• Wind and turbulence profiles with respect to wind energy applications (measurements and theory) including wakes within a wind farm;
• Clouds and aerosol properties with respect to solar energy applications (measurements and theory);
• Marine renewable energy (wind, wave, tidal, marine current, osmotic, thermal);
• Meteorology and biomass for energy;
• Impact of wind and solar energy farms and biomass crops on local, regional and global meteorology;
• The use of numerical models and remote sensing (ground based and from satellites) for renewable energy assessment studies;
• Research on nowcasting, short term forecasts (minutes to day) and ensemble forecasts and its application in the energy sector;
• Quantification of the variability of renewable resources in space and time and its integration into power systems;
• Impacts of long term climate change and variability on power systems (e.g., changes in renewable resources or demand characteristics);
• Practical experience using meteorological information in energy related applications.
The expansion in offshore wind energy leads to more large wind farms being built in close proximity to each other. The wakes of upstream wind farms can influence downstream wind farms, even if they are tens of kilometres apart, exposing them to reduced wind speeds that result in lower power production. To mitigate this effect, it is desirable to reduce the extent and intensity of wind farm wakes by enhancing wake recovery. Suitable mitigation strategies could include adaptations in the wind farm layout, turbine type or operation strategy. However, knowledge on the wake recovery mechanisms of large wind farms is still limited. In particular, the influence of the wind farm setup and operation on different recovery mechanisms is unclear. In this study, we aim to understand the formation of circulation zones inside and in the wake of large wind farms, which can contribute to wake recovery by advecting momentum. We set up idealised large-eddy simulations with a conventionally neutral boundary layer with common offshore characteristics for three different wind farm setups: First, a baseline setup of an aligned wind farm, second, a vertically staggered wind farm with alternating columns of low and high turbine hub heights, and third, a wind farm with alternating columns of positive and negative turbine yaw misalignment. We analysed the budget of the streamwise vorticity to understand which terms are relevant for creating circulation zones. Based on that we identified two different types of circulation zones, inside the wind farm wake area and at the farm wake edges. Inside the wind farm, vorticity is mainly generated because of the rotation of the wind turbines and partly because of spatial variations in Reynolds stresses and buoyancy. At the left and right farm wake edge, vorticity is mainly created by the buoyancy term. The vorticity at the left wake edge is stronger than at the right edge, which is related to the wind farms geographical location on the northern hemisphere. The wind farm setup has a considerable influence on the magnitude of the vorticity. Both, the vertical staggering and organised yaw misalignment cause stronger vorticity inside the wind farm compared to the baseline setup. This enhanced vorticity persists multiple tens of kilometres downstream and contributes to heterogeneities in the wind farm wake. Irrespective of the setup, the dominating mechanism for enhancing and maintaining vorticity in the far-wake of the wind farm is buoyancy. Overall, the results show that wind farms with heterogeneous setups can enhance vortex structures in wind farm flows, which results in a more heterogeneous wake and influences the wake recovery. Given the intrinsic relation between buoyancy and the vertical temperature profile, future work should characterize the effect of atmospheric stratification on the circulations.
How to cite: Gaiser, A., Steinfeld, G., Centurelli, G., and Kühn, M.: Characterisation of vorticity generation in the wake of different wind farms, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-371, https://doi.org/10.5194/ems2026-371, 2026.
Rapid growth of the wind energy industry has led to increasingly congested wind harvesting areas, necessitating that numerical weather prediction (NWP) models better resolve wind farm effects. HARMONIE-AROME is a widely used, operational, mesoscale NWP which has been shown to effectively capture wind farm physics using the established Fitch wind farm parameterization (WFP). The Royal Netherlands Meteorology Institute (KNMI) operates HARMONIE with the standard grid-size of 2 km; however, this scale is relatively coarse compared to wind farms. Within this framework, wind turbines are implemented as a momentum sink, and multiple turbines can occupy one grid point. Various studies have shown that including this WFP is necessary for accurate weather predictions near wind farms. However, a smaller grid-size can potentially improve model performance by better resolving the wind farm–atmosphere interactions.
To this end, we investigate the effect of reducing the grid-size to 1 km for one week when various relevant observations are available to support validation. In the interest of offshore wind farms, we evaluate the model performance by comparing the predicted wind speed to offshore measurements with varying proximity to wind turbine arrays (FINO 1 and 3 masts, Europlatform Lidar, and X-wakes flight). Additionally, we compare to onshore measurements in relatively turbine-free conditions (Cabauw mast) to confirm that the WFP does not degrade the performance for this configuration. Overall, we observed that near wind farms, the WFP improves model performance compared to simulations without WFP, as expected. We also observed that at locations farther from the wind farm, the contribution and effectiveness of the WFP can be dependent on the wind direction. Most notably, we observed that reducing the grid resolution to 1 km did not exhibit a consistent advantage in predicting wind speed at a point location over time. However, we found that increasing the resolution can improve predictions over a spatial trajectory. Therefore, a higher resolution is recommended to effectively study wind farm physics, such as wind speed distribution throughout the farm, wake development, and farm-to-farm interactions.
How to cite: Strickland, J. M. I. and Theeuwes, N. E.: Grid resolution effects on wind farm representation in HARMONIE-AROME, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-596, https://doi.org/10.5194/ems2026-596, 2026.
