Tierra Biosciences Secures $11.4M in Series A Funding to Expand Designer Protein-to-Order Platform

Written Posted on

In addition to being computationally expensive, these models limit the size of forecast ensembles and make it difficult to characterize extreme events accurately.

Tierra Biosciences Secures $11.4M in Series: As an alternative to current methods relying on physics-based simulations, Google AI used SEEDS to produce accurate and computationally efficient ensembles of weather forecasts.

Climate change makes weather forecasts more uncertain.

The weather forecast is traditionally generated by simulating the atmosphere’s behavior using physics-based models. In addition to being computationally expensive, these models limit the size of forecast ensembles and make it difficult to characterize extreme events accurately.

Tierra Biosciences Secures $11.4M in Series A Funding to Expand Designer Protein-to-Order Platform

SEEDS, a generative AI model based on diffusion probabilistic models denoising diffusion, has been developed by Google researchers to solve this problem. By using SEEDS, large ensembles of weather forecasts can be generated at a fraction of the cost of traditional methods, improving the quantification of uncertainty and the prediction of extreme events.

Using generative AI, SEEDS produces ensemble forecasts that are as good as physics-based ensemble forecasts. A numerical weather prediction system can create ensembles efficiently based on only two or three forecasts.

Also Read: Experience ChatGPT without an Account: ‘Anyone Curious Should…’ Try it Now

By capturing spatial covariance and correlations between atmospheric variables, ensemble models can provide more realistic forecasts. With 256 ensemble members per 3 minutes on Google Cloud TPUv3-32 instances, SEEDS significantly reduces computational costs compared to traditional methods.

This scalability enables the generation of large ensembles necessary to assess the likelihood of rare but high-impact weather events.

By comparing SEEDS to operational weather forecast systems and Gaussian models, we can see that SEEDS outperforms Gaussian models when it comes to capturing spatial correlations and accurately predicting extreme weather events.

The SEEDS model generated spatially structured forecasts similar to operational forecasts during the 2022 European heat waves, but Gaussian models failed to capture cross-field correlations.

Also Read: Introducing EVI: Hume’s Revolutionary Conversational AI with Emotional Intelligence

Furthermore, SEEDS provides better statistical coverage of extreme events, enabling the quantification of their probability and sampling of weather regimes in which they would occur.

The paper presents SEEDS as a promising solution to ensemble weather forecasting’s challenges. SEEDS uses generative AI technology to create large ensembles that are accurate at quantifying uncertainty and predicting extreme events.

With this state-of-the-art model, people will make decisions in many areas, from emergency management to energy trading.

Also Read: Revolutionizing Retail Pricing Strategies, An Interview with GenAI Chatbot