Event Report - Seasonal Forecasts and Their Use in the Energy Sector

The first event organised by the Royal Meteorological Society’s new Energy Special Interest Group (SIG) was held on 30 November. The free webinar focused on seasonal forecasts and how they are used by the energy sector and was chaired by Chris O’Reilly (University of Reading). Adam Scaife (Met Office) gave an overview of seasonal forecasts and explained the upcoming winter outlook, followed by four energy sector panellists giving an overview of how they use seasonal forecasts. Speakers were: Laurent Dubus (RTE), Rebecca Jones (bp), Lucy Field (AFRY) and Peter McAward (SSE). The event attracted over 180 attendees from across the energy sector.

Seasonal Forecasts and Their Use in the Energy Sector is the first event in a series of planned engagements with the energy sector. If you want to help us shape future events and programmes of work in this sector, please fill out our short survey on Google Forms. You can also join our email list to keep up-to-date with the Energy SIG’s upcoming activities. Details to sign-up are on the survey or available via our dedicated web page.

Seasonal Forecasts

Adam Scaife began the event by giving an overview of the science behind seasonal forecasts, where he discussed sources of skill at the seasonal timescale. He explained that seasonal forecasting models represent the physics of the atmosphere and broader climate system, including the ocean. Forecasts produced using these models are typically produced as “ensembles”: taking multiple sets of plausible initial conditions (based on knowledge of the current state of the climate) and stepping each forward through time to produce multiple predictions of potential future weather states. This ensemble of possible future states enables the probability of certain outcomes to be estimated. Thus, whilst it is not possible to use seasonal forecasts to predict a single definite outcome (or indeed to rule out outcomes), they can help to inform which types of conditions are more likely or less likely in the upcoming season, particularly when these conditions can be clearly linked to the behaviour of large-scale atmospheric circulation, such as the winter North Atlantic Oscillation. Adam also explained the role of the tropics and the stratosphere in providing important sources of long-range predictability for the UK and Europe.

The Outlook for the Upcoming Winter

A key message from the webinar was that, whilst a wide range of winter weather outcomes remain possible, recent forecasts indicate that December is likely to be colder, stiller and drier than average in the UK. Into January and February, the forecast shows a change of weather type with milder westerly conditions favoured. This was the first Met Office 3-month outlook to report wind speeds, showing that December (and the winter as a whole) is more likely to have calm than windy conditions.

It was stressed that these statements are probabilistic: it remains possible that very different weather conditions could be experienced across the winter period. For example, the possibility of cold and dry late winter was not ruled out but is estimated to be less likely given current conditions.

Image above shows probabilities that temperature, rainfall and wind speeds will be below-average, near-average, or above-average for December (on the right) and the three months from December to February (on the left). Source: Met Office, November 2022.

How Seasonal Forecasts are used in the Energy Sector

We then heard from our four panellists, each giving an overview of how their organisation uses seasonal forecasts.

Laurent Dubus spoke about the challenges faced by RTE in managing France’s grid at the moment. The low availability of nuclear power and hydroelectric generation variability means that demand must be carefully managed, particularly at peak times and during cold snaps. This includes measures to support peak demand reduction by reducing overall consumption and increasing energy efficiency; informing the public of the times of day they should or shouldn’t use energy; and, as a last resort, limiting supply.

Graphic above showing the methods employed by RTE to manage energy demand. Source: Laurent Dubus, RTE, November 2022.

Rebecca Jones informed the audience that seasonal forecasts are very important for setting a benchmark for gas prices and trading a few months in advance. Rebecca explained this via the global nature of energy markets: prices depend on whether the whole northern hemisphere experiences cold weather simultaneously or, for example, if just Europe is colder than usual. She also noted that cold winters tend to be calmer, reducing wind power generation and increasing gas demand. Seasonal forecasts also inform oil and biofuel trading; these markets are less volatile than the gas market, so that traders can plan further into the future. 

Lucy Field focused on how this year’s seasonal forecast is all the more important given the current geopolitical situation. She showed that, due to the mild autumn and concerted effort to maintain full tanks, the EU’s gas reserves are currently fuller than they were at this point in 2021 but have started to decline now the weather has turned colder. A cold snap at the end of winter is particularly likely to increase gas prices, as reserves will already be fairly empty once we reach mid-to-late winter.

Graph above showing the EU’s gas storage inventory in 2020, 2021 and 2022. Source: IEA WDS, ENTSOG, S&P, www.montelnews.com, November 2022.

Peter McAward works in SSE’s renewables division and presented how seasonal forecasts influence the amount of electricity SSE commits to producing for the National Grid over the winter. Based on wind, precipitation and temperature predictions, SSE agrees it will produce a certain amount of electricity from its wind farms and hydroelectric dams. If renewable sources are insufficient to cover energy demands, then SSE must burn gas to make up the deficit. Peter finished by noting that, even though seasonal forecasts are low-skill and come with caveats, they are very useful for the energy sector when used correctly.

The event closed with a discussion between Adam and the four panellists, including questions from the audience. Questions focused on specifics of the seasonal forecast and clarifications on how skilful they are and over what time periods. Additionally, the discussion focused a lot on the uncertainties in seasonal forecasts, how these compare to other uncertainties in the energy sector, and how the panellists deal with this information. Finally, the Net Zero transition was a topic of interest, with the audience asking whether seasonal forecasts would become even more important as society shifts to more renewable energy generation.

Future Events

In the post-event survey, 85% of attendees said they would like to see more events like this, so we hope this will be the first of many events focusing on the energy sector in the future. Stay up-to-date by signing up to our email list.

About the Author

We are sponsoring three 12-month mini Fellowships to engage with our Science Engagement activities, including supporting the delivery of our Climate Action Plan. Our Science Engagement Fellows will have the opportunity to engage with non-academic stakeholders to build multidisciplinary partnerships, develop engagement skills and experiences, enhance their profile with stakeholders and leading academics, and grow their research impact.

Matt Wright

Hi! I’m Matt, and I’ve just started a PhD at the University of Oxford in the Physics department, working on improving seasonal forecasts. My project is partially in collaboration with AFRY’s energy team. I’ve always believed that weather and climate science have the potential to benefit society, and I’ve already found that loads of RMetS’ work is relevant to a variety of stakeholders in the energy sector. I am really looking forward to engaging with the energy industry to show them how meteorology links to their work and how RMetS can better serve them in the future!