The North Sea is becoming increasingly densely populated with offshore wind farms and
this trend will continue in the future. In the coming decades, the power production of these
wind farms will depend on how the wind resources in the North Sea are affected by climate
change. The frequency and intensity of extremely low or high wind speeds could change, which
would significantly impact a renewables-based electricity system. This work focuses on prolonged
periods of low wind speed, so-called persistent lows, since they cause long-lasting drops in the
wind energy production.
In this study, persistent lows are defined in terms of the capacity factor, which is the ratio of
the actual energy generated by the wind turbine to the maximal energy that could be generated
under ideal circumstances. A persistent low of, for example, two days with a capacity factor
below 10% is a period of at least two days in a row during which the daily mean capacity factor
was below 10% each day. This work aims to answer the following three research questions related
to these phenomena.
Firstly, how often do persistent lows occur in the North Sea? To answer this question,
profiling lidar data from 10 stations in the North Sea are analyzed for persistent lows over a
range of capacity factors, which are estimated from the wind speeds using a reference wind
turbine.
Secondly, are persistent lows also captured by reanalysis data and climate models? Here,
CERRA and ERA5 are the reanalysis datasets of interest, and for the climate models the CMIP6
HighResMIP ensemble is chosen. The representation of persistent lows is investigated by first
validating the reanalysis data using lidar data during the active period of the lidar at its location.
Afterwards, the reanalysis data is used to validate the historical climate models. The findings
show that the reanalyses recreate the persistent lows found in the lidar data well, while the
results are model-dependent for the climate models.
Lastly, how does the frequency of these persistent lows evolve in the future? Here, the
future and historical simulations from the HighResMIP climate models are compared. Instead
of focusing on the locations of the lidars, the locations of current and planned wind farms in
the North Sea are selected to make a direct link to the impact of persistent lows on energy
production. Preliminary analysis shows no statistically significant climate signal.
Acknowledgements
This work has been financially supported by the Flemish Government through the Agency for
Innovation and Entrepreneurship (Vlaams Agentshap Innoveren en Ondernemen, VLAIO), in
the context of the CORE project. This study has also been partially supported by the DTWO
project, funded by Horizon Europe grant no. 101146689.
How to cite: Dierickx, A., Gremmo, S., Caluwaerts, S., and Munters, W.: Persistent wind speed lows in high-resolution global climate models: future projections and validation with lidars, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-200, https://doi.org/10.5194/ems2026-200, 2026.
In recent years, wind energy has emerged as one of the most essential energy sources in the transition towards a low-carbon future for climate change mitigation. Both onshore and offshore wind farm installations have expanded rapidly over the past decade, and they are considered a crucial provider in a cleaner energy mix. Specifically, the offshore wind farm installations that are currently operational and those planned for the near future over the North Sea will result in dense structures with a large number of wind turbines.
While many studies, using both measurements and mesoscale models, have shown that wind farm wakes can substantially influence neighboring farms by reducing wind speeds and enhancing turbulence, other flow phenomena also play an important role. Among these are atmospheric gravity waves (AGWs), which can modify the surrounding flow and impact turbine operation. Although the concept of wind‑farm‑induced AGWs has been explored over the past decade, most existing research has relied on idealized large‑eddy simulations or reduced‑order models. Consequently, fundamental questions remain regarding how wind farm-triggered AGWs form under realistic atmospheric conditions, how frequently they occur, and how strongly they influence wind‑farm performance.
In this work, we use lidar observations from five offshore platforms and satellite aperture radar (SAR) data to assist in examining multi-year mesoscale simulations using the Weather Research and Forecasting (WRF) model at a spatial resolution of 1 km. WRF driven by ERA5 over a multi‑year period reproduces realistic atmospheric variability and stability regimes, thereby supporting the assessment of the conditions that help wind-farm-triggered AGWs to develop. Our study suggests the AGW signatures appear to be either triggered or intensified by the presence of a large offshore wind‑farm cluster. Building on these findings, we propose a framework to evaluate AGWs that originate from, or are strengthened by, the 4 GW Belgian–Dutch offshore wind‑farm cluster in the Southern Bight of the North Sea.
Preliminary results show that, although WRF tends to misrepresent the amplitude of AGW-related perturbations relative to lidar measurements, the model reliably captures the spatial and temporal patterns of AGWs in several analyzed events. These events were identified based on AGW‑like structures visible in SAR imagery and in the year‑long WRF simulations, and subsequently validated against the available lidar datasets. To support this identification, we estimate the atmospheric stability conditions and AGW wavelengths, while additional diagnostic methods, such as power‑spectrum evaluation, coherence analysis, and autocorrelation, are applied at the lidar sites. These promising results demonstrate the potential of our approach for advancing the detection and understanding of AGWs in real offshore wind‑farm environments.
How to cite: Palatos-Plexidas, A., Guo Larsén, X., Fischereit, J., Gremmo, S., van Beeck, J., De Cruz, L., and Munters, W.: Wind-farm-triggered Atmospheric Gravity Waves: Bridging Mesoscale Modeling with Observations, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-525, https://doi.org/10.5194/ems2026-525, 2026.
Reliable, high-resolution wind climatologies at turbine hub heights are fundamental to wind energy planning, yet existing products for Austria either lack observational constraint (ERA5, NEWA) or do not provide the continuous hourly time series needed to quantify variability and extremes. We present the Austrian Windatlas, a 25-year (1997–2021), hourly, 1 km observation-constrained wind analysis with option for near-realtime production from near-surface to 220 m above ground.
The atlas is produced via a two-stage pipeline. In Stage 1, near-surface (10 m) wind speed fields are reconstructed from the dense GeoSphere Austria TAWES network (~280 stations) using a shared EOF/rPCA decomposition framework that enables direct, fair comparison of six interpolation families: regression-kriging, Bayesian Additive Models for Location Scale and Shape (BAMLSS), random forest, and three deep-learning variants, each tested with and without ERA5 or CERRA reanalysis backgrounds. Validated against 54 permanently withheld stations, the enhanced deep-learning architecture with CERRA background achieves the best overall skill (RMSE = 1.46 m s⁻¹, r = 0.65 for 2020), outperforming the operational INCA+CERRA analysis by ~8% and demonstrating that the higher-resolution CERRA reanalysis (5.5 km) consistently provides better background constraint than ERA5 (31 km) across all method families. Notably, simple regression-kriging (RMSE = 1.51 m s⁻¹) remains highly competitive at a fraction of the computational cost.
In Stage 2, the gridded 10 m fields are extrapolated to hub heights (80–220 m AGL) using a further zoo of methods — from classical log-law and terrain-adaptive power-law formulations to a NEWA-trained, height-agnostic machine-learning model — with final validation against independent NEWA profile holdouts and operational SCADA data from Austrian wind turbines. The inter-method ensemble spread is retained throughout as a spatially explicit, per-timestep uncertainty estimate, propagated end-to-end from surface to hub height.
The resulting climatology captures long-term variability, seasonal and diurnal cycles, and orographic flow signatures systematically absent from reanalysis-only products. We discuss the complementarity of observation-constrained and NWP-calibration approaches to national wind atlasing, the added value of CERRA over ERA5 as a background field in Alpine terrain, and the atlas as a baseline for future statistical climate-change downscaling of wind resources under CMIP6 scenarios as well as for operational near-realtime analyses.
How to cite: Schicker, I., Lexer, A., Andre, K., Janisch, S., and Bisko, N.: The Austrian Windatlas: An high-resolution Observation-Constrained, Ensemble Windspeed Analysis from Near-Surface to Turbine Hub Heights with near-realtime option, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-549, https://doi.org/10.5194/ems2026-549, 2026.
Extreme wind speeds threaten the structural integrity of wind turbines. As climate change is projected to alter global circulation patterns, it is crucial to understand its impact on extreme winds to ensure safe turbine operation. General Circulation Models (GCMs) and Regional Climate Models (RCMs) are standard tools for assessing future climate parameters. However, they often lack the temporal and spatial resolution necessary to simulate the wind speed variability at sub-daily and sub-hourly scales important for capturing extreme wind events relevant to structural safety.
Convection permitting climate models (CPMs) unlike conventional climate models represent wind spectra accurately when compared to observations, up to their temporal resolution limits (Correa-Sánchez et al., 2025) . Therefore, we utilize an ensemble of CPMs from CORDEX-Flagship Pilot Study on Convective Phenomena over Europe and the Mediterranean (CORDEX-FPS Conv) project. They have a horizontal resolution of 2-3 km and hourly temporal resolution. The CPM domain cover central-southern Europe. The 100 m wind speed from the CPM ensemble is utilised to assess changes in 50-year return period winds at a 10-minute effective temporal resolution, U5010min. U5010min is a key design parameter mandated by International Electrotechnical Commission (IEC) standards for selecting wind turbine classes (I, II, III, and S/T). The spectral correction method (Larsén et al., 2012) is used to estimate U5010min from the hourly CPM outputs.
Future CPM projections suggest large spatial variability in U5010min changes. Most parts show an increase in U5010min, necessitating a higher turbine class requirement in ~7% of the area covered by the CPM domain. While, ~1% will experience a decrease in class requirement. The increases are primarily concentrated along the Ligurian and Adriatic coasts and in areas of complex orography. Moving forward, we plan to determine if existing wind farms are located in regions where a change in turbine class is anticipated. Furthermore, we intend to investigate the added value of CPMs compared to conventional models in capturing extreme wind events.
Correa-Sánchez, N., Larsén, X. G., Fosser, G., Dallan, E., Borga, M., & Marra, F. (2025). Brief communication: Enhanced representation of the power spectra of wind speed in convection-permitting models. Wind Energy Science, 10(11), 2551–2561. https://doi.org/10.5194/wes-10-2551-2025
Larsén, X. G., Ott, S., Badger, J., Hahmann, A. N., & Mann, J. (2012). Recipes for Correcting the Impact of Effective Mesoscale Resolution on the Estimation of Extreme Winds. Journal of Applied Meteorology and Climatology, 51(3), 521–533. https://doi.org/10.1175/JAMC-D-11-090.1
How to cite: Mathew, M., Guo Larsén, X., Müller, S., and Fosser, G.: Extreme wind speed projections from km-scale climate model simulations : Implications for future wind turbine design classes, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-679, https://doi.org/10.5194/ems2026-679, 2026.
As the grid spacing of numerical weather prediction models decreases below the scale of several kilometers, modelling previously neglected spatial details and associated processes becomes more relevant. One such process is the wetting and drying of the large tidal flats which result in semi-diurnal modification of surface temperature, moisture and roughness length over the area whose width, in the specific case of the Wadden Sea tidal flat at the northwest German coast, varies from 3 to 20 km.
To explore the potential effects that including this process has on the near-coast wind field and wind power generation, we use a customized version of the Weather Research and Forecasting (WRF) model (WRF-TIDE). The modification itself was originally developed by a research team from the Kyungpook University of Daegu (South Korea), and was subsequently adapted at ForWind for the area of the German Bight. Research questions we are interested in include the tidal impacts on the long-term and large-scale patterns – such as the coastal wind speed gradients under various stability regimes – as well as transient events like the low-level jets which modify the power generation on the scale of several hours. Therefore, a representative year was simulated using both the default WRF and the modified WRF-TIDE model configuration. We analyze both the mean atmospheric fields and the specific intervals with large differences between the two WRF configurations.
While a comprehensive comparison of the wind field produced by the WRF-TIDE model against fixed LIDAR and mast-based measurements is not the focus of this contribution (although this is a major part of the future work), comparison with aircraft-based measurements show that the surface temperature variation between the periods of flood and ebb are treated in a more realistic manner, as opposed to the fixed-coastline treatment in the default WRF configuration.
How to cite: Golem, P., Cañadillas, B., Voß, A., Steinfeld, G., and Lee, Y.-H.: Assessing influence of tidal wetting and drying on near-coast wind field and wind power production via mesoscale simulations , EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-514, https://doi.org/10.5194/ems2026-514, 2026.
Abstract
Ensemble weather prediction forecasts have been promoted by meteorologists for a long time due to their additional inherent uncertainty information. Despite this advantage over deterministic weather forecasts, their application is still limited, as knowledge of including this uncertainty information in power system operations is not widely available. ProPower is a probabilistic energy market optimisation tool, developed at DLR to demonstrate the advantages of probabilistic forecasts in cost-optimal power dispatch optimisation (Schyska, 2021 and Bents et al., 2024). This study examines whether the ProPower dispatch optimisation benefits by the application of short term lidar forecasts (nowcasts) as well as the potential benefit of more frequent market clearing updates in general.
ProPower is based on a stochastic clearing approach by (Morales et al., 2014) that anticipates balancing costs due to forecast errors. The workflow in ProPower simulates an initial probabilistic market clearing (e.g. day-ahead market) followed by several probabilistic clearings (e.g. intraday market) based on forecast updates. The last step is the evaluation of balancing cost at delivery. For this study, simulations have been performed on a simplified transmission grid topology covering all Germany. ECMWF ensemble forecasts (Leutbecher and Palmer, 2007) are used as input for an initial market clearing and intraday clearings. The latest intraday forecast for the wind park Amrumbank-West is a probabilistic short term lidar power forecast. The clearing is performed 5 to 30 minutes before delivery to capture ramps in wind power caused by local wind flows that are not resolved in NWP forecasts. ERA5 reanalysis data and on-site power measurements at the wind park are used to compute balancing costs due to forecast errors. This approach aims to demonstrate the benefit of offshore lidar power forecasts on reduced power dispatch at neighbouring nodes and less congested power lines. The advantage of using lidar forecast in dispatch optimisation for a simplified network with five nodes has already been demonstrated by (Bents et al., 2025).
Acknowledgment
Lidar forecasts provided by ForWind – Center for Wind Energy Research of the Universities of Oldenburg, Hannover and Bremen.
Funded by BMWK (Windramp II, ref. No. 03EE3101).
References
Bents,H.M., von Bremen,L. and Schyska,B.U. (2024): Using weather forecast uncertainty minimises electricity costs in low flexibility power systems, 23rd Wind and Solar Integration Workshop, WIW 2024.
Bents,H.M., von Bremen,L, Schyska,B.U. and Rott,A. (2025) Evaluating the benefit of probabilistic Lidar-based wind power forecasts in power systems management. D-A-CH 2025, Meteorology Conference, 2025-06-23 - 2025-06-27, Bern, Schweiz.
Leutbecher,M., and Palmer,T.N. (2007): Ensemble forecasting, Journal of Computational Physics, 227
Morales,J.M., Zugno,M., Pineda,S., and Pinson,P. (2014): Electricity Market Clearing with
Improved Scheduling of Stochastic Production, European Journal of Operational Research, 253(3)
Schyska,B.U. (2021): Coaction of Input Parameters and Model Sensitivities in Numerical Power System Modeling PhD thesis, Oldenburg: Carl-von-Ossietzky Universität Oldenburg
How to cite: Goerlitz, A., von Bremen, L., Zech, M., and Schyska, B.: Local impact of short term lidar offshore power forecasts on grid integration, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-687, https://doi.org/10.5194/ems2026-687, 2026.
According to the International Energy Agency, global wind energy capacity is expected to double between 2024 and 2030. Offshore wind developments are a major driver of this increase, with capacity predicted to quadruple over the period. This rapid development of capacity is driven by governments in Europe and elsewhere to accelerate the transition from fossil fuels to renewable sources, in the face of intensifying climate change.
In order to maximize the economic value of these installations, research is needed to understand cost-optimized wind farm layouts and operational strategies. In this context, we present the Microwave Radiometer for the Detection and Assessment of Offshore Wind Resources (MiRadOr) project. MiRadOr assesses the value of use of Microwave Radiometer (MWR) technologies in an operational wind energy context, with a focus on atmospheric stability and offshore applications. The project combines a new months-long MWR observational dataset from the Dutch Meteorological Service’s (KNMI) atmospheric research station in Cabauw, the Netherlands, co-located with supporting observations from a 200 m tall meteorological mast and a Doppler-wind Light Detection and Ranging (LiDAR) system.
