Posted Saturday 01/31/26 @ 12:16 PM — Currently there’s more sunshine than forecast but some high clouds are making their appearance and based on the latest RRFS, I expect to see more cloudiness by 4 PM. The RRFS is also forecasting more sunshine than clouds on Sunday.
The latest Water Vapor imagery shows the deep upper level cold air trough extending down to Florida and an elongated surface low developing off the South Carolina coast. (our snow storm that will pass us by.)
Current satellite water vapor image with superimposed RAP model 500-1000 mb thickness lines (yellow), Mean Sea Level Pressure isobars (black contours) potential vorticity (fine violet contours) with superimposed MRMS RADAR. Water Vapor Imagery reveals emitted water vapor infrared temperature, NOT reflected light from clouds as is the case with visible satellite imagery. (Click on image for a larger view.)
For those of you who were disappointed with our missing the snow on Sunday, the models are hinting at a light snowfall sometime Wednesday. We will be on the northern fringe of this potential storm. Stay tuned
Originally Posted Fri 5:43 PM — This weekend we’ll be dealing with some of the cloud cover from the coastal storm moving to our east and the pressure gradient causing some windy conditions. No snow!
12zECMWF-AI forecast for 1 PM Saturday (Click on image for a larger view.)
Saturday Forecast
Mostly cloudy, but some breaks of sun and/or thinning clouds possible both morning and afternoon. Windy late afternoon
NBM high temperatures: Blue Bell, PA 20º Philadelphia, PA 20º Higher than Average Uncertainty (based on standard deviation): ± 3º
Sunday Forecast
Sunshine early, then periods of clouds and sun for the rest of the day. Windy much of the day, subsiding somewhat towards evening.
NBM high temperatures: Blue Bell, PA 23º Philadelphia, PA 24º Slightly Above Average Uncertainty (based on standard deviation): ± 2º
The outlook for next week: Some moderation of temperatures Monday and Tuesday, but continued cold through the week with a reinforcing cold front. Temperatures remain below seasonal averages.
Posted Thursday 01/29/26 @ 5:34 PM — The models continue the trend in forecasting the coastal storm to move further to our east. Even the Canadian Global GDPS shows no snow for Philadelphia, where it had been showing 8″ a few days ago.
The forecast period now falls in the range of the higher resolution models (84 hours or less). Both the RRFS and the Canadian RGEM forecast to 84 hours and both show no snow for Philadelphia and even no snow for coastal NJ.
Here’s the latest RRFS precipitation forecast for Sunday—
18z RRFS forecast snow – closest approach. (Click on image for a larger view.)
The cold weather continues with temperatures possibly rising into the low to mid 30s by Monday. Another reinforcing cold front, perhaps with some light snow the middle of next week.
Sunday’s “Storm” Looking Less Likely
Posted Thursday 01/29/26 @ 10:24 AM — The latest models through the night and this morning continue to show the storm to develop near the coast to track to our east. Accumulations in the city and westward continue to trend less and less likely.
Here’s the latest Model Blend (NBM) which statistically evaluates over 40 models. 75 percent of the models show zero, or significantly below this value—
13z NBM just became available. Continues a trend towards minimal or no snow in Philadelphia and westward. Several inches possible near the Jersey shore. (Click on image for a larger view.)
NBM median shows half the models show this or less—
13z NBM median snowfall for Sunday (Click on image for a larger view.)
The NBM mean still shows some significant snow, meaning that 25% of the models show much higher amounts skewing the mean higher. They are outliers; still a possibility but not likely.
The storm is now falling into the range of the higher resolution models (84 hours) Here’s the very latest Canadian RGEM, showing the storm totally missing our area—
Today’s 12z Canadian RGEM forecast for Sunday 12 PM. Storm precipitation totally east of us. (Click on image for a larger view.)
Updates later today.
Sunday’s Possible Storm Update
Posted Wednesday 01/28/26 @ 7:34 PM — There continues to be a strong signal that the coastal storm expected Sunday will pass us by. The models in this camp continue to be the GFS, AIGFS, ECMWF-AI, ECMWF and the ICON model.
