This isn't your grandmother's forecast.
Well, I guess it could be if your grandmother is a quantitative commodities trader or supply chain/logistics manager. Or if she happens to be a sales analyst, inventory specialist, or distribution manager. Or maybe she's developing a machine learning algorithm that requires well-formatted global weather data. Well...you get the point.
Weather forecasting models have been using the same atmospheric flow calculations since 1922. It’s a system rooted in plugging a set of initial conditions into physics formulas and then seeing where they may lead over time. Modern technology has allowed us to use better initial conditions and now repeat those calculations millions of times but we still simply pick the least absurd outcomes. This has proven to be effective in short term forecasting but becomes unreliable outside of the 7 to 10 day window. Why does this happen?
Physics models are trying to simulate and predict the conditions of the entire atmosphere. But they can only approximate those conditions. The errors in their approximations compound over time, rendering their longer range forecasts useless. Weather 20/20’s model takes a different approach. We don’t attempt to simulate the atmosphere; We identify cycles in the atmosphere's behavior over time. These cycles can then be used to project forward actual weather conditions over long windows of time. We use what actually happened to predict what is going to happen.
The LRC is Weather 20/20’s proprietary methodology used to identify the foundational cycles of our long range modeling data. The core tenet of the LRC is a unique pattern that establishes itself each year. Key influences of the LRC include (but are not limited to) El Nino/La Nina patterns, harmonics and the Arctic and North Atlantic oscillation cycle.
One of the only companies in the world that offers accurate long range weather data and analytics on a truly global scale.
Everything you need to be successful
Pick a place on the earth, select a time cadance and get weather data. Available in any file size, big or small. Consume at your convenience.
Get access to our long range weather forecast API and plug it into a new or existing solution. It's lightening quick with extremely flexible call rate allowances. VIEW DOCUMENTATION HERE.
Pinpoint key market and drivers. Ideal for localized needs such as energy demand and agricultural outlooks.
See the future in a whole new way. Using best in class engine and visualization software, we highlight key metrics and represent them in a way that is easily discernible giving organizational confidence to take operational action.
We continually analyze our long range forecasts to find events that matter. Set alerts for anomalies such as droughts, floods, severe storms and more. We pour over the data so you don’t have to.
Use Weather 20/20’s advanced data engine and team of data scientists to uncover organizational performance patterns and predict future outcomes based accurate long range forecasting.
Subscribe to our new Weather 20/20 Substack Channel to stay prepared and informed so you can make better decisions for your growing season.
Weather 20/20’s team of meteorologists have studied the LRC® collectively for 75+ years. Apply cyclical nuance to your organization’s weather strategy.
Is it accurate? Yes, it is. This is an intensely simplistic answer to an absurdly over-simplified question. Reducing the complexity of forecasting weather weeks and months ahead for every square inch of the globe to a single numerical representation feels disingenuous. So we're not going to do that. In fact, others in the forecasting industry tend to invent their own systems for tracking accuracy in order to look better in marketing. We do not. We believe there is trust in transparency.
No statistical massaging. No gimmicks. Just raw data.
When you click the link below you will be presented with a map of the earth separated into color coded cells. Each cell is one of our forecast zones and equals 97.6 mi2 (252.9 km2). There are 3 possible accuracy values for each cell that can be toggled via the 3 layers on the left. Each layer represents the accuracy of using the LRC for the 10 year period of 2010 to 2020 for that metric compared to the 30 year average for either high temperature, low temperature, or precipitation. The decimal represents a percentage of accuracy. For example a value of 0.708.. for a cell can be interpretted as being 70.8% accurate. Higher is better.
So, go ahead. SEE FOR YOURSELF.
Here's a quick look at the 3 main components that make our amazing forecasting technology possible. We use cutting edge tools of today to forecast the future.
It all starts with a whole lot of data. Each night we collect trillions of data points for the entire planet from multiple sources like NASA, NOAA, and ECMWF. As a result we have decades of global data at our (and your!) disposal. All that data is what gives us our ability to find patterns that lead to accurate long range weather forecasts.
A massive amount of processing power is required for harvesting, pruning, organizing, analyzing, storing, protecting, and delivering the hordes of data necessary to bring you accurate long range forecasts. Apache Spark and Kubernetes allow us to distribute and scale power as needed while Data Mechanics helps control computing costs. It's a beautifully choreographed data dance.
Once we have all this historical and accurate long range forecast weather data we need to give it a home where it can easily be delivered by any method in any format necessary. And of course that delivery method needs to be fast. A brilliant combination of AWS S3, Apache Parquet, and AWS hosted Apache Cassandra makes it possible and affordable for data enthusiasts on both sides of the pipeline. We can't wait for you to experience it first hand.
A member of sales or support will contact you soon.