high risk

The NAM and especially the NAM3 can be a bit aggressive with the low-level cold air, but I'd give way more weight to them than to the GFS which is notoriously poor with shallow cold air masses. If you look at the hi-res windows, while they differ on the timing, they all bring the front south of DC tomorrow. While I have questions about how far south it will get, I have few doubts that the boundary will make it south of DC by midday Sunday at the latest.

huge differences between the NAM and GFS for Sunday. NAM brings a back door cold front through the area early Sunday and makes it a fairly unpleasant day here. GFS keeps us in the warm sector.

because the snow accumulation (which includes sleet) is from the model microphysics, while the precip type is from the wet bulb temperature profile

yet again, the Euro follows the NAM's lead, and the GFS stays off on its own island.

strong gusts and brief but intense rain in southern Howard County

line of heavy showers now forming along the front

Newest NAM3 still wants to develop a forced line of convection right on top of the DC Metro area during the late morning, with it looking decent for the east side of the area and into northeast MD and the eastern shore. The HRRR brings the overnight convection further east and kills it late which seems to inhibit better development in our area during the morning. The HRRR would be a big whiff; the NAM3 isn't great, but areas to the east could certainly see some stronger wind gusts.

500-750 j/kg of sfc-based cape seems doable for Wednesday morning, and the shear looks good, with strong wind fields in place through the column. With the strong forcing, I expect a squall line as opposed to discrete cells, with a decent wind threat. The NAM3 seems to show this well, as I expect that we'll end up with a slight risk at some point. The risk can go to enhanced if we end up with greater instability.

I answered where you posted this in the model thread.

The HREF takes the most recent NAM3, HiRes Window ARW, HiRes Window NMMB, and HiRes Window ARW2 (effectively the NSSL WRF) and the previous runs of each and aggregates them into means and probabilities. It's simply the max value from the 8 members at each point.

00z NAM suite seems to be removing any concerns raised by the 18z GFS. The only thing I don't like is that they continue to show us losing some initial heavier precip late tonight to sleet, but a quick inspection of the soundings shows that the offending level is just barely above 0C, so it wouldn't take much to have those profiles be all snow.

it absolutely did cut down on precip after 12z Wednesday. could just a be a blip, but it sure wasn't a happy hour run.

NAM3 is a crushing too. I still wish that the NAMs didn't waste some initial precip early Wednesday as IP/ZR, but there are still plenty of hours of SN+ in both.

gempak and lots of cut-n-paste.

one more comment regarding GFS ptype, relevant for this afternoon and for early Wednesday. These were 24h ptype forecasts for the March 2017 storm when there was a lot of IP/ZR in the I-95 corridor. The GFS completely missed the transition zone.