Thu 8:49 AM —Forecast Review — The ECMWF-AI-Single model continues to be impressive. Its forecast for some sunshine through a considerable high layer of clouds has so far verified, in stark contrast to the GFS and NAM-NEST forecasting a dark low cloud day, even with drizzle. After years of ‘doing the weather’, this AI model appears to be the most impressive advance in forecasting I’ve seen.
(Gee, I wonder whether the TV people providing the “exclusive” forecast will acknowledge the blown forecast for today? They often don’t, or they find some area within a 100 mile radius that actually had the dark low clouds forecast.)
Hey, it may be too soon to celebrate, and some low clouds might still roll in from the east. That said, the latest ECMWF-AI just available has clouds thinning and dissipating for even more sunshine this afternoon.
Latest ECMWF-AI-Single at 2 PM—
ECMWF-AI-SINGLE Cloud and Temperature Forecast for 2 PM Thursday. (Click on image for a larger view.)
If this new AI model continues to prove itself, I can reduce my data costs and eliminate many more models I download daily. (I’ve already eliminated downloading the HIRESW-FV3, the HIRESW-ARW, the HIRESW-ARW-MEM2, the RAP, and the SREF. )
Yet Another AI Model ‘Experiment’ Continued
Posted Wednesday 03/12/25 @ 10:52 PM —Tonight’s models are even more emphatic about low clouds Thursday morning into the afternoon. Tonight’s NAM-NEST even shows some drizzle. As I posted earlier, the AI model shows sun through thin high clouds. We’ll see what happens.
Posted Wednesday 03/12/25 @ 8:07 PM — We had a sunny day today (Wednesday), very close to the AI model forecast.
I’m still seeing many of today’s models forecasting low clouds for Thursday, but today’s ECMWF-AI-Single model runs have only high cloudiness for Thursday, albeit with a thicker high cloud layer than forecast yesterday.
The latest ECMWF-AI-Single model just became available and it keeps its sunshine through high clouds forecast intact—
ECMWF-AI-Single cloud forecast for noon Thursday. (Click on image for a larger view.)
Let’s see how it plays out tomorrow!
Yet Another AI Model ‘Experiment’
Posted Tuesday 03/11/25 @ 5:46 PM —The latest ECMWF-AI mode (single) and the latest GFS have very different forecasts for Wednesday and Thursday regarding cloud cover. An easterly flow behind a back door cold front may result in very cloud conditions.
Here’s the latest GFS cloud cover for Thursday noon—
18z GFS forecast for Thursday noon. Very cloudy and damp. (Click on image for a larger view.)
Here’s the ECMWF-AI-Single forecast for Thursday noon—
12z ECMWF-AI Single has little to no low clouds. Plenty of sun through high thin clouds. (Click on image for a larger view.)
The two forecasts couldn’t be more different regarding cloud cover. Ordinarily, with these forecasts, I would go with the GFS model. (The GFS relative humidity fields support a very dark cloudy mid-day. The NAM-NEST also is forecasting low clouds. )
The ECMWF-AI-Single model has really impressed me regarding temperature forecasts. It’s also done well with rainfall. Let’s see how Thursday’s cloud cover turns out. (Or whether the models come together.)
Posted Monday 03/10/25 @ 5:07 PM —A large ridge of high pressure in the central US with a split jet flow moving around this central high will continue to bring above average temperatures to our area.
Current (Monday afternoon) 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. Warm upper air ridge in center of the US. Split jef flow. The disturbance to bring rain late Saturday isn’t visible yet in this image. (Click on image for a larger view.)
Dry weather is expected to continue until late Saturday afternoon or evening.
Previously Posted Sun @ 5:03 PM — —The big story (and actually the only story) for the weather this coming week will be the warmer temperatures expected. Additionally, winds and wind gusts will be considerably lower than recent days.
For Monday, here’s the latest model blend (NBM). What’s interesting is how close the the ECMWF-AI model is to the NBM model with less computational effort.
