November 2023 - General Weather Discussion

[LakeEffect]

Nothing much to report for the last two days of October other than sun, wind, and cold air... November here we come!

[okatty]

But 70s for the weekend? I’ll take it if that comes to pass!

[Bunty]

But 70s for the weekend? I’ll take it if that comes to pass!

70s for the last Bedlam would be really cool! Meanwhile, it's so cold that today the highs will only be up to around the average lows.

[BG918]

Quiet week ahead with high temps back in the 80’s through Wednesday then a cold front and chance of rain Thursday. Most models right now are showing slight rain chances across northern and central OK and better chances SE. Cooler more fall-like temps in the 60’s for the weekend. Next good chance of rain across the state is about 2 weeks away

[Anonymous.]

Rain chances for OKC are basically gone for Thursday. However, the cool-down will still occur to drop us to basically our average temps for this time of year.

[BG918]

Yep this system is diving further south so just SE OK should see rain Thursday. Another quiet couple weeks. GFS rainfall through the weekend:


Still a ways out but there are model signals for a strong storm system to affect the area the weekend of 11/18-19. Will look at potential rainfall totals early next week.

[BG918]

Another dry week with high temps in the upper 60’s and lows in the 40’s - very typical Oklahoma fall weather. The next chance of rain is Sunday 11/19 and again Tuesday 11/21. Thanksgiving looks dry.

[Anonymous.]

Slight chance of rain coming Sunday into Monday. But otherwise dry, fall weather continues. Looks like a stronger cool-down is coming around Thanksgiving timeframe.

[kukblue1]

Anyone else notice the SPC put I-35 east under a 15% slight risk area for Sunday. I haven't looked into yet might tonight.

[Anonymous.]

We may see it get backed up further W into C OK.

[fortpatches]

For all the weather nerds here, just came across this blog post and figured yall would be interested.

In the study, GraphCast demonstrated superior performance over the world's leading conventional system, operated by the European Centre for Medium-range Weather Forecasts (ECMWF). In a comprehensive evaluation, GraphCast outperformed ECMWF's system in 90 percent of 1,380 metrics, including temperature, pressure, wind speed and direction, and humidity at various atmospheric levels.

GraphCast utilizes what researchers call a "graph neural network" machine-learning architecture, trained on over four decades of ECMWF's historical weather data. It processes the current and six-hour-old global atmospheric states, generating a 10-day forecast in about a minute on a Google TPU v4 cloud computer.

Remi Lam et al., Learning skillful medium-range global weather forecasting.
Science 0, eadi2336 DOI: 10.1126/science.adi2336



Learning skillful medium-range global weather forecasting | Science

[jedicurt]

For all the weather nerds here, just came across this blog post and figured yall would be interested.


Remi Lam et al., Learning skillful medium-range global weather forecasting.
Science 0, eadi2336 DOI: 10.1126/science.adi2336



Learning skillful medium-range global weather forecasting | Science

one of my tech podcasts was covering this. but they did say the limitation was in localized forecasting, and so while this might become a tool for much quicker regional modeling, it's certainly a tool and not the sure fire end all for how we will get weather forecasting in the future.

[kukblue1]

We may see it get backed up further W into C OK.

IS that for 8am? That be crazy

[fortpatches]

one of my tech podcasts was covering this. but they did say the limitation was in localized forecasting, and so while this might become a tool for much quicker regional modeling, it's certainly a tool and not the sure fire end all for how we will get weather forecasting in the future.

I apologize, I don't know much about weather forecasting, but what is meant by "localized forecasting"?

In the paper above, they state "A single weather state is represented by a 0.25° latitude/longitude grid (721 × 1440), which corresponds to roughly 28 × 28 km resolution at the equator (Fig. 1A), where each grid point represents a set of surface and atmospheric variables."

And they call "GraphCast" a method for "global medium-range weather forecasting".

It is localized because if breaks down the globe into just over 1mil segments?

[BG918]

EURO forecasted rainfall through Tuesday. Severe weather possible across central and NE OK Sunday afternoon

[Anonymous.]

Rain associated with this system is basically done for C OK.

Morning lows will be noticeably colder beginning Wednesday. Highs in 50s, lows near freezing. Another shot of cold air comes in Friday into Saturday, could see some cold rain and perhaps some light snow mixed in over the weekend, but does not look significant at this time.

[Timshel]

Got a nice little 5 minute shot of marble+ sized hail in Nichols Hills late last night. Looked like a tiny cell, but was surprised to say the least.

[Anonymous.]

A couple of the forecast models are trying to make the weekend a little more interesting with some wet snow. Right now, both the GEM and CAN are showing a cold rain changing to snow Saturday evening across the NW half of the state. However, nearly every other model is forecasting snow only for far NW OK. Will keep monitoring trends.

[BG918]

After a cool day today with wind chills in the 30's and another freeze tonight south winds will bring high temps near 60 through Thursday. Another system moves through Oklahoma Thursday night into Friday morning bringing widespread rain mainly east of I-35. Friday will be cloudy and chilly with north winds. The weekend looks decent with temps back close to 60 and sunshine. GFS forecasted rainfall totals through Friday:


After tonight's freeze it doesn't look like any really cold air will be coming toward Oklahoma for the next 10 days. Nighttime lows should also stay above freezing.

[Anonymous.]

Thursday afternoon into evening: