Fri 11:09 AM —Forecast Review — Snow?? Some of us woke up to a light layer of snow (less than 0.3 “). Since I didn’t see this coming (nor did the “Exclusive” forecast I watched on TV around 11 PM), I did some forensics to see what happened.
Going back to yesterday’s afternoon models (18z), both the RRFS and the NAM-NEST showed light radar shadows around 1 -3 AM; so light, I usually discount that level as model ‘noise’ especially from the NAM-NEST. The 18z NBM showed nothing.
I went off-duty 😉 after I posted here at 6 PM.
The next model runs (00z) (available after 8:30 -9:45 PM) of the RRFS showed no precipitation, except for a ‘short wave’ at 700 mb (10,000 feet) and high moisture content around 1 AM.
The 00z NAM-NEST and 00z HRRR DID forecast light snow in the 1 AM-4 AM time period, close to what we received.
00z NAM-NEST accumulated snow by 7 AM this morning. (Click on image for a larger view.)
So the take-away is the RRFS in its current form (the NOAA scientists have intentionally damped the convective precipitation in the current version because it was forecasting too high during development) is possibly now insensitive to light precipitation events. Something to keep in mind. NOAA is actively working on the next RRFS version which will have an entirely different ‘core’.
As for why the “Exclusive” forecast didn’t say a word about the snow… we’ll have to wait to hear if anything is mentioned.
Friday and Weekend Outlook
Posted Thursday 01/01/26 @ 5:42 PM — The models are in general agreement that the next few days will be dry and quiet weather-wise.
Earlier this week, it had appeared that several systems might affect us, but they now are forecast to move off to our south on Saturday and our north on Monday. The trend will be slowly moderating temperatures over the weekend with partly to mostly sunny skies.
By next Thursday and Friday, the AI models are forecasting highs 55º-60º. Wouldn’t that be nice! Unfortunately, it appears that rain will be in the picture on Friday.
Some additional insights. The TV weather people love to talk about weather “patterns” and often throw around that term when they really should just be saying there will be a “change in the weather”. With the ECMWF-AI forecast shown below for next Friday, there really is a sign of a true change in the large scale weather pattern.
The central part of the country has been incredibly mild, as a dip in the jet over the Northeastern US has left a ridge over the Midwest and South central regions. The forecast for end of next week suggests the dip in the jet will be in central US and we’ll be in the upstream warmer ridge and possibly wetter flow.
It will also open us up for Gulf moisture and the potential for coastal storms. To soon to know for sure, but expect changes beginning the end of next week.
Today’s ECMWF-AI forecast for NEXT Friday evening. If the high pressure system sinks straight south (1), a pattern change is underway for the second week in January. If it follows path 2, the pattern change may just be a weather change. (Click on image for a larger view.)
Forecast Review
Happy New Year!
Thu 7:03 AM —Forecast Review and 2026 Preview — Yesterday’s RRFS forecast for a snow squall this morning was quite good. Anyone awake about 5 AM briefly saw whiteout conditions as the predicted snow squall moved through. Many areas now have a 0.25″-0.5″ of snow covering. Here’s the 18z RRFS graphic from yesterday that I did not have time to post at that time—
Yesterday’s 18z RRFS forecast showing the squall line and 1 hour prior precipitation at 6 AM. (Click on image for a larger view.)
Here’s the radar image at 5:19AM—
Radar Image with NWS Snow Squall warning box. Regular readers might notice that I’ve changed from over 20 years of using WeatherTap.com for live radar to the new WeatherFront app on iPad and iOS. Excellent live weather data and radar. I’m paying for a subscription to support the developer, but much of the data on the app is available for free.I highly recommend this app. (I have no financial or any other connection to WeatherFront.) (Click on image for a larger view.)
2026 promises to be exciting from a weather data perspective. AI models continue to be advanced and developed . The RRFS and the REFSensemble models are expected to become operational [early] this year. The NBM will be upgraded to version 5.0 sometime this Spring.
Several models will be retired in 2026 including the NAM, SREF, the NARRE, all three variants of the HIRESW along with the HREF models
Long-time visitors here know I’ve loved the NAM for snow forecasts, but the RRFS looks to be a much better model.
So here’s to 2026!
