snow. It'll be memorable for sure. Keep the faith ocean temperatures are near their annual minimum. The EPS also has weak support for this "event" - at least north of DC

Over the last 3 EPS cycles (00 and 12 UT today and 12 UT yesterday) the percent chance of exceeding 3" has increased from 10-20% for the northern part of our forum; while the chance for the DC-area has remained at 10-12%. Of course 4 cycles ago (00 UT yesterday) the chances throughout the area were ~4%

8-14 DAY OUTLOOK FOR MAR 26 - APR 01, 2019 ... calls for above normal temperatures and precipitation Anomalous southerly low-level winds increase chances of above normal temperatures for parts of the eastern CONUS. ... the largest probabilities of above average precipitation in the East are restricted to near the coast as the storm system predicted in the 6-10 day period shifts eastward during the week-2 period. FORECAST CONFIDENCE FOR THE 8-14 DAY PERIOD: Below average, 2 out of 5, due to fairly good model and tool agreement over most of the domain, largely offset by significant differences between today's GEFS and ECMWF ensembles in the Eastern U.S.