The Canadian GDPS with “AI Spectral Nudging” model still maintains a heavy snowstorm for us and the Global Ensemble models are still forecasting snow in the 3-5 inch range for Philadelphia.
The storm is still in the 96-120 hour forecast period and only the synoptic-scale global models extend their forecasts out that far; the higher resolution models come into play at 84 hours and less.
The latest model blend (NBM) statistical take on this storm is interesting.
The Median Snowfall is below and is currently my choice (perhaps with some blend of the 75 percentile) for this storm—
19z NBM median snowfall accumulation. 50% of the models show this or less. (Click on image for a larger view.)
The NBM 75 Percentile also captures the next 25% of the models—
19 NBM 75 percentile. 75% of the models are predicting this or less than this. Since the first 50% are predicting zero, for Philadelphia, only 25% are predicting 1-3 inches, the rest zero. (Click on image for a larger view.)
Here’s the NBM MEAN snowfall which suggests that the remaining 25% of the models are predicting very high snow totals and is skewed high—
19z NBM Mean snow totals is skewed high due to 25% of the models showing heavy snow. (Click on image for a larger view.)
One of the models skewing the NBM mean high is the Canadian GDPS. It’s an outlier—
12z Canadian GDPS (Click on image for a larger view.)
Right now, I’m leaning towards the the NBM Median or a blend of the 75 percentile (zero-1″ in Philadelphia) based on the several models in that range. Stay tuned.
Sunday Storm Less Likely
Posted Wednesday 01/28/26 @ 11:13 AM — The current possibility of another major storm for Philadelphia is fading but not gone, especially for NJ. This current forecast of a Miller Type A nor’easter has demonstrated the power of a AI weather models with the AI models first showing the likelihood of a storm and also first showing the trend towards a more easterly track at our latitude.
The current forecasts, particularly with the ECMWF, ECMWF-AI and even the GFS and AIGFS have trended towards the storm entirely missing Philadelphia and/or only affecting mostly eastern NJ on Sunday. The German ICON model also has the storm well east of us.
A likely scenario is captured by the ECMWF-AI —
01-28-26 06z ECMWF-AI forecast for 1 PM Sunday. (Click on image for a larger view.)
On the other hand is the AI-ensemble hybrid HGEFS which still shows 2-3″ in Philadelphia. The westward extent is mostly due it the GEFS ensemble component to this model.
01-28-26 06z HGEFS forecast for 1 PM Sunday. Still a possibility of 2-3 inches in Philadelphia, but not a high likelihood. (Click on image for a larger view.)
This morning’s GFS, just available, captures the trend—
01-28-26 12z GFS forecast for 1 PM Sunday. The snow shield just brushes the coast of NJ. (Click on image for a larger view.)
The only model still showing substantial snowfall for Philadelphia is the Canadian GDPS with “AI Spectral Nudging” but it’s an outlier at this time.
The NBM mean snowfall incredibly shows 10″ in Philadelphia, but the current NBM does not include any of the AI models. The NBM median is closer to what some AI models are forecasting—
01-28-26 13z NBM median snowfall (Click on image for a larger view.)
So still some uncertainty. But a significant storm for Philadelphia continues to look less likely at this time based on the AI models. (Click on image for a larger view.)
Sunday Possible Storm Update
Posted Tuesday 01/27/26 @ 5:26 PM — Earlier I posted what was then the latest ECMWF-AI forecast, showing the expected nor’easter to graze the NJ shore this Sunday; the snow would not make it into Philadelphia.
The regular ECMWF and our own GFS models are currently moving in that direction. The latest GFS shows the storm’s snow shield barely making it into Philadelphia (but still showing ~1- 2 inches of snow.) The GFS predicts snowfall over 8-10 inches along the NJ shore.
01-27-26 18z GFS forecast for 1 PM Sunday. Snow barely makes it into Philadelphia, but significant snow possible at the shore. (Click on image for a larger view.)