During the month of March in 2026, the MWR observations were supplemented by a period of intensive observations (IOP), featuring approximately 100 radiosonde ascent profiles, recording temperature, pressure, relative humidity, wind speed and direction. We compare these profiles against those retrieved by the MWR, LiDAR and meteorological mast over the month, including stability-relevant weather events such as cyclones and front systems. In particular, we focus on the observations of Low-level Jets (LLJs): fast moving masses of air, close to the surface (typically within the first 300 m) which are highly relevant to wind energy operations. Additionally, we compare radiosonde and other observations with forecast products from KNMI, DWD (the German Weather Service) and the ECMWF’s integrated forecast system.
How to cite: Juchem, H., Shenolikar, J., Schrepfer, J., Mu, F., Gottschall, J., Czekala, H., Cimini, D., Nilo, S. T., and Fiedler, S.: The MiRadOr Project: enhancing the assessment of offshore wind resources with microwave radiometers, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-314, https://doi.org/10.5194/ems2026-314, 2026.
The wake behind large wind farms simulated through large-eddy simulations (LES) in shallow conventionally neutral boundary layers is characterised by an asymmetric behaviour in the turbulence kinetic energy (TKE). A streak of increased TKE forms at one edge of the wake only, and it extends over several tens of kilometres.
A larger-than-expected TKE in the wind farm wake has non-negligible consequences on several aspects, from impacting the life expectancy of turbines in neighbouring wind farms to enhancing mixing of several quantities of the atmospheric boundary layer (ABL) with consequences for the local microclimate.
The primary objective of our work is to determine the physical nature of the increased TKE by means of LES. We initially identify the Coriolis force as the primary symmetry-breaking feature in the resolved physics by observing the TKE streak switching the side of the wake at which it appears when simulating a real wind farm in the German Bight in a shallow northern hemisphere ABL (NH) and in a southern hemisphere ABL (SH).
Moreover, we implement a method to simulate the same wind farm in three additional fictitious ABLs: one with identical shear profile to NH but without Ekman veer, this isolates the contribution of the Coriolis force induced by the velocity reduction in the wake; an identical ABL to NH but without Coriolis force, to isolate the effect of veer; one ABL where both veer and the Coriolis force are removed.
Thanks to such a framework, we identify the presence of veer as the necessary conditions for generating a noticeable and long-lasting asymmetry in the wind farm wake TKE. When veer is present, the wind direction inside the wind farm wake is different from that in the free stream. This causes a convergence at the side where the lateral velocity component in the free stream points towards the wind farm wake region. This effect is the most noticeable in the region above the hub height of the rotors.
Such a convergence in the upper part of the ABL induces a larger vertical shear in the main velocity component that causes a larger TKE production.
Our simulation setup allows for further qualitative comparison with satellite-based measurements of the sea-surface collected in the German Bight, showing asymmetric streaks in the measured friction velocity similar to the modelled TKE.
Our LES suggest that, in the TKE streak, downward momentum transport is enhanced. Furthermore, a similar pattern to the TKE asymmetry also appears in the friction velocity when using the Charnock parametrisation instead of a constant surface roughness at the domain bottom boundary. Thus, the asymmetric behaviour observed from the satellite could stem from the TKE asymmetry. However, more simulations with different ABLs are required to demonstrate this statement.
How to cite: Centurelli, G., Peinke, J., Djath, B., Schulz-Stellenfleth, J., and Steinfeld, G.: Asymmetric behaviour of turbulence kinetic energy in the wake of wind farms caused by the Coriolis effect, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-702, https://doi.org/10.5194/ems2026-702, 2026.
How to cite: Krutova, M., Malekmohammadi, S., Cheynet, E., and Reuder, J.: The LOLland offshore Lidar EXperiment (LOLLEX) dataset for studying wind farm flow with a ship-mounted lidar, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-477, https://doi.org/10.5194/ems2026-477, 2026.
Convective cold-pool gust fronts represent extreme inflow conditions for wind-energy systems, yet their impacts on turbine wakes and loading remain poorly quantified. This study presents a high-resolution numerical case study of a strong cold-pool gust front simulated with the WRF model coupled to a generalized actuator disk representation of a wind turbine. The event is characterized by three distinct atmospheric-boundary-layer regimes: a marginally convective pre-gust boundary layer, a highly turbulent gust-front passage, and a strongly stabilized post-gust period.
During the gust front, hub-height wind speeds more than double relative to pre-gust conditions, reaching values close to 24 m/s. This wind ramp is accompanied by rapid wind direction changes and a near-surface temperature drop of approximately 6 K. The enhanced turbulence during the gust-front passage substantially reduces the turbine wake velocity deficit and accelerates wake recovery, while the stabilized post-gust environment leads to a stronger and more persistent wake. Changes in rotor-layer shear and veer during the event would have important implications for wake structure.
Analysis of the wake structure indicates that the cold-pool passage strongly modulates the turbine wake. In the marginally convective pre-gust environment, the wake deficit exhibits strong temporal variability. During the cold-pool gust front passage, due to the fact that turbine reaches rated power and due to enhanced turbulence dramatically reducing the wake velocity deficit, there is a smaller wake in terms of horizontal extent, indicating rapid wake recovery.
The analysis of out-of-plane blade root bending moments indicates that mechanical loading responds primarily to changes in turbine operating regime rather than directly to variations in wind shear associated with the gust-front passage. These results highlight the importance of convective cold-pool dynamics for wind-turbine performance and wake behaviour.
Comparisons to extreme observed cases at the DLR operated wind park, WiValdi, located in northern Germany, are presented.
