weatherwiz

Kind of explained it in a post above but I'll add more. We should be focusing on the resources needed to better what we currently have instead of just adding more tools to the toolbox. Now, if AI will be used to better the initialization/parameterization process, that would be amazing. But at the end of the day we still need better computing technology (which we have, we just need it within the field). Quantum computing is going to go a sizable way I think here. Parameterize better and improve initialization, those two alone will go a great way in forecast model accuracy and hopefully reduce inconsistency. If AI models are just going to add to the list of potential outcomes and increase uncertainty, then what good is it?

Bingo. That's exactly what I could see happening. People in charge of making critical decisions will take AI at face value and that will set the stage for some disastrous decision making. And I'm not even talking in the sense of general public - this could be anything, logistics, supply chain, resource distribution, etc

As much as I don't like AI, that doesn't mean I don't believe there is value in it, however, I don't think its value outweighs the "bad" and that's where I see the problem. In terms of weather, I think AI could have tremendous value in the nowcasting (<6-12) hour window, particularly when it comes to severe weather and flash flooding potential. In terms though of using AI as like medium/extended range (maybe even short range), I see little value - AI models in this range will not tell us what a great forecaster will not already know. The only way we will ever greatly improve forecasting skill in this range is to better understand how the atmosphere and how it evolves and better understanding physics and processes...then it's taken that and quantifying that numerically so computers can process this information. Computer forecast models struggle as you move away from initialization because of compounding error. This was why for a prolonged period of time the euro was by far and away the best, it had superior initialization skill - the euro would have very few error inside of 72 hours which resulted in greater accuracy through 5 to even 7 days. AI isn't going to solve medium-to-long range error or reduce inconsistencies...AI will not do this until we ourselves are better able to understand this and teach the algorithm to do this.

His question was valid given my open disdain about AI weather models lol

Those float around Facebook too...and there's this huge circulation too of videos of like destructive weather or chain reaction accidents. I guess at least there is wording that says, "This is AI generated and only for your entertainment" but it's stupid and dangerous.

AI in general. It's scary what can be done and so quickly. A friend of mine and I took a picture together at a Wolf Pack game on opening night and one of our friends made some AI video of it in two seconds and it was not something rather appropriate...and it looked very real.

This is getting out of hand and becoming dangerous very quickly.

I mostly use social media nowadays for sports news/rumors but even those are a joke. There is this stupid Boston page that does does for Bruins/Red Sox and the authors are a bunch of click bait tools with their ridiculous titles. I wish it could be blocked from my google searches for news/rumors. I'd also like to tell them off.

The other great part of blogging and posting your reasoning for forecasts is you have something to look back on. This is why I wish I was more active with blogging but really dropped off over the last several years. So I try to rely somewhat on memory which isn't always the best metric. But when I'm forecasting I will try to think of not necessarily similar events but I try and visualize what the response will be to certain things. One thing I learned from Ekster way back to the eastern days was (and this is about thunderstorms) to try and visualize how a thunderstorm will respond to the environment it is in. But I try and extrapolate this concept across weather phenomena. With winter storms, I always try and paint a picture in my head of how the radar presentation will look at how it will evolve based on how the atmosphere is evolving. This concept has really helped me produce some solid snowfall forecasts over the years, but (in the cast of this past storm), if I am underplaying/overplaying certain aspects I will be way off.

That's precisely the reason I love blogging too. It help keeps my sharp and I kind of use it as a way of talking to myself to ensure I am understanding of what's going on meteorologically (under the hood).

Yup. I know every now and then we all debate on would you rather have one big monster storm for the winter and little outside of that or a winter which is composed of many, but small events. There was a time I would have preferred the former, but I think I would prefer the later (I mean the true preference would be many smaller storms but a few big monsters lol). In general, winter and cold suck...but getting snow makes it tolerable. Give me several 2-5" events and winter will be more fun to deal with

Can we just get like 5-6 more systems like this past Saturday, except maybe a few inches beefier?

It's not an excuse...it's just a product of life. Even myself I wish I got to be a bit more invested with local forecasts. But I have to do forecasts for all over the country and its not really forecasting for people so I don't get to do specific forecasts like this (making snowfall maps/forecasts for example). I just do this for fun...but I am so tired after work stuff I barely have the energy to do these maps/forecasts/blog posts. I've been wanting to switch and do videos but I am not a technological person or meant to be heard (why I didn't pursue broadcasting lol). But for my severe weather class this past semester, we had to do a video presentation for our final project and as part of the class I got camtasia which is super easy to use...so I may switch from written blogs to video...might be easier/less time consuming.

I think what you put forth was pretty damn good as what did 4 Seasons and Seymour. I wouldn't even be surprised if some or a chunk of the snow reports were inflated a tad (by like 0.5"), if people measured in the grass when you also account for fluff factor. When I was driving along 84 yesterday, the stretches of median which are flat and grassy...you could easily see grass poking through. It wasn't until like towards Waterbury/Route 8 where that wasn't the case (especially closer to Seymour/Ansonia). But these maps are pretty damn good. There will always be some that end up along the upper end of the range or just over and some that end up along the lower end or just a tad under

Great calls! Ryan had a great forecast for CT too. Edit: Looking back at some posts FXWX had a great call too.

Definitely a D- or an F on my forecast for this. I also try to factor in the backend/reasoning to the forecast into the verification and weigh that much more than the outcome itself. For example, if I did a forecast somewhere of 3-5" and said that would come on the front end of a storm, and it turned out that happened on the back end with a CCB...well that would grade extremely low, but the reasoning was incorrect. The most incorrect aspect of my thinking here was a stronger band developing and impacting Long Island would cut back on totals farther north across CT (subsidence) but that did not happen. While I don't know what everything looked outside Saturday night in terms of snow growth and flake size, it was clear there was just enough lift, combined with a deeper DGZ, to utilize better ratios. I definitely underplayed that component. In terms of assessment, I don't think I would do anything different if I had to do this over. But what I would do is certainly go a bit more "aggressive" (putting aggressive in quotes because we were really dealing with 1.5-2.5") and maybe do a range of 1-3" over a larger area to cover uncertainty better. Or maybe even C-2" but that's a range I hate.

Well I mean they were flying for the Pats...unfortunately so for the Bills lol

And Pats TDs

Guessing snowfall totals outside of the shoreline generally in the 1-2” range…with some locally higher amounts? Definitely more than I anticipated, even up north near me. Up to a coating bad call for inland.

Highway wasn’t terrible but saw two separate accidents on route 8 that looked like spin outs. One was on the southbound side and the other on the north bound side…probably even a few miles from each other. Definitely more snow headed through Seymour and in Ansonia than along 84. Beautiful scenery, everything caked in snow

Meeting a few friend for brunch at Copper City Bar and Grill

Eclipsed 1” here! Have to drive to Ansonia so hopefully highways aren’t bad

If we could reel off like 5-6 events like this in a months stretch we’d be good

Light snow falling. Just measured right around 3/4” so 0.8”

Yup, we went to school together! One of my best friends, great dude