Seasonal Wind Prediction for the Energy Sector

Seasonal Wind Prediction for the Energy Sector

Understanding future wind conditions through spotting patterns in seasonal wind prediction data

Overview

What was made: 

Designed for the wind energy sector, Project Ukko aims to support wind farm operators and wind energy traders in making decisions, such as when to plan maintenance and where to invest in new wind farms.The interactive browser application allows exploring the data. For example, users can; 

  • Understand the wind level predictions through visualised patterns of data collected since 1981
  • Locate regions with relatively high or low wind level predictions 
  • Identify whether a specific region has existing wind farms

This was enabled through intuitively and beautifully visualised interface (designed by Moritz Stefaner);  Lines on the map show three things at the same time. Opacity shows the predicted strength based on data in the past, its width demonstrates wind speed, and its angle and color predict change in wind speed. Each line can be selected for more information on a specific area.

Background and objectives: 

"The natural environment and human health, infrastructure, food security and access to fresh water are all vulnerable to variations in climate, this presents a hazard and also an opportunity that will increase as the climate changes. Predictions for the coming months or season can help answer the urgent need to adapt to extreme weather events." The art and design company FutureEverything, climate scientists from Met Office and the Barcelona Supercomputing Centre worked together in order to understand the challenge of creating climate services that are useful, usable and effective. They are partners in EUPORIAS, Europe’s flagship project supporting the uptake of climate forecast services.

Outcome: 

-

How it was made

User Workshop & Data Exploration

The Project Ukko team ran a user workshop in order to understand what information users want to draw out from the data. The team also explored data in order to identify waht insight and patterns that they can get from the 51 datasets.

Feeling during this step: 
0
No
Data processing

Project Ukko uses the ten-metre wind speed forecasts from the ECMWF System 4 (Molteni et al., 2011) operational seasonal prediction system. It uses 51 ensemble members of the wind speed prediction. A callibration technique was used in order to minimise the lack of probabilistic reliability. The technique was done through cross-validating predicted and observed values over the period (1981-2014) with the Ranked Probability Skill Score (RPSS). RPSS is used to validate the cumulative probabilities. It also uses wind farm data from www.thewindpower.net/

Open Dataset: 
Feeling during this step: 
0
Yes
Interface Design

Data visulisation designer Moritz Stefaner experimented with dataset through visulising it. This was crucial step in order to present the complex information intuitively, and also beautifully. Mortitz explored the datasets through a data visualisation tool, Tableau with custom code. Through creating several prototypes quickly, he understood how to best present the data through the texture; smoothness, distribution and outliers. (Image courtesy of Moritz Stefaner at Truth & Beauty)

Feeling during this step: 
0
No