The latest AIGFS is similar to the GFS forecast above. So we may be spared from a major snowfall. A 50 mile deviation in track would change this forecast dramatically.
Regarding uncertainty, here’s the latest ECMWF-Ensemble still showing high uncertainty. (Compare to yesterday.)
01-27-26 12 z ECMWF-ENSEMBLE mean sea level pressure standard deviation. (Click on image for a larger view.)
Posted Tuesday 01/27/26 @ 2:06 PM — I wanted to just do a quick update with the latest ECMWF-AI model which just became available. The ECMWF-AI keeps the storm to our east, a marked change from yesterday’s model runs.
01-27-26 12z ECMWF-AI model snow accumulation by Monday morning. The storm only brushes the NJ shore. (Click on image for a larger view.)
One model doesn’t make a forecast in this time horizon of near 120 hours. But I thought I’d post it to counter the NBM model’s suggestion of a possible but uncertain substantial snowfall. Stay tuned!
Sunday Storm Increasingly Possible
Posted Tuesday 01/27/26 @ 12:08 PM — As mentioned yesterday, there is much uncertainty regarding the forward speed and track of a Miller Type A Nor’easter expected to approach our area on Sunday morning. That said, additional models runs since my last posting (00z, 06z and 12z) increasingly show the storm to affect our area. The latest GFS model which just became available shows the storm hugging the coast to our south and then moving northeastward.
The latest Hybrid AI- GEFS (HGEFS) is showing significant uncertainty in the track of the storm, with a strong signal towards a westward track.
01-27-26 06z HGEFS pressure with standard deviation (uncertainty) at 7 AM Sunday. (Click on image for a larger view.)
Even with the above uncertainty, the HGEFS shows 3-5 inches of snow in our area by Sunday evening.
Several other models are also showing a similar 3-5 inches of snow at this time. Based on the trends, I expect it to run much higher, but it’s too soon to be really be sure.
It’s way too early to post snow total possibilities but here’s the latest NBM 13z MEAN snow accumulation forecast for Sunday!
01-27-26 13z Mean Total Snow Accum by 7 PM Sunday. The fine line contours are labeled in 1″ increments (Click on image for a larger view.)
The totals above will likely change in either direction.
There’s incredible uncertainty reflected in the models that comprise the NBM. The MEDIAN snow shows nothing for us!
01-27-26 13z NBM MEDIAN Snow totals (Click on image for a larger view.)
Rule of statistics: When the Median and the Mean are so different, there’s extreme uncertainty.
Stay tuned.
Another Snow Storm?
Posted Monday 01/26/26 @ 4:36 PM — Many of us probably just got done clearing snow after yesterday’s storm and hoping that would be end of it for this winter season. We may be in for another storm this coming Sunday.
The models started showing this possibility two days ago but several had vacillated back and forth with the track. Today’s models are moving towards forecasting a major Miller Type Anor’easter beginning early next Sunday. (Miller classified nor’easters into type A and B based on their origin. This storm originates in the Gulf of Mexico and takes form off the South Carolina coast.
The latest Canadian GDPS with “AI Spectral Nudging” illustrates the path of this storm, a typical Miller Type A—
01-26-26 12z Canadian GDPS with “AI Spectral Nudging” shows a monster Nor’easter early Sunday morning. (Click on image for a larger view.)
There is still uncertainty about the track and its closeness to our area.
Here’s the latest Hybrid AI GEFS (HGEFS) —
01-26-26 12z HGEFS at 1 PM Sunday (Click on image for a larger view.)
Here’s the latest ECMWF-AI—
01-26-26 12z ECMWF-AI forecast for 1 PM Sunday (Click on image for a larger view.)
The AI models have Philadelphia on the western fringe, but with heavy snow in NJ. If it happens, this will be an all-snow event.
It’s too soon to be concerned about this storm. The latest ECMWF-Ensemble (statistical) shows extreme uncertainty in its forward speed and path—
01-26-26 12z ECMWF-ENSEMBLE Sea level pressure uncertainty. This is as high as it typically gets. It suggests it might arrive sooner and or be closer to the coast or be further east. (Click on image for a larger view.)