How to cite: Kilroy, G., Thayer, J., Menken, J., Hach, O., and Krimm, F.: A High-Resolution WRF Case Study of a Thunderstorm-Induced Cold Pool and Impacts on Wind Turbine Power Output, Wake Structure and Mechanical Loading, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-682, https://doi.org/10.5194/ems2026-682, 2026.
Highly renewable power systems can regularly experience system stress derived from weather variability, as weather is a driver of both electricity demand and renewable energy generation. Stress events emerge when periods of high demand coincide with periods of low renewable electricity generation. As European power systems become more interconnected, cross-border electricity exchange using these electricity interconnectors offers the potential to reduce the severity of these stress events.
This study investigates the role of interconnectors in mitigating energy drought events in a future European power system representative of 2030. To this end, the DestinEE power system model is driven by 75 years of meteorological data from the Pan-European Climate Database to construct a long-term time series of renewable generation, electricity demand, and inter-regional power flows. Periods of extreme stress events are identified in this time series, and the electricity exchanges across interconnectors are analysed during these events. Great Britain and Ireland are used as case studies due to their position at the edge of the European network, and the analysis includes regions which are connected to those two countries.
The results present that events occur during an extended winter period, from October to March, and that their duration can reach up to five days. A second part of the analysis investigates the co-occurrence of events between neighbouring regions. This reveals that simultaneous stress events are common across northwestern Europe, although their frequency varies greatly between regions.
During energy drought events, neighbouring regions can contribute to mitigating system stress through electricity imports. In well-connected areas, interconnectors offer a consistent reduction in stress, supported by multiple links to regions with diverse generation portfolios, including dispatchable and complementary renewable sources. In contrast, regions with less connectivity show a more variable response, as support depends on both the availability of neighbouring generation and the risk of simultaneous stress or transmission constraints; while most cases show a mitigating effect by interconnectors, occasional coincident events and limited transfer capacity can lead to reverse flows, temporarily reducing the overall benefit. As a result, interconnectors generally provide effective mitigation of energy droughts, although their benefits vary from one region to another.
Consistent with these results, spatial connectivity plays a key role in managing weather-driven energy risks, while also revealing clear regional differences in the benefits provided by interconnectors. In future European power systems characterised by high renewable dependence, the resilience benefits of interconnection are therefore likely to depend on both network structure and the occurrence of concurrent stress in neighbouring regions. These findings contribute to a better understanding of the vulnerabilities and resilience of interconnected power systems under climate-driven variability.
How to cite: Morin, B., Maimó Far, A., Delubac, R., Flynn, D., and Sweeney, C.: The role of interconnectors to mitigate energy droughts in a highly renewable power system, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-613, https://doi.org/10.5194/ems2026-613, 2026.
Natural gas remains the dominant source of winter heating demand in the UK, with 85–87% of homes relying on gas-powered central heating boilers. Accurate gas demand forecasting is therefore critical for ensuring supply reliability, optimising operational costs, and informing both short- and long-term infrastructure planning. In the medium and short term, temperature is the single most influential driver of demand variability relatively small forecast errors can translate into significant supply-demand imbalances and price exposure. As a result, improved weather forecasting has become a strategic priority for energy retailers seeking to manage price volatility and maintain supply adequacy. In this study, we examine the financial impact of incorporating sub-seasonal to seasonal (S2S) weather forecasts generated using a Sequential Learning Algorithm (SLA) and benchmarked against ECMWF outputs into demand forecasting and hedging activities for British Energy Markets, across lead times of weeks 3 and 4 for the winters of 2020–2025. SLA outputs are used to forecast Non-Daily Metered (NDM) demand covering residential and small business consumers using a Composite Weather Variable (CWV) that demonstrates greater skill than climatological benchmarks at weeks 3 and 4 lead times. Four demand estimation approaches are evaluated, spanning both deterministic and probabilistic SLA frameworks, to assess their relative value for hedging decision-making and financial risk management. The net cost associated with each approach is quantified by comparing gas procurement contracts across daily, weekly, and monthly time horizons. Thus, this study provides evidence that S2S weather forecasts, generated through a Sequential Learning Algorithm, offer measurable financial value for gas demand forecasting and hedging decision-making at sub-seasonal lead times with implications for how energy retailers manage supply risk and procurement costs during winter periods.
How to cite: Roy, R., Mammatt, J., Fox, S., Brayshaw, D., Das, A., Sharma, S., O’Reilly, C., and Frame, T.: Evaluating the financial value and decision-making benefits of sequential learning algorithms (SLA) for medium range gas demand forecasting and hedging during UK winter periods: Case study for British Energy Markets , EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-264, https://doi.org/10.5194/ems2026-264, 2026.
Test Reference Years (TRY) are a high-resolution temporal and spatial climate dataset that reflects the average meteorological weather pattern of a year in Germany. This climate dataset represents the expected character of the respective climate period at the respective location and, with its diurnal and seasonal variations, simulates the typical synoptic pattern.
The TRY are continuously being further developed. The dataset encompasses both the current and future climate in Germany. For the current climate, a period from 2006 to 2024 has been defined. Using an advanced method, individual segments are selected to represent the TRY. The latest version focuses on improved representation of local climatic effects (e.g., the UHI effect) and greater spatial and temporal continuity. In addition to a grid-based approach, there will also be a variant based on natural climate regions.
