The real world is neither black, nor white. It is made from shades of grey. The weather forecast can contain many possibilities, but forecasters usually try to distil it down to the most likely outcome. Why should a weather forecast try to determine a single ‘black’ or ‘white’ outcome, rather than present a more well-rounded picture of all of the shades of the weather that could possibly come?
Up to now, traditional weather forecasting has tended to rely upon determining a single (most likely or deterministic) solution; discarding many other potentially useful possibilities in the process. This method has been based on what forecasters can best provide, rather than necessarily what users of the forecast may actually need.
But times are changing. We will soon be giving customers the forecast that best suits their needs by using more of the available information to create more accurate and a more useful forecast that enables better decisions to be made. Providing different possibilities for the weather forecast uses a method known as ensemble forecasting.
Some decisions could be mistakenly regarded as simple, not requiring too much detailed weather forecast information. Ken Mylne is an ensemble forecasting expert. He said: “Even a simple decision like ‘should I take an umbrella tomorrow’ will depend on how much you’re concerned about getting wet.
“If you are dressed up for an important event then any likelihood of rain is sufficient to encourage you to take an umbrella, but if you were dressed casually for a walk with the dog you may have fewer concerns about getting a little wet.”
But what about an energy company concerned about whether strong winds will trigger power outages? Here lots of decisions need to be made. Should more engineers be put on standby to cope with power outages? Should generators be available to provide energy in certain locations? Should vulnerable customers be warned? All of these actions can have hefty cost implications, so being able to assess the risk ahead of making those decisions provides greater confidence.
Customer benefit
These are cases where the customer would benefit from having access to much more of the information to make informed decisions, rather than relying on the conventional ‘black’ or ‘white’ outcome. For example, knowing the chance of wind speeds approaching the tolerance of power cables to withstand them will allow more balanced decisions to be made. Some customers find it vital to know if there is only a small chance of significant impacts.
Ken Mylne concluded: “We believe the greatest value of ensembles will be realised with so called HILL events – these are weather events which although deemed to be low likelihood can have extremely high impacts. These events can include windstorms, thunderstorms, snowfall and intense rainfall.
“That is why we are moving away from the single ‘deterministic’ outcome to a process called ensemble forecasting, where many forecasts are produced to see not only the most likely outcome but also to assess the risk of otherwise unlikely events causing considerable impacts. We will still be providing simple forecasts for those customers who don’t need more detailed information, but for those who do our forecasts will reflect their needs.”