Originally Posted Mon 10:47 AM — A cold week in store for us.
I’m already getting requests for comment on a major storm for next weekend. As of this morning, only the European AI and regular ECMWF were showing it affecting us. The GFS, GEFS, AIGFS, HGEFS, NAEFS, Canadian GDPS-AI and the German ICON have it moving east of us, not affecting us.
Unlike this past storm, where the models were more in-sync in this lead time frame, there are too many ‘votes’ at this time showing next week’s storm missing us. I’ll be updating later this afternoon with some new model runs. Stay tuned.
Mon 10:19 AM —Forecast Review — As the models moved into the final 24 hours, the forecast was pretty good.
The storm was first forecast over 8 days in advance by the AI models.
By Saturday, the models had refined the transition to sleet around noon to 2 PM, which was spot on. Freezing rain was over-forecast.
The total precipitation was about 1.3″ water, less than the 1.6″ forecast, so snow totals were lower than forecast.
The NBM mean snow forecast was closer than the median forecast.
The storm ended about when it was forecast that day (9 PM in the city) but earlier than previous day’s forecasts which had it extending into Monday and Tuesday. Some additional precipitation occurred through 11 PM.
The RRFS model , soon to become operational, which was very low on the precipitation forecast for several days in advance, but came closer to the actual by late Saturday in its 24 hour forecast.
Official NWS snow Totals—
Storm Review and Other Thoughts
Posted Sunday 01/25/26 @ 5:03 PM — The latest models show the mix tapering off significantly over between 6 and 9 PM near the city. A visitor to the site informed me that my blog settings had closed the comment feature to additional comments. I’ve corrected that issue. The same visitor asked about sleet freezing rain and snow—
“I’m wondering what the difference is between sleet and freezing rain. How do they impact things like shoveling and just getting around after this storm”
Well there’s a simple graphic that explains the difference and it comes down to this: the precipitation type depends how thick a layer of warm air (air temperature above 32º) the precipitation must fall through before reaching the ground.
Different thicknesses of air above 32º will result in different types of precip as shown below. (Click on image for a larger view.)
As for the following advise, I’m no expert. But the worst thing is letting the freezing rain and sleet hit the cold concrete. So I like to wait until the storm is over before shoveling, keeping the ice-sleet as the top layer. Otherwise you’ll have to contend with removing ice bonded to the ground. Of course one has to balance the total depth/weight and your ability to shovel it, so the incremental approach might work for some. I think ice is worse than snow when walking, driving etc.
As for the storm, it looks like the models over-estimated the snowfall. The near 10″ turned out to be close near the city, but still high. Here’s the official snow depth measurements for our area.
Changing Over
Posted Sunday 01/25/26 @ 12:17 PM — The snow is changing over to sleet early in the expected time window about 12 noon in my neck of the woods. We have about 7.5″ at the time of the changeover.
The RRFS runs every hour. Here’s the latest precipitation type forecast. I’ve drawn in the temperatures at around 6000 feet and about 3000 feet—
15z RRFS forecast for just after noon with precipitation type and thermal level temps. (Click on image for a larger view.)
A Note
Over the years, readers have written to me asking “what is the best radar and weather data app on the Apple App Store”. After trying so many apps, I recently discovered a contender for the best live weather and radar data— WeatherFront. The NEXrad radar on this app is as good as anything out there and its animations are real time. I pay for the added features and to support the developer, but an incredible amount of the data is available for free. Try it out! (I have no financial interest or gain in this app nor do I know anything about the developer. )
WeatherFront Screen capture live MRMS national view. (Click on image for a larger view.)