The further development of the TRY data is being carried out as part of a project at the German Meteorological Service (DWD) on behalf of the BBSR (Federal Institute for Research on Building, Urban Affairs and Spatial Development) and is intended to build upon the existing developments of the dataset. The aim is to further develop existing methods and integrate new data sources and concepts into the final product.
The Test Reference Years serve as a basis for thermal building simulation and for the design calculations of heating, ventilation, and air conditioning systems. They are also used to estimate the energy consumption for building climate control. TRY data is an important component for both the planning of new buildings and the renovation of existing ones.
How to cite: Spitzer, A., Dostal, P., and Scholz, O.: TRY - Further development of the test reference years in Germany, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-497, https://doi.org/10.5194/ems2026-497, 2026.
Tropical regions exhibit high solar potential, making them well suited for photovoltaic deployment. Moreover, in non-interconnected zones, such as islands, solar power generation can contribute substantially to electricity supply. In this context, accurate intra-hour forecasts of surface solar irradiance facilitate the operation of photovoltaic plants and their smooth integration into the power grid. However, short-term (0-1h) irradiance forecasting remains particularly challenging on tropical islands due to the high variability of cloud cover.
Cloud cover forecasting techniques based on satellite observations are of great value for ground-level solar irradiance forecasting. Among them, optical flow algorithms for deriving Cloud Motion Vectors (CMVs) are now considered a standard in the field. Recently, deep learning approaches, especially U-Net architectures, have shown strong promise in improving the short-term anticipation of cloud positions. However, the extent to which these models can enhance subsequent irradiance forecasting in tropical environments remains to be investigated.
This study introduces a lightweight U-Net-based model designed to simultaneously forecast four future clear-sky index (Kc) maps, with lead times of up to one hour. The model takes four past Kc maps as input, which are processed from Meteosat-9 visible channels (Indian Ocean Data Coverage, 45.5°E). First, distinct models are trained using 2.5 years of satellite data. While sharing the same architecture and training procedure, each model is tailored to a specific territory in the South-West Indian Ocean through dedicated training data. Then, independent evaluations are performed on a separate year. Beyond the assessment of forecast Kc map quality, surface global horizontal irradiance (GHI) is derived from model outputs at specific station locations. These values are then compared with ground-based GHI measurements from the Indian Ocean Solar Network (IOS-net).
The results demonstrate that our U-Net-based models improve spatial Kc map forecasts by approximately 10-15% over standard CMV methods, in terms of mean absolute error (MAE) and root mean square error (RMSE). Consequently, local GHI forecasts are typically enhanced by 5-10% across all territories tested. In addition, the models exhibit promising cross-island generalisation, suggesting that they capture robust location-invariant patterns. These findings point to the potential for developing a unified model for tropical islands, as well as for leveraging transfer learning approaches, which could simplify the practical deployment of such models from an operational perspective. Overall, this study establishes U-Net-based models as valuable candidates for improving intra-hour GHI forecasts over tropical islands in the South-West Indian Ocean, with the potential to further increase photovoltaic penetration in the local energy mixes.
How to cite: Caron, C., Voitus, A., Adrien, B., Turpin, M., Sébastien, N., and Schmutz, N.: Satellite-Based Clear-Sky Index Nowcasting with U-Nets to Improve Solar Irradiance Forecasting in the Indian Ocean, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-148, https://doi.org/10.5194/ems2026-148, 2026.
We introduce a novel framework that captures spatiotemporal dependencies for intraday photovoltaic (PV) power forecasting and employ it to systematically assess seven nowcasting models. The evaluated approaches cover a broad spectrum, ranging from satellite-driven deep learning models and optical-flow techniques to physics-based numerical weather prediction systems, and include both deterministic and probabilistic formulations. Their performance is examined with respect to accuracy, reliability, and forecast sharpness. The evaluation is conducted in two stages. First, forecast skill is assessed at the irradiance level using satellite-derived surface solar irradiance fields as reference. These irradiance predictions are then translated into PV power estimates through a station-specific machine learning model, which incorporates local irradiance together with solar azimuth and elevation angles as input features. This setup enables a consistent conversion from irradiance forecasts to power forecasts, which are subsequently validated against observations from 6434 PV systems distributed across Switzerland. To the best of our knowledge, this study constitutes the first nationwide assessment of spatiotemporal PV power forecasting. We further introduce new visualization techniques that reveal the impact of mesoscale cloud dynamics on PV generation at hourly and sub-hourly timescales. Our results show that satellite-based approaches outperform the Integrated Forecast System ensemble (IFS-ENS) at short lead times, although their skill decreases more rapidly with increasing forecast horizon. Among the evaluated methods, SolarSTEPS and SHADECast achieve the highest overall accuracy for both irradiance and PV power, with SHADECast also demonstrating the most reliable ensemble spread. While the deterministic IrradianceNet model yields the lowest root mean square error, probabilistic forecasts from SolarSTEPS and SHADECast provide better-calibrated uncertainty estimates. Forecast performance is found to deteriorate with increasing elevation, and conditions characterized by cloudiness and high variability remain particularly challenging. At the national scale, satellite-driven approaches accurately capture daily aggregate PV generation, achieving relative deviations under 10% on 82% of days over the 2019–2020 period. This demonstrates their strong robustness and potential for use in operational settings.
How to cite: Lanzilao, L. and Meyer, A.: Benchmarking satellite-based and numerical weather prediction models for national-scale intraday PV power forecasting, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-96, https://doi.org/10.5194/ems2026-96, 2026.
