Category Archives: Commentary


This Sunday morning,  Eastern Daylight Saving Time changes to Eastern Standard Time — the early morning hours get lighter (at least initially), the afternoons gets darker earlier, and the daylight hours that are already shortened seem even shorter.

What could possibly be good about this?

Well, there’s at least one good thing —  the weather models (which are run on Universal Time”UTC” (previously called Greenwich Mean Time “GMT”) are available an hour earlier! On the east coast, this makes a giant difference.

Every evening, the National Center for Environmental Prediction begins incorporating the latest weather data measurements including weather balloon radiosonde data at 00 UTC, which is 7 PM Eastern Standard Time and 8 PM Daylight Saving Time.

The incredible amount of data to ingest and the number of calculations needed for models to output takes 2 – 4 hours to complete, even on super computers.

Since last March (the beginning of DST), every meteorologist, TV station, etc. on the east coast has had to make their evening  and 11 o’clock broadcasts without the latest GFS model data, simply because the first 24 hours of GFS forecast data hadn’t become available until 11:38 PM EDT — after the 11 o’clock news was over.

With the change to Eastern Standard Time, the GFS now becomes available at 10:38 PM EST— plenty of time to update the weather on the 11 o’clock news.

Similarly, the NAM whose 24 hour forecast data became available about 9:57 PM EDT will now become available at 8:57 PM EST.

For those interested, the exact model time availability can be found here.

For amateurs who enjoy weather forecasting and for those who do it professionally, this one hour difference makes all the difference in the world.

It’s a small conciliation for the shorter days and the colder weather, but it’s something. 🙂

Over the years, I have talked about this on previous blog posts—here,  and here.

[su_note note_color=”#d9f2da”]BTW, the GFS is showing a frigid outbreak for the time period around Nov 12th.[/su_note]


I find it fascinating that most of today’s models didn’t predict the rain that developed in a narrow, persistent band over Philadelphia. Interestingly, little rain developed at the shore to this point.

Radar 4:30 PM Friday, courtesy of

Going back to last night’s models, the WRF-NMMB and to a lesser extent the WRF-ARW were the only models to predict this rain last night.   However, the same models run this morning didn’t show it as much and the GFS, NAM really were off about this.

The unexpected rain was the result of some vorticity and vertical lift in the mid levels of the atmosphere, captured on this afternoon’s 2PM RAP model analysis—

RAP Model 2 PM Friday Analysis showing lift and vorticity at the 500 mb level (~18,000 feet)

I don’t know why, but as I mentioned last week, hurricanes seem to throw off the model forecasts, often in big ways.

This rain is expected to clear out later today.  The weekend looks to be beautiful with sunny skies and comfortable temperatures in the 70s.  I’ll do a brief update later this evening.



So what happened to the storms and flash flooding expected for today?

Clearly, the models didn’t do a very good job here.  On Sunday, the models overwhelming predicted a severe weather event.  They continued to do this on Monday, but by Monday evening there was a lack of agreement, with some models showing the insignificant showers we ended up with.

By this morning, the severe weather parameters had become very unimpressive and I posted that change this morning.  Still, the Rapid Refresh model (RAP) available at 9:35 AM showed a line of storms about 4 PM.

While at work, the rest of this morning’s models became available between 10:15 and 12:40.   The NAM NEST, WRF-AWF, WRF-NMMB, RAP, HRRR and HRDPS are the models I look at for thunderstorms. (Hey, it’s almost a full-time job, but it’s still a hobby.)

By this morning’s model run, the models had really backed off considerably with even the rain, shunting anything developing to our south.   Too late to make an updated web announcement, while at work.

Interestingly, they were still talking about severe thunderstorms on the radio while driving home at 6:30 PM.

It’s always a tough call to cancel the call for severe weather when even the slight possibility could endanger people if it occurs.

Anyhow, the “elusive” search for the model that’s always correct is elusive for a reason.

Looking back, even the GFS, a large-scale model, did better yesterday than some of the high resolution models.  And if I had to hang my hat on a model yesterday, the Canadian HRDPS probably called it the best at the earliest time.  But it’s not always right…


Today, with its morning model run, the National Center for Environmental Prediction (NCEP) will have upgraded the GFS model (version 14) to the new and improved FV3-GFS (GFS version 15.1.1). The FV3-GFS becomes the main global model of the National Weather Service today.

The FV3-GFS (Finite Volume Cubed Sphere) has been in development for well over 10 years and has been in testing for the past three. It uses a different three dimensional geometry to reduce errors inherent in all numeric weather models. It includes different modeling physics and new parameters.   Significant information on the FV3-GFS can be found here.  Background information is here.

The model data has been available on university web sites in a very limited fashion for awhile.  Over the past two weeks, the NCEP made the data available on the main weather model download server, “NOMADS”.

The FV3-GFS has been heavily evaluated and is considered “equal or better” than the current GFS.


My post from last year about Daylight Saving Time and the forecasts you see on the 10 and 11 PM TV news is useful information to check out.

The short version is that the latest runs of the GFS and other global models are not completed nor available in time for the late night TV weather segments when we switch to Daylight Saving Time on the east coast.

Indeed, the first “products” of the GFS model first become available about 11:32 EDT.

With the exception of the NAM, the forecasts you see on the 10PM or 11PM (east coast)  news/weather can only be based on older model runs from the short range models and [possibly] the newest NAM.   Indeed, some of the short range models at those times are, in turn,  “initialized” from the earlier afternoon NAM model, not necessarily the latest!