Solving major environmental forecasting gaps for climate adaptation planning.
Addressing the Gap
Planette’s patent-pending forecasting technology is built for the pace of decision-making and grounded in 15 years of near-term climate prediction science, enhanced and operationalized with the speed of AI. Our forecasts operate when critical business decisions need to be made: quarterly, seasonally, annually...up to 5 years in the future.
The predictive power of physics with the acceleration and efficiency of AI.
We build high-resolution simulations with state-of-the-art Earth system models, initialized with key scientific observations of the global oceans.
The Planette climate sciences team utilizes Fourier Neural Operators and other cutting-edge AI methods to boost our forecasting ensemble size and accelerate forecast production.
Actionable Climate Insights
Weather forecasts can only make skillful predictions up to two weeks in advance; other forecasting solutions on the market only offer limited forecast variables up to one year.
Planette's near-term predictions deliver climate intelligence over seasonal, quarterly, annual, and multi-annual time horizons, making our climate forecasting engine essential for business decision-making.
Planette's Climate Forecast Products
Planette delivers critical climate insights to a wide range of businesses and industries. Our forecasts are updated quarterly and automatically connect to your internal systems with our data API. Planette forecasting variables are available over a 25-km grid covering the entire globe so your business can incorporate localized environmental insights for the specific locations affecting your business operations.
Planette assesses forecast quality by evaluating how well our climate model predicts the past ("Hindcasts"). Our Hindcasts analyze historical data based on the locations specified by each customer. We share these Hindcasts with potential customers to establish trust - so you yourself can evaluate whether our forecasts are sufficiently accurate for your business needs.
Planette Precision 101
Hindcast* MVP - Q1 2024
Forecast MVP - Q2 2024
Precision 101 Forecast Variables:
100-m Wind Speed
Surface Solar Irradiance
Planette Precision 201
Hindcast* MVP - Q4 2024
Forecast MVP - Q2 2025
Precision 201 Forecast Variables:
Extreme Precipitation Risk
Severe Convective Storm Risk
Severe Hail Risk
Quarterly to 5 years in the future.
Our forecasts predict ups and downs in key environmental variables, at 25-km resolution, covering the globe.
Planette's Climate Forecast FAQs
How is it even possible to forecast at seasonal, quarterly, annual, or multi-annual time horizons?
Seasonal, quarterly, annual, and decadal weather prediction is a relatively new area of climate science research, developed incrementally over the last 10 years. At these time horizons, the state of the ocean and its predictability enables Planette to predict conditions for key environmental variables over land masses.
How is this different from weather forecasting?
Weather forecasts generally allow for predictability of meteorology up to 14 days into the future, based on a set of initial meteorological conditions. Planette's precision forecasts, on the other hand, allow for predictability over much longer time horizons because we harness the predictability of the ocean. While we cannot tell you on what day an extreme weather event will happen in the future, we can tell you the probability of such an event for a given month, up to 60 months (5 years) in the future.
Why isn’t NOAA, NASA, or some other Federal Agency already doing this?
These federal agencies primarily focus on advancing scientific research, not providing actionable data for businesses. In fact, NOAA does provide coarse seasonal and annual outlooks for temperature and precipitation, but these forecasts don’t include extremes, and don’t extend beyond the continental U.S., making their usability limited for business purposes.
How do you know that Planette forecasts are accurate?
The Planette climate science team assesses forecast quality by evaluating how well our model predicts the past ("Hindcasts"). Our Hindcasts show that our model is 80% accurate over the globe (for 1-year forecasts), and over 90% accurate over many regions. We share these Hindcasts with potential customers to establish trust - so they themselves can evaluate whether our forecasts are sufficiently accurate for their business needs.
How can the Planette climate sciences team achieve these amazing results?
AI is revolutionizing weather forecasting, and we know that it also has the potential to revolutionize seasonal, quarterly, annual, and decadal forecasting. AI and ML have the power to find non-intuitive patterns where traditional methods fail. Planette uses patent-pending AI algorithms to improve, accelerate, and amplify forecast production from physics-based models, making our scientific forecast predictors commercialization viable.