iiinova — where your digital self keeps evolving

K Griffin @kgriffin.dev · 25/09/2026

In a weird twist of storm locations, NHC/CPHC are issuing advisories whose public time stamps span -five- time zones. And none of those are EDT nor CDT. Thanks, +ENSO? Fay - GMT Gonzalo - CVT Polo - CST (~= MDT) Odalys - PDT Nolo - HST

060

K Griffin @kgriffin.dev · 25/08/2026

one of the busier EPAC/CPAC maps you'll see...

0309

K Griffin @kgriffin.dev · 10/08/2026

This really needs to be framed from the perspective of what changed (that is here, what was removed), not the fraction of basic societal decency that they kept in place.

081

K Griffin @kgriffin.dev · 10/08/2026

The award for terrible weather UI design goes to this prompt from Gemini. 69°C|°F, whatever that means.

060

K Griffin @kgriffin.dev · 03/08/2026

State Line and the Union, hard to beat that day!

010

K Griffin @kgriffin.dev · 03/08/2026

I hear some of y'all are gonna be hanging out with this cool cat later this week.

060

K Griffin @kgriffin.dev · 28/07/2026

goes live. An ideal world has a period of time where old and new models run in parallel - or, with these quick-to-run AI models, even a short backcast is doable. Whatever can be done is quite helpful and insightful and lets users prepare for the changes, both qualitatively and quantitatively.

010

K Griffin @kgriffin.dev · 28/07/2026

Fair question! It's not uncommon to have systems built up around the general biases and recent errors of a forecast model for particular places. If the underlying details of how that model performs changes, it's good to reset the stats if you can't build up a set of new stats before the change 1/2

210

K Griffin @kgriffin.dev · 28/07/2026

For real, you can't be making "small" changes (like the entire loss function??) to your AI models and only announce it after the change is live. That's basically a whole different model!

130

K Griffin @kgriffin.dev · 28/07/2026

The number of posthumous SCNs from NOAA lately is too damn high.

192

K Griffin @kgriffin.dev · 06/06/2026

The B-52 flyover is always worth the wait. #USAvGER #USMNT

010

K Griffin @kgriffin.dev · 06/06/2026

This Magic School Bus publication from the late 90s leaves a lot to be desired. Or maybe I just don't know my, ahem, "Fujito" scale 🫠

010

K Griffin @kgriffin.dev · 01/06/2026

Sometimes "no" is hard to define, especially for large and powerful organizations.

011

Reposted by K Griffin

Katharine Hayhoe @katharinehayhoe.com · 26/05/2026

Cannot 👏 emphasize 👏 this 👏 enough

12561113

K Griffin @kgriffin.dev · 25/05/2026

If you didn't know any better, you might look at this corner of the ORD departure queue and think that Fort Wayne and Madison must be very important, busy destinations.

110

K Griffin @kgriffin.dev · 31/03/2026

These are insane statistics. For NSF Geosciences, there have been just 74 grants awarded in the ~6mo of FY26 so far, 19.6% of FY21-24 avg 377. The obligations ($) are larger per grant, but the comparison even to FY25 is even more striking. With only $75M awarded, it's long way to the floor of $1.2B.

0146

K Griffin @kgriffin.dev · 24/01/2026

Extreme Cold Warning for lows of 21F. Don't mind me being out there in short sleeves on Monday and Tuesday... (I know, I know...)

040

K Griffin @kgriffin.dev · 15/01/2026

My secondhand excitement for your Racing has become very firsthand.

110

K Griffin @kgriffin.dev · 10/01/2026

Welcome back to The Magic Of The Cup™ Season.

010

K Griffin @kgriffin.dev · 21/12/2025

The solstice isn't even for another 90 minutes....no.

020

K Griffin @kgriffin.dev · 26/11/2025

Some of the coolest science I've heard in a while. Pretty solid evidence given the lack of ability to directly measure the environment on Mars. www.npr.org/2025/11/26/n...

npr.org

At long last, the mystery of lightning on Mars is solved

A chance discovery by a NASA rover on Mars shows that the red planet has a form of lightning, which researchers had suspected for decades but never seen.

000

K Griffin @kgriffin.dev · 22/11/2025

And big bonus for the WET conditions.

010

K Griffin @kgriffin.dev · 22/11/2025

The dozens (?) of NVDOT traffic cameras along the strip are a fun way to watch F1 qualifying without worrying about your ESPN subscription - all have real-time video feeds. www.nvroads.com

131

K Griffin @kgriffin.dev · 20/11/2025

That's never a good way to start your day.