18z NBM forecast high temperatures for Monday (Click on image for a larger view.)
By Tuesday, warmer temperatures push in from the southwest. We’ll be in the 67º-69º —
ECMWF-AI Single model temperatures at 2 PM Tuesday. (Click on image for a larger view.)
It’s been well-advertised about the 70º temperatures for this coming Saturday. The latest AI model suggests we may not reach that.
The change to Daylight Saving Time always takes some adjustment. For me, the biggest adjustment is the one hour later availability of weather model data. One hour doesn’t sound like much, but it brings models out a bit too late in the morning to be as useful and too late in the evening to make the 11 PM TV weather. I’ve written about this many times over past years—
Posted Friday 01/31/25 @ 9:32 AM — A slight shift in the timing of the heavier rain today— late afternoon through this evening, ending after midnight. Total rainfall in the city and surrounding areas is now forecast to be less, closer to 0.30″-0.40″. Heavier rainfall in South Jersey.
NBM 12z Accumulated Rainfall for Friday into early Saturday. (Click on image for a larger view.)
The potential mixed precipitation storm I prematurely mentioned last evening has changed timing, from possibly next Saturday to possibly next Thursday. That’s a big change, suggesting the models, including the AI models, don’t yet have a good handle on things. Nonetheless, the setup ingredients appear to be there. Stay tuned.
Friday’s Rain Update
Posted Thursday 01/30/25 @ 6:01 PM — The rainy forecast for Friday hasn’t changed appreciably and rainfall totals are similar to the ECMWF forecast graphic posted yesterday. Rain begins about 3 AM Friday, earlier south and west and continues through about 9 PM Friday evening. It will be windy and gusty at times during the day, especially late afternoon when the heaviest rain is expected to fall.
The weekend looks as follows: Sunny on Saturday and cloudy on Sunday as a disturbance passes to our north.
No major storms until NEXT weekend, when things may get interesting, according to the AI models and the ECMWF—
The ECMWF forecast for Superbowl Weekend (NEXT Saturday) Very far off in the future, but the setup is for a mixed bag of very heavy precipitation here. (Click on image for a larger view.)
Friday’s Rain
Posted Wednesday 01/29/25 @ 7:42 PM — A storm bringing only rain to our area will move in several hours before daybreak on Friday. It will be rainy all day and the rain will taper off during the evening.
Continuing our ongoing trend towards lower than average rainfall, we’ll only see 0.50″ of rain in the immediate Philadelphia area according to the AI model. Areas not far to our south will receive over an inch, according to the latest models.
Staying with the AI model, here’s its forecast total rainfall for Friday—
Today’s 12z ECMWF-AIFS total accumulated rainfall by Saturday morning. (Click on image for a larger view.)
The NBM is forecasting heavier rainfall for Friday—
Today’s 18z NBM rainfall forecast (Click on image for a larger view.)
AI Model Looks Quite Good in Simple Experiment
Posted Wednesday 01/29/25 @ 5:47 PM — The temperature reached at Philadelphia airport today was 52º, 53º based on hourly records. ( If 53º was reached, it occurred off-hour and will be posted on the NWS site in a few hours.)
It’s very impressive that the AI model correctly forecast this range a week ago. Today’s NBM forecast was a still a degree so too cool.
Posted Wednesday 01/29/25 @ 9:39 AM —Over the past week, I’ve been conducting a simple observational experiment regarding today’s high temperature forecast. (Also see here.)
I chose to follow the AI model temperature forecast for today, because it was below freezing here at the time and the majority of the models, including my go-to model for temperatures (the NBM) were forecasting highs in the mid 40s for today, while the ECMWF-AIFS at that time was forecasting mid 50s! Quite a difference.
Incredibly, the regular model forecast have creeped upward towards the 50s and closer to the AI model while the ECMWF has shave off a few degrees to a high of near 53º.
Here’s the latest NBM —
Today’s NBM (12z) high temperature forecast for today. (Click on image for a larger view.)