New Year’s Day
Posted Wednesday 12/31/25 @ 5:40 PM — One more quick update. The latest RRFS, just available, shows a fast moving squall line of light snow moving through about 6 AM Thursday morning. A bit more coverage of light snow than depicted earlier with the NBM. I guess we’ll see how good the RRFS is.
Posted Wednesday 12/31/25 @ 4:37 PM — Happy New Year! A quick update. The light flurries predicted for this evening, New Year’s Eve, are no longer forecast.
A strong cold front will move through before daybreak Thursday. The light snow Some light snow showers are expected with this front and the model blend (NBM) shows low probability of a coating of snow in a few areas northwest of the city—
18z NBM forecast snow accumulation (mean) by 6 AM Thursday morning. Extremely light, if anything. Note the position of the 32º line. This sinks further south throughout the day as temperatures fall through the day! (Click on image for a larger view.)
The latest ECMWF shows a bit more coverage than the NBM, but still very light and scattered. For New Year’s Day, the models are predicting somewhat higher wind gusts, over 30 mph.
Temperatures drop after 6AM. Here’s the calculated wind chills based on the RRFS model for location: Center City Philadelphia.
RRFS wind chill forecast. Into the teens! (Click on image for a larger view.)
Into the new year, dry through Tuesday with a warm up by Tuesday. Rain, not snow later Tuesday.
New Year’s Eve – Cold, Less Windy with Flurries
Posted Tuesday 12/30/25 @ 4:51 PM — It really felt cold today with the continued gusty winds that are finally expected to diminish somewhat on Wednesday and especially New Year’s Eve.
Water Vapor image with superimposed RAP model data. Cold front is drawn with snow showers showing on radar ahead of the front. (Click on image for a larger view.)
However, another reinforcing cold front is poised to move through here later New Year’s Eve, preceded by snow flurries and light snow showers. The NBM shows these flurries as accumulating a dusting to a coating Wednesday evening after 8 PM .
Additional flurries and snow showers into 8 AM Thursday morning.
19z NBM mean snow accumulation by 2 AM New Year’s Eve. (Click on image for a larger view.)
As mentioned yesterday, New Year’s Day looks quite cold and windy with highs [maybe] just making it to 32º.
For the Mummer’s Parade, wind chills will be near 20º
Windy and Cold
Posted Monday 12/29/25 @ 9:56 PM — No big storms on the horizon. Windy and cold through New Year’s Day and beyond. A mixture of instability clouds and sunshine. The winds today have been impressive—
6:15 RTMA model record wind gusts. There have been higher wind gusts at other times. (Click on image for a larger view.)
Speaking of New Year’s Eve/Day, a disturbance may bring some snow showers early New Year’s Day morning as a reinforcing cold front moves through. The Mummers Parade will have to deal with 30 mph winds and below freezing temperatures.
The next chance of snow (or another “mixed event”) appears to be Monday, according to the most recent ECMWF-AI model.
Previously Posted Sun @ 8:19 PM — —A warm front moves through tonight followed by a sharp cold front moving through between 11 AM and 1 PM. Some light rain is expected Monday morning with the warm front and ahead of the cold front.
18z HGEFS (AI hybrid) forecast for Monday at 1 PM. (Click on image for a larger view.)
Temperatures will be quite mild and peak about 11 AM, but with the cold front, temps will drop during the afternoon and will be below freezing by 7 PM.
RRFS Wind and Gust forecast meteogram for Blue Bell, PA (Click on image for a larger view.)
It will also be increasingly windy in the morning and strong wind gusts will occur with the cold front passage into the afternoon, possibly to 40+ mph.
The word for the week will be “cold’. Cold high pressure will dominate through New Years Day. There’s a slight chance of snow flurries or a coating of snow New Year’s Eve. No major storms are forecast this week, but things may become more interesting over the weekend.
Sat 8:40 AM —Forecast Review — The models (and the forecasts) had trouble with this storm. There was much more sleet and freezing rain than snow over our region than the day-before forecasts had suggested. While the models did correct themselves over the day Friday, several morning model runs were still predicting inches of snow.
Regarding forecasts, I was surprised to see so much sun this morning; it was supposed to be cloudy. The RRFS and the HRRR are still predicting low clouds for most of today. We’ll see.
