counters

10 ms·

91.
by counters 1y ago
> Hurricane risk has been grossly exaggerated for years Year-over-year, economic impacts and disruptions due to tropical cyclones are dramatically rising. Most of this is an exposure issue. But long-tail events - like Andrew's utter
92.
by counters 1y ago
The companies I'm referring to are (generally) not publicly traded, so it's not quite that simple. Is it possible that some sort of backroom shenanigans are going on here? Yeah, absolutely, especially as several knowledgable folks
93.
by counters 1y ago
> To anyone acting as if this is a surprise or they're suddenly caught out and have to switch to another provider, I have to wonder, with the writing on the wall for 8 years now, how have you not already updated your plans? That doe
94.
by counters 1y ago
It's not - it actually the core mechanism through which the "Weather Enterprise" works. Over 20 years, an important report from the National Academies [1] laid out how an enterprise comprised of public, private, and academic
95.
by counters 1y ago
Nothing but 1xL4 are even offered on Cloud Run GPUs, are they?
96.
by counters 1y ago
> This is an opportunity for private industry to step up and step in, while drastically reducing the size of government. Did... you actually read the comment you're replying to? They're explicitly stating that there is a large
97.
by counters 1y ago
> I won't doubt it was fun and educational but I can't in all honesty pretend that is a good value for the taxpayers. The students who work on these types of projects go on to create technology, companies, and jobs. The skills
98.
by counters 2y ago
GDM's GraphCast/GenCast require an "analysis state" of the atmosphere. This is a 3D, gridded dataset with key variables like temperature, humidity, and winds. Generally speaking, an analysis is produced by a physics-base
99.
by counters 2y ago
Not really; it was a gimmick. They used standard forecast post-processing techniques to bias correct global/regional weather models. There is virtually no evidence they actually used device data in this process.
100.
by counters 2y ago
Extremely poorly, because forecasting the weather is all about forecasting the deviations from the expected seasonal patterns - the "eddies" in the atmospheric flow which give rise to storm systems and interesting, impactful weath
101.
by counters 2y ago
> Is there a big Clearinghouse for this data? The short answer is, "no." There are some projects like the "NNJA-AI" project at Brightband[1] which is attempting to create such a clearing house in order to focus resear
102.
by counters 2y ago
1.5 degrees is perfectly fine for predicting large-scale (synoptic) weather patterns. They're not just "global trends." But yes, typical global NWP models and their MLWP competitors are run at 0.25 degrees or finer. All forec
103.
by counters 2y ago
> Google has a local nowcast model that uses raw observations, in production, but that's a different genre of forecasting than the medium-range models of Aardvark. It's very clear from the MetNet announcement blog[1] that they
104.
by counters 2y ago
Then it should be pretty easy to demonstrate fraud and abuse, hmmm?
105.
by counters 2y ago
Actually, the point is that the normal rules _do_ apply, and the normal rules very much encourage the publication of the names and likenesses of civil servants. I'm kind of surprised that no one has made the argument that there is some
106.
by counters 2y ago
I'm not sure I can point you to a single reference, but a good starting point would be the UK Met Office's "Unified Model", which provides a framework for configuring model simulations that scale from sub-mesoscale rapid
107.
by counters 2y ago
> Here, since weather certainly changes its fundamental patterns over time, there is no way of reliably predicting out-sample performance. This seems overly pessimistic. The weather doesn't change its fundamental patterns over time;
108.
by counters 2y ago
> That's actually quite a long time for an AI model. Sure, but it's several orders of magnitudes smaller than the parent, physics-based weather model it's emulating, which may take 2-3 hours of wallclock time on an HPC sys
109.
by counters 2y ago
That's part of the story. The real reason why this is a domain where AI models can thrive is actually because we already have extremely complex, well-developed, physics-based simulations of the atmosphere which can take all of these ob
110.
by counters 2y ago
We don't need to; dynamical meteorology is an incredibly mature field and our understanding of the fluid dynamics of the atmosphere grossly exceeds the resolutions and limitations of coarse, 0.25 degree global numerical models.
111.
by counters 2y ago
The current crop of AI weather models captures a surprisingly robust array of "core dynamics". E.g., recent work by Hakim and Masanam [1] reproduce classic dynamical core test cases on AI weather models and show that in many cases
112.
by counters 2y ago
> Weather and climate models have their own physics, which at the very least means that the solution is physical for the universe that particular model inhabits. The boundary conditions are parameterized, and those can be tweaked as clim
113.
by counters 2y ago
Being the NOAA administrator is a particularly not-fun role if you're primarily a research scientist. I _highly_ doubt Cliff Mass is on any short lists for this job. Several folks from private industry immediately come to mind as likel
114.
by counters 2y ago
Many companies offer customized, high resolution weather simulations that resolve those sorts of features. It's purely a cost vs value proposition - only a tiny sliver of users actually need this sort of spatial resolution, and typical
115.
by counters 2y ago
Sort of. The motivation here is that numerical weather prediction (NWP) models are used for a lot more than just weather forecasting - they're critical research infrastructure and tools. NWP allows researchers to simulate complex atmos
116.
by counters 2y ago
I haven't heard anything about NOAA having any new issues with workforce retention. The trend you're seeing in private sector expansion / growth is linked to the massive investments in Earth observation and derivative applica
117.
by counters 2y ago
It doesn't take care of the errors. They still "accumulate" over time, leading to the same divergence that traditional physics-based weather models experience. In fact, the hallmark that these AI models are _doing things righ
118.
by counters 2y ago
No one is claiming that there is "new knowledge" here. The entire class of deep learning or AI-based weather models involves a very specific and simple modeling task. You start with a very large training set which is effectively a
119.
by counters 2y ago
I agree with other posters that this is overly nitpicky. Take as an example playing a musical instrument. Perhaps you play cello in a community orchestra, and you're presented with a piece that has a passage that you can't reliabl
120.
by counters 2y ago
> One thing we believe, however, is that the reanalysis step in weather forecasting is unnecessary in the long term, and that future (ML) weather models will eventually opt to generate predictions based on un-assimilated raw data and wil

More ›