I should add that all the weather model graphics you see on this site are created using my own downloaded and post-processed model data and displayed with the weather model grib2 display app called LuckGrib on MacOS. I’ve customized the graphics displays, which LuckGrib fully facilitates. LuckGrib is also available for iPad and iPhone. LuckGrib does supply a range of model downloads at no charge, although I don’t use that feature. LuckGrib is a gem. Highly recommend it. (I have no financial interest or gain in this app nor do I know anything about the developer. except the developer is a top class programmer. )
Storm Update -Sunday
Posted Sunday 01/25/26 @ 10:21 AM — The snow began as forecast between 2 and 4 AM and the storm is playing out as forecast.
Based on a review of last night’s models and comparison with current conditions in the RTMA model, it appears that the NBM median snow totals is a best estimate. (plus or minus 2 inches) Here’s the NBM 07z median forecast, just before the snow started—
07z NBM model median forecast. Median = Half the models comprising this ensemble are showing more than these values, half are showing less. (Click on image for a larger view.)
Transition from snow to sleet is predicted by the 13z NBM to occur near the city between 12PM and 1 PM
13z NBM snow to sleet transition betweeen 12 and 1 PM (Click on image for a larger view.)’
Between 7 and 8 PM, the NBM shows this transition to have occurred—
13z NBM precipitation type between 7-8 PM (Click on image for a larger view.)
I wanted to get this online. More info coming.
Sunday’s Storm Forecast Update
Posted Saturday 01/24/26 @ 4:09 PM — I’ve looked at multiple models and here are the trends for Sunday’s snow/sleet storm—
Depending upon the model, light snow starts near the city between 1 AM and 4 AM, earlier in southern Chester county.
Very heavy snow expected Sunday morning. The snowfall may exceed 1 inch per hour.
From noon to 3 PM, the snow mixes with and changes to predominantly sleet, although a mix may re-develop if dynamic cooling brings the temperatures aloft down.
Temperatures rise from 3 PM to 10 PM but remain below freezing from the I-95 corridor north and west. Sleet followed by freezing rain likely from the city and north and west. Dangerous icy buildup possible.
East, in NJ, a transition to rain is expected during the 3 PM to 10 PM time frame.
Many models have the precipitation end about midnight. The HRRR has periods of sleet transition back to snow as late as 9 AM Monday.
As for snow totals, they have remained fairly constant over today’s models with the 10″ number highly predominant. The graphics below capture the likely snow totals—
12z ECMWF-AI Total snow accumulation. (Click on image for a larger view.)19z NBM median snow totals (half the models are above these levels, half are below this levels.) (Click on image for a larger view.)19z NBM MEAN snow totals. (Click on image for a larger view.)
The earlier GFS model also clustered around 10″.
Some thoughts: The soon to be released RRFS model seems to be an outlier with snow totals in the 4″ -7″ range and a rapid end to the storm. Either the RRFS is not ready for prime time or it is so advanced that it ‘knows’ some thing the other models don’t. I guess we’ll find out.
Another thought: Regular visitors here know that I often post expected rain totals. Those numbers are never exactly correct. They can only be ballpark estimate.
For some reason, when it comes to snow, weather forecasters (and I guess I have to include myself in that) erroneously believe we can nail down snow accumulations to the inch. The irony is that snow totals are based on the expected precipitation (basically rain snow equivalent) where any error is magnified by a multiplier of typically 10 x. Put another way, nobody gets upset if we get 0.2″ of rain instead of 0.4″ of rain. But with snow, it’s 2 inches vs 4. Just something to think about.
Storm Forecast Update
Posted Saturday 01/24/26 @ 9:32 AM — The storm is now well within the forecast time frame of the higher resolution models. The latest HRRR, just available, shows the snow starting around 4 AM (Blue Bell grid point) and ending for the most part by 10 PM with a changeover to a mix of sleet and freezing rain about 2 PM. A snow sleet mix is likely for an hour or two more.
Despite the changeover, it’s predicting accumulations in line with yesterday’s NBM median and mean forecasts—
Todays 12z HRRR snow accumulation through 10 PM. (Click on image for a larger view.)
The reason is most of the accumulation of the snow will occur before the changeover, essentially before 2 PM Sunday. However, an additional 1/2+ inch of ice and sleet will fall, which will accumulate on top of the snow after 2 PM. This will be heavy, and will likely be a hazard to trees and power lines.