Here’s the latest AI (ECMWF-AIFS) forecast for today—
Today’s ECMWF-AIFS high temperature forecast for today. (It should be noted that the ECMWF forecasts in 6 hour increments and the version I download does not include the day’s maximum temp. The only data is the actual temp at 1 PM as shown above. Since high temperatures typically occur around 3 PM, the actual max temp for today may be somewhat higher
What’s impressive is that the ECMWF-AIFS was forecasting this range of max temperatures almost a week ago! It took the computational models almost a week to approach its forecast.
Posted Tuesday 01/28/25 @ 10:53 AM — Radar shows some snow showers to our far northwest. Superimposed RAP model vertical motion (“omega”) shows strong downward motion, likely dissipating these snow showers greatly by the time they get to Philadelphia—
Radar shows some snow showers to our northwest. Superimposed RAP model vertical motion (“omega”- yellow dashed lines are negative) shows strong downward motion, likely dissipating these snow showers greatly by the time they get to Philadelphia (Click on image for a larger view.)
Posted Tuesday 01/28/25 @ 9:17 AM — When I went to bed last night, the forecast for Philadelphia on our news radio station called for “‘”snow showers, possibly reducing visibility” and clouds. Did I miss something with my update yesterday? What were they talking about?
We’ll have a mix of sun and clouds today. Several weak disturbances are in the pipeline for us over the next few days. Snow/rain showers are a slight possibility.
WV image with superimposed RAP model parameters. Disturbances 1-3 are in the pipeline to affect us. The upper level low (4) will bring rain on Friday. (Click on image for a larger view.)
There’s a chance that the storm on Friday may start as a bit of freezing rain or sleet here before a change to all rain. Not a certainty.
Posted Monday 01/27/25 @ 5:55 PM — A slight warmup for Tuesday and Wednesday. The ECMWF-AI model has us in the low 50s on Wednesday. The model blend (NBM), my go-to model for temperatures, now has a forecast high of 48º in Philadelphia with a rather large standard deviation of ± 2.8º.
A series of weak fronts and upper air disturbances come through over the next few days, specifically before daybreak Tuesday and Tuesday night. A rain maker is on the way for Friday, although the heaviest rain may move to our north and south.
The front and the system bringing rain on Friday are visible on water vapor imagery—
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. The closed low in the southwestern US will move into our area on Friday. (Click on image for a larger view.)
A note for those of you who are visiting this site from “X“. Due to the highly unbalanced political environment on X, I’ve considerably cut back auto-posting there. To receive all forecast postings, I recommend following me on mastodon.social or bluesky.social
Originally Posted Mon 10:45 AM —A split jet flow and a still somewhat chilly northwesterly flow will keep temperatures slightly above average through Wednesday, becoming somewhat colder Thursday, then warming again Friday with rain.
Today’s 06zNAEFS jet stream wind speed forecast for Wednesday morning. The upper closed Low (L) will eject towards us bringing rain by Friday. (Click on image for a larger view.)
Regular readers here know that over the past week, I’ve been watching the AI version of the ECMWF which has consistently forecast high temperatures on Wednesday to be in the low 50s! Every other model has us in the mid to upper 40s. So this is a little experiment to see if the AI model has any special insights into high temperatures—
Today’s 06z ECMWF-AIFS forecast temperatures on Wednesday. Much higher, and consistently so. then the regular models. (Click on image for a larger view.)
It appears we may have some rain on Friday into early Saturday. This feature has been in the forecast for several past days, with each model run moving it earlier. (It was originally forecast for this rain to move in Saturday into Sunday.)
Today’s 06z NAEFS forecast for Friday at 1 PM. The white contour is 32º, the red, yellow and magenta lines are critical thickness lines, indicating the average temperature of different lower levels of the atmosphere. With all of these lines to our north, we will have rain, not snow. (Click on image for a larger view.)
The weekend looks to be dry.
No significant snow storms are on the horizon for us. For that matter, we’re still running a rain/snow deficit and the dry conditions of last Fall have not shown signs of ameliorating to this point.