Originally Posted Fri 8:05 PM — Following the exit of the sleet, freezing rain, and some snow, low level clouds will be with us for much of Saturday. An upper level ridge will bring potentially milder temperatures very late Sunday into Monday and an approaching cold front will bring rain Sunday night into Monday.
18z AI-GFS forecast for Sunday. Upper level ridge with briefly warmer temperatures by late Sunday evening into Monday before a cold front moves through Monday, bringing back frigid temperatures. (Click on image for a larger view.)
Saturday Forecast
Mostly cloudy for most of the day. Continued cold. Clearing towards sunset. Somewhat windy.
NBM high temperatures: Blue Bell, PA 34º Philadelphia, PA 36º Low uncertainty (based on standard deviation):± 1.3º
Sunday Forecast
Considerable cloudiness with some sunny breaks possible. Increasingly cloudy late afternoon. A few sprinkles in the evening. Rain at night.
NBM high temperatures: Blue Bell, PA 39º Philadelphia, PA 41º Average Uncertainty (based on standard deviation): ± 1.8º
Posted Friday 12/26/25 @ 4:27 PM — The latest radar and short range models increasingly suggest the snow will be the least important aspect of of this winter event. The earlier HRRR animation (below) along with the NAM-NEST model are consistently showing minimal snowfall. This will be increasingly a sleet, freezing rain and ice event.
Current Radar at 4:20 PM shows very light snow and freezing rain to our west. (Click on image for a larger view.)
The latest NAM-NEST captures my current view of this storm—
18z NAM-NEST shows minimal snow with tonight’s storm.
Since the NBM did so well with our last snowfall, I was inclined to go with it again. However, the NAM-NEST has consistently been predicting very low snowfall, with the storm producing more sleet and freezing rain.
Additionally, the latest NAM-NEST had reduced the total precipitation from 0.40″ water equivalent to closer to 0.20″ water equivalent.
Continuing with the NAM-NEST, here’s the expected precipitation type at 11 PM—
18z NAM-NEST precipitation type at 11 PM (Click on image for a larger view.)
Friday’s Snow Sleet Ice Update
Posted Friday 12/26/25 @ 12:13 PM — The latest HRRR shows a 2-5 PM onset. The HRRR is showing more freezing rain than sleet at the onset.
16z HRRR time sequence 3-5 PM. The way models work, this is not like radar. The actual precipitation shown is the for the previous hour, not the instantaneous hour. Which means that what you see may occur 1 hour earlier. Red is freezing rain Pink is sleet Purple is snow (Click on image for a larger view.)
Snow Sleet Ice Update
Posted Friday 12/26/25 @ 9:21 AM — The latest models continue with the trend from last night, predicting slightly more snow accumulation further south than previous models had shown.
Here’s the current forecast and uncertainties—
Snow begins between 3 and 5 PM, moving in from the northwest.
Most models show a changeover to sleet in most locations near the city and the immediate surrounding counties.
The NAM-NEST consistently forecasts very light accumulations of snow with an easterly wind bringing mild air into the region before midnight. The NAM-NEST shows most areas less than 1/2 inch or less!
The REFS (experimental) shows significantly more snow and colder temperatures.
Precipitation ends between midnight and 3 AM Saturday
Latest HRRR forecast:
12z HRRR Snow depth parameter (Click on image for a larger view.)
The latest REFS (experimental)—
06z REFS snow accumulatioin forecast (Click on image for a larger view.)
The latest NBM (13z) with updated precipitation data. Median Snowfall—
13Z NBM forecast snow accumulation (median) (Click on image for a larger view.)
To a large extent, the major issue with this storm will be icing and sleet, freezing rain in some areas. Predicting snow totals with compaction, ice on top, etc is just an exercise.
Friday Winter Weather Update
Updated Thursday 12/25/25 @ 11:18 PM — Tonight’s early models show an increase in the southern extent of the snow accumulations. Add about 1 inch to the map posted below. Updates tomorrow morning.
Posted Thursday 12/25/25 @ 8:43 PM — Additional models during the day have the following trends—
Total precipitation (total water equivalent) in our area has reduced to the 0.30″ to 0.40″
A warmer influx aloft will quickly change the majority of the precipitation to a wintry mix of mostly sleet mixed with snow in the north and freezing rain in the southwest of Chester County.