HRRR snow accumulation meteogram for grid point Blue Bell – Wings Field. Snow starts about 4 PM. The major portion of the snow depth occurs before 2 PM. (Click on image for a larger view.)
More updates to follow. Stay tuned.
Storm Forecast Update
Posted Friday 01/23/26 @ 11:16 PM — Another quick update. The 01z NBM has the median snow totals about 2-3 inches lower than the previous 19z model graphics posted earlier. The reason- the changeover to sleet may occur as early as 2-3 PM. Updates in the morning.
Posted Friday 01/23/26 @ 8:49 PM — One more thing… I’m noticing the changeover to sleet and freezing rain may occur earlier in the afternoon Sunday and the push northwestward is all the way to Berks and Lehigh counties during the evening. This may further reduce snow totals.
Updates in the morning…
Storm Forecast- Late Friday Update
Posted Friday 01/23/26 @ 4:27 PM — I’m reviewing various models. As explained yesterday models have various ways of providing snow accumulation data. Many are not all that useful. Basically, models calculate snow from expected total water precipitation and using various algorithms involving snow-water ratios (temperature-dependent through the lower atmosphere) they compute the snow (referred to ASNOW or accumulating snow) Some only compute snow depth which includes density, compaction and melting (SNOD- snow depth) . Some only provide the water equivalent of accumulated snow depth (WEASD). It’s a mixed bag. And some models (like the ECMWF) only provide snow accumulation to paying subscribers.
As one of the people leaving a comment today accurately recalled, I used to use the NAM model data preferentially. Specifically, the NAM total precipitation in inches of water, multiplied by 10. I’d average that with the GFS precipitation similarly multiplied by 10. Based on temperatures at certain levels, I’d change the multiplication factor.
Very primitive but it works. But it really only works when it’s an all-snow event. This storm will have a sleet mix in that throws that simplistic approach out the window.
The model blend (NBM) uses very advanced statistical approaches in analyzing multiple model precipitation. The NBM is updated and run hourly, but the 01z, 07z, 13z, and 19z model runs have a special emphasis on precipitation. I’ve found it to be quite good and that’s what I’ve been leaning on.
Enough about that. What about the snow? Here are some model snow forecasts from this afternoon and this morning. The trend is downward, not because of less total precipitation (water equiv) but because part of it will fall as sleet and freezing rain—
Model Name
Blue Bell Grid Point
Comments
ECMWF-AI
9.0″
SRATE based
ECMWF
~ 9.0″
Derived from Snow Water Ratio
GFS
9″
SNOD
AIGFS
10″
No Snow Parameter Derived from Snow Water Ratio
Canadian RGEM
9.8″
SNOW RATE BASED
RRFS
7.0″
SNOD
German ICON
8.2″
HGEFS
~ 9.0″
No Snow Parameter Derived from Snow Water Ratio
Canadian Global AI
11.4″
Snow Rate calculation
The latest 13z (2 PM EST) NBM model has became available. Here’s the NBM MEAN snow accumulation forecast—
01-23-26 19z NBM MEAN snow accumulation (calculated) The fine line contours are labeled in 1″ increments. (Click on image for a larger view.)
The NBM Median Snow forecast—
01-23-26 19z NBM MEDIAN snow accumulation (calculated) Median = Half the models comprising this ensemble are showing more than these values, half are showing less. (Click on image for a larger view.)
With the median and the mean so close together, it suggests a normal distribution and a higher than average confidence.
The latest GFS shows the snow transitions to sleet and some freezing rain between 4 PM and 9 PM. Here’s the northern-most extent according to the GFS—
Latest GFS (18z) shows northern most extent of changeover to sleet occurs at 9 PM. Then transitions back to snow! (Click on image for a larger view.)
Updates tomorrow morning. Tonight if things change significantly. Stay tuned.
Storm Forecast Update
Posted Friday 01/23/26 @ 11:51 AM — I only had time this morning to review the latest NBM 13z data which has updated precipitation info.
