Wind forecast in open oceans: the PredictWind success story

How PredictWind compared wind observations from weather stations and weather models to create two proprietary weather prediction models.

Download the case study

PredictWind success story

PredictWind provides the sporting and leisure maritime community with high-quality forecasts. The company created two proprietary weather prediction models with resolutions as fine as one kilometer. It provides two alternative forecast models – ECMWF and GFS — for comparison.

Spire’s model ranked #1 for both wind speed and direction.

The challenge

PredictWind needed to identify the best weather forecast provider to equip sailors with accurate and detailed maps, weather routing tools, a departure planning guide, forecast alerts, and live wind observations.

Custom forecasts that prioritize relevant variables were a constant challenge in the maritime sporting and leisure industry.

The solution

PredictWind compared wind observations from weather stations and wind forecasted by weather models for various locations around the world in order to rank weather models in terms of wind accuracy.

This study focuses on the following weather models: ECMWF, SPIRE, UKMO, and GFS.

The study covered 4 months (February to June 2020)

Hourly data used
out to 7 days

Around 166 buoy
stations evaluated

The results

PredictWind’s comparison ranked Spire as #1 for both wind speed and direction. The fine vertical resolution and high quality of Spire’s Radio Occultation observations combined with Spire collecting 10,000 profiles a day results in a homogenous, unbiased, observational data set that is able to cover the entire planet.

Spire Data is not collected in a single altitude, but from the ground up to 120 km. Weather events are observed as they are forming at very high altitudes, giving unique medium-range forecast capabilities. Spire’s satellites collect data all over the planet – including open oceans and remote areas which make for a more comprehensive and accurate weather forecast.

Let’s take a look at the tested metrics:

Wind speed

Plotted in this chart is the wind speed mean absolute error in knots for Spire, ECMWF, UKMO, and GFS. Lower values are better and Spire’s model consistently has a lower mean absolute error (in knots) than the other test models over the course of a 7 day period.

Wind direction

This chart shows the comparison of the mean absolute error, this time in degrees, for wind direction (again, lower values are better). Results are even more conclusive than the wind speed results: Spire’s weather forecast model ranks #1 in this metric for each day tested.

“The Spire Weather Forecast excels for open ocean weather forecast accuracy. The level of development and speed of innovation that Spire is putting into its models continues to impress us.”
Nick Olson, Marketing & Business Development Manager, PredictWind

Share this quote

Download the case study