Exclusive: A new generative AI weather model from Nvidia Research significantly advances the accuracy of short-range weather forecasting, helping to predict hazards from flash floods to tornado outbreaks, the company first told Axios.

Why it matters: This is the first AI model to demonstrate improved skill at simulating extreme weather events down to the kilometer scale, Nvidia claims.

• If this advance is replicated by other researchers, it could usher in a new era of even more accurate short-term forecasts.
• That could save lives and money by helping people stay safe and protect their property.

The big picture: Until now, AI weather and climate models from Nvidia, Microsoft, Google and researchers elsewhere had demonstrated advances in using AI and machine learning to produce medium-range, global weather projections that rival or beat conventional, physics-based models run on supercomputers.

• In addition to accurate forecasts, the new model, outlined in a study out today as a preprint, could help scientists take global climate change projections and more accurately apply them to local scales.
• This is called "downscaling," and it too has historically been computationally limited.

What they're saying: "I'm convinced we're at that moment now where AI can compete with physics for storm-scale prediction," study coauthor Michael Pritchard, a climate scientist at Nvidia, told Axios.

Zoom in: The new model, discussed in a paper submitted for peer review, is known as StormCast. The paper is written with scientists from the Lawrence Berkeley National Laboratory and University of Washington.

• Nvidia Research claims the model can operate at the mesoscale level, which means it can resolve small-scale weather systems and includes some of the complicated physics of how thunderstorms form, intensify and dissipate.
• Such "convection-allowing models" are used to predict the evolution of derechos, squall lines and deadly flash flood events, among other applications.
• According to the study, the model already beats the accuracy of the top U.S. short-range weather prediction model, known as high-resolution rapid refresh, or HRRR, when it comes to predictions of how storms will evolve on weather radar.

Yes, but: Pritchard cautioned that physics should not give way entirely to generative AI but could be used to help test new models' "physical credibility that [enables] us to trust AI predictions and to understand their boundaries."

What's next: Nvidia is working with The Weather Company as well as Colorado State University to test the new model, and may eventually make it more widely available.

Read more

We're very curious to see how Democrats shape their energy and climate message at their convention that opens today.

Why it matters: Party conventions are high-profile, prime-time chances to pitch big themes — and maybe even offer real policy hints.

A few things we're watching...

⏱️ How much time the top speakers devote to the topics. To date Kamala Harris and Tim Walz have put scant focus on climate, a sign they see other topics like health care and food costs being far more important to voters.

👀 Whether Harris goes there. Republicans are frequently reminding voters of her 2019 calls to ban fracking, which her campaign has walked back.

• I highly doubt she'll address it head-on. But keep an eye on whether Harris offers any views on the domestic oil and gas industry — on prices, on climate, on production, and more.

The latest: The Democratic National Committee unveiled its 2024 platform this morning.

• Platforms are nonbinding messaging documents, but offer clues.
• The new one backs aggressive climate steps, but it's also heavier on energy affordability than the 2020 document.
• It doesn't repeat the 2020 goal of banning new oil and gas permitting on federal lands.
• There are a few new policy ideas, like a "disaster resilience tax credit" to help families and small businesses.

What's next: Look for Democrats to mostly stitch climate into their economic pitch, casting Biden-Harris policies as ways to support domestic clean tech industries and jobs.

• Convention organizers told E&E they expect a part of one evening to highlight climate messaging.

Equity funding for climate tech companies in public and private markets is slumping in 2024, per new data from the research firm BloombergNEF.

Why it matters: These companies provide — or are developing — goods and projects that help cut emissions.

Threat level: Macroeconomic trends like higher interest rates help explain the slowdown, BNEF analyst Musfika Mishi tells Bloomberg.

• But the report also suggests that investor "fascination" with AI startups may be redirecting some cash.

The big picture: The U.S. has a larger piece of this smaller pie, moving ahead of China as the top market with $6.7 billion mobilized, BNEF said.

• The report said the U.S. is increasingly seen as a lucrative region thanks to the IRA, which helps provide a market for climate tech.
• Total funding in the first half totaled $22 billion, about half of 2023's first half.

The bottom line: The slowdown is pronounced but only captures one feature of the finance landscape.

• Overall investments in manufacturing and energy projects are rising domestically amid unprecedented U.S. tax and grant support.

The IRA and other policies are driving U.S. CO2 cuts, but higher project costs and other forces are slowing the pace, a new study shows.

Why it matters: The Princeton-led REPEAT Project provides some of the most sophisticated U.S. policy analysis out there.

The big picture: It sees a 32%-34% cut in U.S. greenhouse gas emissions by 2030 relative to 2005 levels, reaching 40%-43% by 2035.

• That's a lot! But it lags the White House goal under the Paris Agreement of a 50% drop by 2030.

Threat level: Updated assumptions yield smaller emissions cuts than the 2023 version of this annual study.

• There's a stronger economic outlook, higher energy demand, and higher project finance and capital costs.
• There's also more info project announcements to work with.
• Add it up and 2035 emissions are 9-12 percentage points higher than in last year's version.

The intrigue: Recent regulations — like EPA CO2 rules for power plants and cars — would claw back a chunk of those losses if they survive the election and the courts.

The bottom line: Lots of uncertainties — like levels of data center demand — are stuffed into these kinds of estimates. But it's an important guide.

Go deeper: 2024 analysis...lead author Jesse Jenkins' summary thread on X.

"The excitement of the early days has not lived up to the hype."