Here’s an important insight from the NBM (Model Blend): It continues to show additional accumulation past midnight into Monday morning, adding a few inches in some areas.
I’m switching to the mean snowfall, away from the median, since the median is forecasting too high at this time. More about this later this afternoon.
01-23-26 13z NBM mean snow accumulation forecast (calculated) by 9 AM Monday morning. (Click on image for a larger view.)
Storm Forecast Update
Posted Friday 01/23/26 @ 8:09 AM — Interesting differences in the forecast have developed between models regarding the transition to sleet and freezing rain. Let me list the issues—
The total forecast quantity of precipitation (water or ice/snow-water equivalent) has changed little and remains quite high, on the order of 1.3″ water to 1.8″ water. If this were to be all snow, we’d still be talking about a 20+” snowfall. The models differ with the degree of transition, the northern-most extent.
The storm starts between 2 AM and 5AM Sunday morning (NBM) and about 5 AM (RRFS).
The models differ with the end of the storm. The NBM suggests snow continues to accumulate well past midnight Monday as a return to snow into Monday morning. The RRFS has it cutting off about midnight.
The models differ significantly in timing and degree of transition to sleet and freezing rain near the city. The graphics below capture some of these differences–
06z GFS snow to sleet transition occurs late, about 9 PM. The transition does NOT move very far north before retreating. (Click on image for a larger view.)06z RRFS snow to sleet transition. This transition moves well into Montgomery and Bucks counties at 8 PM (Click on image for a larger view.)
The NBM model blend shows the greatest degree of change—
12z NBM rain sleet snow transitions at 7 PM . (Click on image for a larger view.)
Despite these differences, the NBM mediansnowfall is higher than what I’m hearing on the radio. I’m using the median because it best reflects the transition to sleet/freezing rain. Since the NBM does not provide the median over the time period of an entire storm, the graphic below is based on my own algorithm—
07z NBM median snowfall (calculated) (Click on image for a larger view.)
The next NBM with hourly precipitation data is the 13z run. I’ll update this later this morning, probably towards noon with new morning data. Stay tuned.
Storm Update
Posted Thursday 01/22/26 @ 7:44 PM — I had a chance to review some additional models since my last post.
First, the snow will start near the city between 11 PM Saturday and 3 AM Sunday. Heaviest snow will occur around noon into early afternoon. A mix with freezing rain is likely after 4 PM according the NBM and 8 PM according to the RRFS, which is forecasting sleet. It ends during the morning on Monday, and no longer appears to linger through the day.
Here are some snow totals from individual models:
Model Name
Blue Bell Grid Point
Comments
ECMWF-AI
11.4″
ECMWF
9.5″
Derived from Snow Water Ratio
GFS
11.6″
AIGFS
12.2″
No Snow Parameter Derived from Snow Water Ratio
Canadian RGEM
8.9″
RRFS
11.4″
German ICON
9.2″
HGEFS
7.0″
No Snow Parameter Derived from Snow Water Ratio
So why are these numbers so low compared to what was posted from the NBM? I don’t exactly know, but I’ve been doing this for years and the built-in snow parameters (Snow Depth- parameter SNOD) in the models are often way off ; they assume snow pack compaction, melting etc.
The NBM and the RRFS have a parameter called Accumulating Snow (ASNOW) which I find to be more useful.
Generally, what works in this region is computing actual snow from snow-water ratios. What makes this so difficult this time around is the snow-water ratio at the start of the storm may be 16:1 or more but when it’s mixed with freezing rain, it might drop to 6:1 or less.
In the past two snowfalls this season, the NBM did the best and snow accumulations fell above the mean and much below the 75 percentile.
Some other information: The NBM shows freezing rain about 4 PM. The latest RRFS shows a transition to sleet, not freezing rain about 8 PM in the city. Too soon to know.
My next update will be Friday morning after the 13z NBM becomes available. I’ll update earlier if tonight’s models show major changes.