083

K Griffin @kgriffin.dev · 14/11/2025

One little thing is that the MBP has 3 USB-C ports + HDMI, vs the MBA with only 2. But, these are the two models that just made their way into the laps of each of the two scientists in my house in the last few months, so neither will be a bad choice.

000

K Griffin @kgriffin.dev · 12/11/2025

Another sign of an intense event - when SWPC goes _hard_ down. Reminds of the NHC site during big events way back in the day.

140

K Griffin @kgriffin.dev · 12/11/2025

I already got some great clear-sky shots earlier this evening before the clouds came back in a second time, but views like this of the current substorm are just a tease

0130

K Griffin @kgriffin.dev · 31/10/2025

This photo from Sept 2022 was surfaced to me and, uhm, I think we're going to need more of this beverage all too soon...

050

K Griffin @kgriffin.dev · 28/10/2025

Nauseatingly intense winds. That ML150 value is likely what supported the 10a ET bump to 160kt. Remember dropsonde winds are the equivalent of gusts, but 172kt gust at the surface is beyond insane.

070

K Griffin @kgriffin.dev · 28/10/2025

Finally a dropsonde message got out. 892mb. Lightning continues across much of the western semicircle.

030

K Griffin @kgriffin.dev · 28/10/2025

Must've been some mess that got made on the way through the eyewall to have this long of a hold inside the eye. Typically it's either a cleanup or, on occasion, an equipment check. The up-and-down blue lines and low red line values in the top left chart indicates the time spent circling in the eye.

051

K Griffin @kgriffin.dev · 28/10/2025

The satellite images remain unparalleled - yet also utterly sickening when you see the outline of Jamaica just a half degree of latitude (so, ~30 miles) north of the northern edge of the pink truck tire.

060

K Griffin @kgriffin.dev · 27/10/2025

Lightning only now beginning to encompass the entire eyewall of Melissa, or at least an uptick in lightning counts if you don't fully trust the GOES lightning mappers.

020

K Griffin @kgriffin.dev · 27/10/2025

I've seen a lot of dropsondes over my years, and a peak wind of 210kt - even at the top of/just beyond the boundary layer - is wild to see. Might be up there in the ranks of all-time strongest dropsonde-measured gusts as well.

49821

K Griffin @kgriffin.dev · 19/09/2025

The chargers on the AF 223 fleet has a leg up on the KLM 195s. Also makes sense given that AF bans all in-flight use of power banks.

030

K Griffin @kgriffin.dev · 14/09/2025

Maybe it's a bit hyperbolic

000

K Griffin @kgriffin.dev · 14/09/2025

It might seem modern to only have the "universal" plug but these USB-A ports went out of style 5-10 years ago for charging. I don't even have a cable worthy of traveling that is compatible.

250

K Griffin @kgriffin.dev · 04/09/2025

This is one of the saddest public transit headlines I've ever seen. Major sporting event? Either drive home from the massive parking lot, or take a train that can only run because of a sports betting company. The SEPTA funding situation is so beyond sad at this point.

050

Reposted by K Griffin

Timmy Cagg @axilla2.bsky.social · 01/09/2025

Inter Miami is disgusting. Luis Suarez is disgusting. Spitting on a Sounders coach. #LeaguesCup #MLS

3631183

Reposted by K Griffin

Nahel Belgherze @wxnb.bsky.social · 16/08/2025

Unbelievable footage from NOAA’s Hurricane Hunters inside the eye of Hurricane Erin. Simply incredible.

632715920

K Griffin @kgriffin.dev · 16/08/2025

I saw it at 8-1 as well and thought oh well, the burger motivation was enough. Then turned on the game at 8-8 just two innings later. Just magic. 👼

110

K Griffin @kgriffin.dev · 14/08/2025

I know which two centers split on that latitude, lol

020

K Griffin @kgriffin.dev · 11/08/2025

Ahhh! (replies crossed in the dark) That at least makes more sense, even though I'd suspect most gauge sites would have issues with under-reporting than over, but I'll admit that I don't know enough about microscale wind patterns - and in the end, wrong is wrong.

000

K Griffin @kgriffin.dev · 11/08/2025

I know the staffing had been thin at periods - having left AOS in 2016, all I knew is that it sat on 13 and still (seemingly) had active staff. Given that one of the prospective record-breaking measurements came from AOS faculty on that same floor, was kinda confused by that number!

000