Precipitation begins light, between 3 and 5 PM
Precipitation ends for the most part before daybreak, but snow flurries are likely along with clouds on Saturday.
I’m using the latest NBM 00z, just available—
00z NBM PTYPE forecast for Friday at 4 PM (Click on image for a larger view.)
By 8 PM—
00z NBM Ptype forecast for Friday at 8 PM. (Click on image for a larger view.)
By 5 AM, the precipitation has mostly ended, except for flurries. Here’s the NBM median forecast total snow. THE ABOVE DOES NOT INCLUDE AN ADDITIONAL 0.25″+ WATER EQUIV SLEET AND ICE FROM FREEZING RAIN—
00Z NBM Median (not mean) snow totals. (Click on image for a larger view.)
I’m using the median vs mean amounts because it more closely approximates the NAM-NEST and the Canadian RGEM.
Despite the low snow totals, traveling will be treacherous.
Updates tomorrow morning.
Weather Update
Posted Thursday 12/25/25 @ 11:42 AM — Merry Christmas and Happy Holidays! First let’s look at today’s weather. There was more cloudiness this morning than predicted by the forecasts I heard last night. A disturbance passing to our south caused the cloudiness. Clearing should develop in the next hour or so.
Satellite with superimposed radar and RAP model Omega at 11:30 AM. Clearing on the way. (Click on image for a larger view.)
As for the Winter Weather for later Friday, the models are coming together for more of an sleet/snow storm here in Philadelphia and the immediate suburbs, while areas to our northeast, starting with Trenton and Doylestown may have significant snow accumulation. What’s saving us from a 4-8 inch snowstorm is warm air that is expected to move in about 5000 – 6000 feet above ground level. Ground level temperatures are still expected to remain below freezing.
Below, I’ve posted the latest NBM median snow depth at 1 AM. Don’t be fooled by the low numbers; a significant accumulation of sleet and ice will develop that is not included in the calculated values below.
12-25-25 15z NBM MEDIAN Snow accumulation by 1 AM Saturday. The NBM shows no further accumulation (Click on image for a larger view.)
A rapid changeover from snow to sleet will occur from south to north.—
NBM precipitation type at 10 PM (pink sleet. blue snow) (Click on image for a larger view.)
Precipitation begins 3-5 PM Friday. Ends Saturday before daybreak.
Friday Winter Storm Update
Posted Wednesday 12/24/25 @ 5:34 PM — A review of the new models available reveal the following trends—
The NBM mean and median snow accumulation have come closer together. The amount of snow forecast is lower than previously shown here due to a faster changeover to sleet and a mix in Philadelphia and the immediate suburbs.
Temperatures in the city and surrounding counties remain below freezing. Dangerous icy conditions are likely.
There’s significant differences between the European/Canadian models and the NOAA models regarding temperatures in the upper atmosphere which relate to whether what falls will be rain, freezing rain, snow or sleet.
The new GFS AI model and the experimental RRFS are colder and suggest more snow than forecast by the NBM.
Let’s go to some specifics. The GFS shows it’s sleet snow line with along the classic 700-850 mb critical thickness line—
18z GFS forecast for 9 PM Friday. Warm upper atmosphere (red line) and changeover along the 1540 thickness line (a measure of temperature about 6000 feet). (Click on image for a larger view.)
The experimental RRFS shows more snow initially with a colder forecast. I should add that the RRFS is showing a narrow band of precipitation that is unlike any other model. Perhaps that’s why it’s still experimental…or perhaps it knows something the other models don’t —
18z RRFS shows more snow and a critical thickness line that is further south, keeping our area in the snow region. (Click on image for a larger view.)
The NBM does the heavy lifting. Here’s its median forecast (see earlier discussion regarding mean vs. median)—
19z NBM median snowfall forecast. (Click on image for a larger view.)
What’s my current take on all of this? I’m actually leaning towards higher totals than shown above. How much higher to be determined.
Updates tomorrow. Happy Holidays and Merry Christmas!
Friday Winter Weather Update
Posted Wednesday 12/24/25 @ 11:01 AM — As promised, here’s the latest Model Blend (NBM 13z ) with updated precipitation (mean snowfall)—
13z NBM accumulated mean snowfall (Click on image for a larger view.)
As was the case earlier, the mean and the median are very different. This indicates that there’s uncertainty and the statistical distribution of all the models comprising the NBM are skewed and asymmetric.
While our previous snowfall about 10 days ago had a more similar mean and median, making the forecast highly certain, this system’s model’s means and medians are very different. (Likely a handful of high rollers are skewing the mean towards a higher value.)
In this case, the rule of thumb is to lean towards the median. Here’s the median forecast. (Half the models forecast more than this value, half the models forecast less.) It better captures the rain/sleet/snow transition and the ‘warmer’ models in a skewed distribution—
13z NBM median snow accumulation. (Half the models have more than this value, half the models have less.) (Click on image for a larger view.)
I expect the mean and the median come together as we get closer to Friday.
In all cases, the snow starts 3-5 PM Friday afternoon with a [possible] changeover to a mix south of the city. It ends Saturday morning from south to northeast.
I’ll update again later this afternoon.
Friday Winter Weather Update
Posted Wednesday 12/24/25 @ 8:03 AM — The latest model forecasts are less in agreement than yesterday’s models. Here are the trends—
The storm will split into two lows, diminishing the total water equivalent falling. Many models have backed off from 0.70+” to 0.30″ water equivalent. That reduces potential snowfall.
The temperature profile from the latest HGEFS (AI) is significantly colder, suggesting whatever precipitation falls will fall more as snow.
The current Model Blend (NBM) has large differences between the mean snowfall and the median snowfall, suggesting large differences.
Here’s the latest HGEFS (AI) forecast for 1 AM Saturday. The HGEFS does not explicitly predict snow, but PTYPE as snow can be inferred from its other temperature parameters—
06z HGEFS shows two lows, with 32º temperature contours at critical atmospheric levels for snow all below freezing, reducing the chance of a mix north of these lines and suggesting more snow and less ice. (Click on image for a larger view.)
Here’s the latest NBM (Model Blend) with mean snow totals—
07z NBM (Model Blend) with mean snow totals. Note that several models do NOT show the precipitation maximum in NYC. The blue contours are 1″ increments. Additionally, the NBM mean and median predictions are very different, indicating uncertainty. (Click on image for a larger view.)
The next NBM run with updated precipitation data is the 13z run, available later this morning. I expect it to be different from the above. I’ll update by noon. Check back.
More on Friday’s Potential Storm
Posted Tuesday 12/23/25 @ 7:57 PM — The models have good consensus regarding a moderate precipitation event for later Friday afternoon into Friday night. As much as 0.40″ to 0.60″of water will fall, as snow sleet or freezing rain/rain.
An all-snow event would give us 4-7″ of snow, BUT temperatures in the atmosphere between 5000 and 10,000 feet will rise above freezing during the storm, reducing accumulations but making for a potentially high ice condition.
Here’s the latest RRFS model for the precipitation type expected—
Today’s RRFS forecast “PTYPE” at 1 AM Saturday morning. This will likely change as the week progresses, but it doesn’t look good, as ice can be worse than snow. (Click on image for a larger view.)
Stay tuned for updates.
“Interesting” Weather Setting up for the Weekend
Posted Tuesday 12/23/25 @ 4:08 PM — A strong jet stream and subtropical jet stream flow of very moist Pacific air has been affecting areas in California. The energy and moisture are poised to move in for the weekend and will cause the development of several areas of low pressure affecting the Mid-Atlantic region over the coming days.
Water Vapor image at 4 PM with superimposed RAP model pressure (black). 500-1000 mb thickness (yellow) and potential vorticity contours (violet) Dark Blue areas are high level moisture. The white arrows capture the current model track for these areas of energy and moisture. (Click on image for a larger view.)
Over the last several years, the possibility of snow in our area seemed elusive. Not so much right now with the current pattern.
As shown with the water vapor image above, we’re set up for disturbances going over the upper air ridge in the center of the country and then descending upon us in an area that’s an upper air trough. We’ll likely be on the cold side of the storms that develop from this pattern.
Last week, the NWS-NOAA made their AI versions of the GFS and GEFS operational. Of particular interest is a third model now known as the HGEFS, a Hybrid model that combines the new AIGFS and the AIGEFS.
The HGEFS consists of 62 models, more specifically 30 “perturbations” of the GEFS and 30 “perturbations” of the AI (Machine Learning) GEFS plus 2 ‘control models’.
