WEMC’s Reanalysis Working Group: Maximising the Value of Reanalysis Data for the Energy Sector

How much will a wind farm produce? How heavy should offshore turbine foundations be? How to design an optimum weather-driven energy grid? For answering these questions, engineering use historical weather datasets called “Reanalyses”. Originally designed by atmospheric scientists, for atmospheric scientists, Reanalysis datasets are now widely used in the Energy sector where they form the backbone of crucial engineering and commercial activities.

We caught up recently with the members of WEMC’s Reanalysis Working Group (RWG) a group of practitioners from the Wind- and Power Grid engineering communities who aim at maximising the value of reanalysis data for the energy sector.

Reanalysis Matters

Reanalysis datasets are fundamental to the success of offshore wind and power system projects. Since weather and climate conditions are key drivers for both the design and operation phases, accurate environmental data is essential. Reanalysis products, which combine observations with numerical models to generate consistent climate records, are used either directly or as inputs for downscaling techniques to achieve finer spatial and temporal resolution. Their role is critical in supporting decisions throughout the project lifecycle, from early feasibility studies to operational optimisation.

Figure 01: Illustration of the famous 1953 storm over the North Sea, in the ERA5 reanalysis. The image shows surface wind speeds in blue and pressure contour lines. Reanalysis are “maps without gaps”, and are produced using weather models fed with historical observations. Source: C3S updates its ‘maps without gaps’ for 1959-1978

Specifically, reanalysis datasets are extensively used to characterize wind, waves, water levels, and ocean currents, which are all vital parameters for wind farm design and project planning. They also provide the basis for predicting long-term wind farm production, helping developers assess future energy yields and financial viability.

Beyond individual renewable projects, reanalysis data are also crucial for planning and modelling broader power grids, ensuring system stability and efficient integration of renewable energy sources. Their versatility and comprehensive coverage make them indispensable tools across the energy sector.

Figure 02: Reanalysis (here ERA5) is often used as baseline for assessing the impact of climate change on future, weather-driven power systems. Source: Weather- and climate-driven power supply and demand time series for power and energy system analyses

Improving the value of Reanalysis

The RWG is a self-organised initiative supported by WEMC, bringing together wind, power grid, and marine engineers and scientists. It facilitates collaboration between reanalysis centres and end-users to maximise the value of reanalysis data for the energy sector.

The RWG has four main goals:

Collection of user input and feedback (see the example below, with the testing of ERA6 together with the ECMWF)

Reanalysis validation using sector-specific in-situ measurements, and incorporation of electric sector meteorological data into reanalysis products. The RWG promotes the use of publicly available data such as the ones listed on the Wind Resource Assessment Group wiki page.

Fair and proper use of reanalysis for Wind Energy and Power System Modelling

Advocacy for reanalysis

The RWG is a spin-off from the International Energy Agency Wind Technology Collaboration Program (IEA Wind TCP) Topical Expert Meeting 111, held in April 2024.

“The wind energy practitioners have shown great interest in co-developing the next-generation global reanalysis ERA6 from the European Union Copernicus Climate Change Service (C3S). Being a user driven programme we were delighted by such an interest and openly engaged with them. As a consequence of this interaction, specific new products will be added, such as height-level data. The wind energy community has also provided us with key feedback on the test products we made publicly available for that purpose. We look forward to continued fruitful exchanges and welcome their support, as we are now in the last preparation stages for launching ERA6 production.“

 Carlo Buontempo, Director of C3S.

Figure 03: Participants to the TEM111 in Denmark, where export from wind, power-grid and reanalysis communities exchanged and discussed collaboration opportunities

ICEM 2025 Workshop on Reanalysis

A dedicated workshop on reanalysis was held at the ICEM 2025. It was held by Justin Sharp (EPRI), a founding member of the RWG and renowned Energy Meteorologist, and Hannah Bloomfield, Academic track fellow in climate resilient energy systems. The workshop gathered approximately a 50 persons, most attendees not being experts in Reanalysis, but relying on Reanalysis for their work.

