SMART MODEL, WE'LL SEE

By now, I assume that most of you are familiar with the idea of artificial intelligence, a field of study that focuses on creating machines and computers that can perform tasks that usually require human intelligence. AI systems use math and logic to simulate human reasoning, and can actually learn from data to improve accuracy going forward. AI can perform a wide range of tasks, including using existing data to make predictions or take actions. That is why AI driven weather models are being developed to improve weather forecasts. Some are already in use on a test basis.

While the GFS (the U.S. based model) has come a long way, the superior model in terms of skill scores remains the ECMWF, a European forecast model that I call "the EURO". It and the GFS are highly sophisticated data driven numerical models currently used in what's currently known as the IFS (integrated forecast system).

The EURO has long been my go-to model at crunch time, and now it has expanded from its IFS based past to an artificial intelligence integrated forecast system that goes by (AIFS). Unlike older weather models, AIFS uses complex algorithms to analyze past weather patterns and make predictions based on those learned patterns, allowing for potentially more accurate forecasts, especially for complex weather situations. Although they are still produced at relatively coarse resolution, AIFS forecasts already show higher skill than IFS forecasts when measured by a range of standard forecast scores (an example is shown in the image below).

As a man who is always looking for an edge, I like the idea of a cutting-edge AI system developed by ECMWF to improve weather forecasting accuracy by learning patterns from large amounts of historical weather data. That is why I'm playing around with the new EURO AI. Having made forecasts the traditional IFS way my whole life, I'm a little leery of the AI route since It's still a work in progress, but it has shown promise. Sometimes you need to embrace new ideas, like cars over horses, or a TV over a radio. I'm standing on the fence right now, but I've got a pretty good feeling where this is leading me.

SEE OUR NEW WEBSITE & HOLIDAY DEALS AT OUR GALENA AIRBNB

Santa's secret destination to relax after his big night!

Having laid this all out, I wanted to let you in on what the EURO AI is depicting for temperatures the next 15 days in 5 day increments, each getting progressively colder.

November 20-25th (5-days)

November 25-30th (days 5-10)

November 30-December 5th (days 10-15)

The last 10 days of the AI are quite cold, especially the last 5 days to start December. The average daily departure for that final 5-day period is roughly 20 degrees below normal. Again, this is largely based on past weather patterns and historical data going back to 1979.

Here's the traditional day 10-15 day IFS departures on the latest EURO. While cold, they are nowhere near as chilly as what the AI showed for that period above. This will be an interesting case comparison.

I'm not advocating for one solution over the other, but the AI outcome would bring an extremely cold start to December. Without much experience with the AI and its accuracy, I'm being very cautious about the intensity of the cold. It's in the back of my mind, but until I see more consistency with other sources, I will let it set on the back burner. Just wanted you to know some of the insights you get with this site.

That brings us to our current situation, which so far has played out the way it was envisioned. Most of us saw our first snowflakes Wednesday, which is always an exciting event for me. I took this shot of a heavy snow squall around 11:30pm Wednesday night from my deck in Dubuque.

Some additional snow showers will linger into Thursday, with the most likely locations to see them east of the Mississippi, especially in NW Illinois. Considerable cloudiness is expected, but there may be a few breaks for sunshine. The HRRR and 3k NAM show this for snow amounts. We are missing out on a nice snow event that impacts our neighbors just to the east. A tough break for a snow man like me!

The HRRR

The 3k NAM

Thursday will be a very raw, uncomfortable day, with NW winds on the order of 30-45 mph. A few places east of the Mississippi could gust to 50. Wind headlines are likely from the NWS. The HRRR shows these 10 meter gusts.

Highs should remain in the upper 30s to low 40s, and wind chills all day will never get out of the 20s. Some of the coldest will be found around 7:00am when upper teens to low 20s are likely.

Conditions will slowly improve going into the weekend with dry weather back on the table and some modification in temperatures. After highs in the 40s Friday and Saturday, readings Sunday could pop into the low 50s. That will feel good.

Things could get interesting around Thanksgiving or slightly thereafter when another burst of energy rolls out of the west. Where it tracks and how it phases will be important aspects to monitor in coming days. Just for kicks, the EURO AI shows a system to our south, bringing snow Thanksgiving night and the following morning. Something to watch.

Hopefully you found this to be an intelligent, data driven post. If not, I gave it my best shot. Hold on to your hat today and roll weather...TS