It’s been a long time since things seemed so dry outside. And the current extended range forecasts keep thing dry for us over the next two weeks!
Yes, there will be two cold fronts moving through this week. The first one will result in a familiar pattern of having any rain shear off to the far north and west of the city—
Today’s ECMWF shows showers only far north of the city. (Click on image for a larger view.)
Another cold front moves through Saturday. The models had been forecasting some light showers for us, but those have all but disappeared in the forecast—
Today’s GFS forecast for Saturday. Grey black is cloud cover. Essentially no rain with this front. (Click on image for a larger view.)
I’m still playing with the highly experimental AI-Machine Learning version of the GFS. It still seems to show the first chance of any substantial rain here is November 3rd!
Graphcast GFS (AI Machine Learning model ) forecast for November 3rd. (Click on image for a larger view.)
There’s almost no reason for a weather forecast for this weekend and certainly not a reason for many updates. You’ve heard the details on the radio or TV.
Sprawling dominant high pressure and a persistent upper air ridge will give us sunny skies and increasingly warm temperatures. I don’t see any rain here in the Philadelphia area until the end of October, possibly on Halloween, according to the NAEFS and into November according to the GFS AI model.
During this downtime, I’ve been exploring programming downloads and post processing of some the the experimental AI models.
Regular readers here have seen that the ECMWF-AIFS was quite impressive with its forecast of Helene and Milton.
NOAA also has what they call an experimental “machine learning model” which is a combined effort with Google (Graphcast) and uses the weather database of the ECMWF. It compares current conditions to a weather database going back to the late 70s with several parameters at 13 vertical atmospheric levels to create a forecast.
Over the past week or so, I’ve been working on programming the download and post processing of what’s called the Experimental Graphcast-GFS model. (Weather model downloads are much more complex than simply “click here” for the model.)
Here’s its forecast for October 31st at 8 AM, showing the first area of rain on the map for awhile. Notice that the rain is blocked to our west by high pressure.
10-18-24 00z Experimental Graphcast GFS for Halloween morning. High pressure still blocking rainfall. But it suggests a pattern change within several days of this date. (Click on image for a larger view.)