The workshop began with a tribute to our friend and colleague John Zack (1952 – 2025), founder of MESO Inc., AWS TruePower (now part of UL) and world-famous expert in Energy Meteorology. John passed away shortly before the ICEM, he will be dearly missed.

This workshop began with a general presentation by Justin on the current and future challenges related to the use of reanalysis datasets for wind energy and power systems, along with an introduction to the newly established RWG.

Two short presentations from Reanalysis users were given. First Laurent Dubus from the French TSO RTE, entitled “Use of reanalysis data at RTE”, and then a second presentation from Matti Koivisto from DTU regarding “Post-processing to increase spatial and temporal resolution of reanalysis wind speed data”.

The workshop then moved into an interactive discussion with the audience, focusing on two key actionable topics: how to enhance the collection of user input and feedback for existing and future reanalysis products, and how to better advocate for the fair and proper use of reanalysis datasets in wind energy and power system modelling. The session aimed at engaging participants in shaping a collaborative path forward to address these challenges.

In effect, with increasing weather sensitivity in modern power systems, planners require high-resolution, long-term datasets that accurately capture complex space-time relationships among critical meteorological variables. Although a range of reanalysis datasets exist — often blending observational data with numerical weather prediction models or enhanced through physical or machine learning approaches in the coming years — users often lack clear guidance for selecting datasets fit for specific applications. Consequently, choices are frequently based on institutional habits or dataset popularity rather than suitability, risking flawed planning outcomes.

Drawing on the 2023 ESIG report Weather Dataset Needs for Planning and Analyzing Modern Power Systems authored by Justin Sharp, and the 2024 IEA Topical Expert Meeting #111, the session featured a panel discussion focused on identifying the major gaps in current datasets and the most promising pathways to improve dataset quality and user decision-making for future energy system planning.

Three main items were discussed the workshop:

1.- Future datasets for energy-meteorology applications should explicitly communicate their limitations, especially as reanalyses are increasingly treated like pseudo-observations. There is a clear demand for high spatial and temporal resolution, though the specifics of “high” vary by user need. Many global or continental data providers may not be best suited to generate the hyper-local, niche variables required—such as soil temperature, relative humidity, or wind at 100m and above—which are essential for the energy sector. Users must also be informed when further calibration of reanalysis may be required, for example to improve the representation of local diurnal cycles. When performing these calibrations (against available observations) they must be statistically robust. multivariate bias correction methods are also crucial to better capture compound extremes, particularly during critical periods of high energy demand and low renewable output.

2.- Choosing the right dataset depends heavily on the use case. A large-scale power system modeller may find hourly reanalysis data sufficient given the overarching uncertainties in energy policy and infrastructure development. In contrast, site-specific applications like wind farm assessment might require sub-hourly resolution. Different temporal scales and metrics (e.g., short-term dunkelflaute vs. seasonal storage needs) bring different error sensitivities, and it remains unclear whether ultra-high-resolution datasets meaningfully improve long-term averages. Understanding and addressing these varied needs is key to effective data application. The workshop attendees were mainly from a power system background, which clearly highlighted the importance of education around this point, and an understanding of how far you can ‘push’ reanalysis based products.

3.- Some improvements are relatively easy to pursue. Making existing observation networks (particularly from renewables sites) more open and better quality-controlled could reduce duplication of effort and significantly enhance calibration datasets. There’s also a pressing need for clarity and transparency in data shared through platforms like ENTSO-e, especially with recent changes in contributions from countries like the UK. Finally, AI offers practical help in automating data cleaning and metadata generation, allowing researchers to concentrate on scientific insights rather than repetitive preprocessing tasks.

In the coming weeks, the RWG will follow-up on any actions proposed at the workshop and continue working towards increasing the value of reanalysis for the energy sector.

On behalf of the RWG, Rémi Gandoin, Justin Sharp, and Hannah Bloomfield.

climate change, climate data, climate impact, climate risk, data scientists, energy sector, ICEM, workshop