Thursday Storm Update
Posted Thursday 01/22/26 @ 5:26 PM — The storm continues to present challenges to a correct snow forecast.
As posted earlier, there’s an increasing trend for warm air to infiltrate at levels between 6000 and 10,000 feet. As a result, the models are showing a partial transition to freezing rain and snow about 4 PM Sunday near the city.
This will play havoc with the snow accumulation totals, which up to now, were assuming an all-snow event at low temperatures. While big snow totals are always impressive, the likelihood of ice accumulation may further increase the hazards of this storm.
Here’s the latest NBM showing a transition to freezing rain and snow at 4 PM Sunday—
18 NBM Precipitation Type at 4PM. This does NOT show coverage like radar……It shows conditional probability. (Click on image for a larger view.)
The latest NBM median accumulation is very similar to the mean calculation, suggesting a higher confidence in these numbers (for now).
01-22-26 19z NBM MEDIAN snow accumulation. Median = Half the models are showing more than this, half are showing less. These totals are very high despite a mix of freezing rain near the city. Not sure how this will all play out. (Click on image for a larger view.)
The NBM 75 percentile snow accumulation is below. Don’t hang your hat on this. (75% of the models are showing this value or less; 25% of the models are showing this value or more!)
01-22-26 19z NBM 75th percentile snow accumulation. For those who believe we’ll have an outlier event. Not likely. (Click on image for a larger view.)
I’m probably going to post some additional info this evening. The 01z NBM becomes available about 9:45 PM. I’d like to look at the latest ECMWF and the RRFS which is now coming into forecast range. Stay tuned.
Originally Posted Thu 9:59 AM — I’ve decided, perhaps prematurely, that this impending snow storm deserves its own posting, rather than fitting into my regular “Weekend Weather Forecast” and “This Week’s Weather” formats.
Those of you visiting this site for the first time might want to review my earlier posts on this storm here for some background insights.
This storm was first picked up by the AI models last weekend. Timing has been remarkably unchanged but the details have continued to evolve.
The expected storm onset has just begun to enter into the coverage of some of the higher resolution models (84 hours). So we’re getting more insight.
Here’s the latest GFS forecast for midnight Sunday (Monday) —
01-22-26 06 GFS forecast for Midnight Sunday into Monday. Lingering system! (Click on image for a larger view.)
Here are the current trends—
The storm continues to show increased intensity and development closer to our area. The double center low pressure system described yesterday is still in the picture, but the GFS continues to show the secondary low off the coast to be predominant.
Based on the higher resolution RRFS, the storm starts between 2 AM and 4 AM Sunday.The NBM has it starting 4 hours earlier.
The latest GFS and ECMWF has the storm lingering well into Monday. The ECMWF-AI and the HGEFS (Hybrid-AI) all concur. This suggests considerable additional accumulation on Monday.
With the closer approach of the system, it appears that warm air will be brought as a layer around 10,000 ft. This suggests that some areas will see a mix of sleet and even freezing rain with the snow later Sunday afternoon. The ECMWF is warmer than the GFS. (The GFS has a tendency to be too cold. We’ll see.)
01-22-26 06z ECMWF-AI shows critical thickness lines (red, magenta) north of our area at 7 PM Sunday. This shows a layer of warmer air at 6000-10,000 feet causing a mix to freezing rain and sleet for a time. (Click on image for a larger view.)
Here’s the latest NBM forecast Snow Totals—
07z NBM Mean snow totals. by Tuesday. The NBM keeps the precipitation predominantly snow despite the warm layer mentioned earlier. (Click on image for a larger view.)
The above snow totals are impressive and the NBM shows a major burst in accumulation late afternoon on Sunday.
Here’s the latest NBM 13z just off the wire. The totals have increased and extended into late Monday/Tuesday—
13z NBM Increased snow totals through Tuesday. (Click on image for a larger view.)
The NBM 75 percentile is near 30″. I won’t post that yet, since it suggests an outlier event.
This storm is going to be memorable and promises to be a forecasting challenge. I’ll update late this afternoon. Stay tuned.