The 30 “perturbation” models of each are creating by complex mathematical vector breeding of the main “control” models. These models statically compensate for unknowns in the data input, errors from the mathematical blending of the grids, and rounding/truncating errors from equations that are infinite series.
All 62 comprise this model group or what’s called an Ensemble model. (That’s the “E” in GEFS). The HGEFS promises to be an advance in longer range forecasting. Details can be obtained here from NOAA.
So enough with the technical stuff. Here’s what the latest HGEFS shows for late Friday into Saturday—
Todays’ 12z HGEFS with 1000-500 mb thicknes (red) 850mb (5000 ft) 32º line (violet) and 925mb ( 2000 ft) 32º line (yellow) . The white line is the mean 32º near ground temperature and the fine white lines show standard deviation of uncertainty in the mean 32º line. Notice that Philadelphia is below freezing at several levels as the storm approaches at 7 PM. (Click on image for a larger view.)
This storm has the potential for more than 4 inches of snow, according to the HGEFS and the regular GFS. Stay tuned.
Early Wet Mix Near the City
Posted Monday 12/22/25 @ 5:54 PM — The latest models continue with a similar forecast as posted yesterday. The immediate Philadelphia area and south will be above freezing when the precipitation starts. However, a layer of cold air about 4000 feet aloft will cause any snow to turn to a mix of wet snow and sleet and rain near the city and the immediate suburbs.
The GFS captures the general picture. The white (32º) line near the surface will be north of the city at 7 AM. However. the magenta line running through Honey Brook (below) shows that north of the line, precipitation will fall as a frozen mix of snow and sleet mixed with rain.
Today’s 18z GFS showing light precipitation has started at 6 AM and accumulated somewhat by 7 AM. The near ground freezing line and critical thickness are shown. (Click on image for a larger view.)
By noon, most of the precipitation has ended having changed to rain in most areas by 9 AM to 10 AM. Here’s the latest NBM with updated snow accumulations—
19z NBM mean snow accumulations. Even the 0.5″ snow line may be overstated. (Click on image for a larger view.)
This forecast is complicated by the following factors:
Ground temperatures will not be near freezing
Accumulations mostly on grassy surfaces near the city.
Total accumulation is light ranging from 0.09″ water equivalent to 0.15″ water equivalent (AI models)
Frozen wet mix accumulations are difficult to get correct and are dynamic with packing, compression and melting; once the changeover to light rain occurs, accumulations will significantly reduce.
The snow and snow/sleet mix starts just before daybreak and ends around noontime.
Originally Posted Sun 6:06 PM —High pressure moves off to our south on Monday as a disturbance to our west moves to our north on Tuesday.
GFS forecast for 11 AM Monday. Clouds move in from the west. (Click on image for a larger view.)
Precipitation associated with this disturbance will bring some light snow mixed with rain or sleet before daybreak Tuesday.
Today’s 18z RRFS forecast for 6 AM Tuesday. White line is the 32º line at the surface. North of the magenta (700-850 mb) and/or yellow (850-1000 mb) lines, the atmosphere is cold enough for sleet and/or freezing rain. We will be north of the yellow thickness line, Much of what falls will be melting, but may freeze as it falls through this lowest freezing layer. (Click on image for a larger view.)
Temperatures will be above freezing for much of the immediate PHL area. Total water equivalent is less than 0.15″ and many models show less than 0.07″ water equivalent.
The Model Blend (the NBM) did well with the last snowfall, so I’m leaning on it for this ‘snowfall’, if we can even call it that.
19z NBM forecast snow accumulation (mean accumulation) (Click on image for a larger view.)
The NBM provides percentiles for snow accumulation. The above graphic is the mean calculated on 30-40 models.
Below is the 75 percentile accumulation, meaning 75% of the 30-40 models forecast the amount shown or less. (It does NOT mean that 75% of the models forecast this amount.) Thinking about this, it also means that only 25% of the models forecast this amount or more.
19z NBM 75 percentile snow totals. Likely too high. (Click on image for a larger view.)
While the last snowfall actually was correctly captured by the 75 percentile forecast, this is system is very different. Even the lower ‘mean’ forecast may be too high.
Either way, calling this a ‘snowfall’ in our area is stretching it.