The Copernicus Atmosphere Monitoring Service (CAMS) operated by the European Center for Medium Range Forecast (ECMWF) provides open-data access to surface solar irradiance (SSI) through its CAMS Radiation Service (CRS). In CRS, observations from geostationary meteorological satellites are combined with modeled data for aerosols, water vapor, and ozone from the ECMWF CAMS Integrated Forecasting System (IFS) to derive irradiance at the Earth's surface. CRS data is operationally provided for the Meteosat satellite field of view (Europe and Africa) and the HIMAWARI satellite field of view (Asia and Australia). Time series in 1 min, 15 min, hourly and daily temporal resolution are produced 'on-the-fly' on user request at the desired location. In the expert mode, all input parameters e.g. aerosols, total column water vapour and ozone, cloud optical properties, and surface albedo can also be accessed as additional time series.
Recently a probabilistic error model was developed by the CAMS team (Lezaca et. al, 2025), which is designed to provide uncertainty information on the CRS irradiance estimates. For this model, an extensive database of quality controlled irradiance ground observations was exploited to condition the cumulative distribution function on the CRS estimates deviations (CRS estimates - observations) of selected CRS model inputs. By applying this uncertainty model on CRS user requests for solar irradiance time series, a complete error probability distribution can be delivered on every location and time step requested.
We have started to explore ways in which this uncertainty information should be presented to the CRS users for a straight forward/efficient implementation into typical processes and applications. In this work, some propositions for the metrics to be given alongside the CRS irradiance estimates will be shown, with the objective of an extensive and critical scientific exchange with CRS users, in order to fulfill as much as possible the solar energy community needs. We will also show typical applications in which the value of this new uncertainty information is demonstrated and quantified. We investigate the use in typical energy system applications (e.g. PV plant production and feed in, energy dispatch, or instrument calibration) and the interaction or implementation with well established commercial PV modeling software (e.g., PVsyst, PVSol, or PVlib) will be tested.
How to cite: Schroedter-Homscheidt, M., Lezaca Galeano, J. E., Saint-Drenan, Y.-M., and Roy, A.: CAMS radiation service news – implementation of uncertainty information to the irradiance estimates , EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-533, https://doi.org/10.5194/ems2026-533, 2026.
In the context of climate change, there is an urgent need to develop sustainable and reliable energy systems, particularly renewable energy.
For the planning and operation of such systems, accurate and trustworthy forecasting is essential. This study addresses prediction uncertainty in machine learning-based solar forecasting by quantifying it and systematically analysing its sources.
The uncertainty of a prediction arises from various sources. On the one hand, models may fail to represent the underlying relationship or may not be optimally parametrised. This type of uncertainty is referred to as epistemic uncertainty. On the other hand, uncertainties arise from the data itself, for example due to measurement errors and missing information, or from inherent variability in the physical system. This is also referred to as aleatoric uncertainty. In order to reduce uncertainty, its source must be identified.
In this study, a solar forecasting model is employed and combined with uncertainty quantification (UQ) methods to assess the reliability of its predictions.
A network of measurement stations across Germany provides irradiance time series data, which are modelled within a graph framework. The model produces short-term forecasts of solar energy, using the data from all stations and potentially additional meteorological input. It is based on Graph Neural Networks (GNNs) to incorporate both temporal and spatial information in the data.
Different UQ frameworks are integrated with the forecasting model to assess predictive uncertainty and analyse its underlying sources. Particular emphasis is placed on uncertainty arising from the input data. Classical UQ approaches are compared with perspectives from trustworthy AI, notably adversarial machine learning. In this context, small perturbations of the input are identified that induce significant changes in the model output. This approach is explored as a tool to characterise model sensitivity, and to better understand the influence of input uncertainty on the forecast, thereby complementing established UQ methods.
How to cite: Horstmann, S., Vercauteren, N., Quinting, J., Bojchevski, A., Crewell, S., and Akhondzadeh, S.: Uncertainty Quantification and Analysis in Solar Forecasting Using Machine Learning, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-585, https://doi.org/10.5194/ems2026-585, 2026.
Accurate short-term solar energy forecasting is essential for the safe and stable integration of a growing share of photovoltaic electricity generation into power grids. Intra-day forecasts up to a 6h prediction horizon are essential for grid management, trading on the electricity spot market and energy storage exploitation. Estimate irradiance using images from geostationary meteorological satellites is particularly appropriate for intraday projections, giving better performances than numerical weather prediction models.
The deployment of third-generation satellites provides unprecedented spatial and temporal resolution by delivering images every 10 minutes on a 500 meter grid at nadir. This high-frequency data offers a significant opportunity to improve short-term irradiance nowcasting via precise cloud displacement analyses. However, standard methods based on cloud motion vectors (CMV) computation do not inherently benefit from this finer resolution. Instead, they face increased sensitivity to parallax and cloud shadow effects alongside difficulties in extracting global motion tendencies. Recently, deep-learning has been extensively applied to satellite based forecasting tasks. ConvLSTM architectures, combining temporal recurrence with spatial convolutions, proved effective in reproducing the evolution of complex cloud structures.
This study explores novel forecasting approaches using Meteosat Third Generation (MTG) imager data. Firstly, a CMV based model has been improved by applying a parallax and cloud shadow correction to the input satellite images. Secondly, a ConvLSTM model has been set-up on MTG data to take advantage of the finer spatio-temporal resolution.
High-resolution images from the year 2025 captured every 10 minutes by the visible narrow channel centered on a 0.6μm wavelength of the MTG sensor have been used to produce CMV forecasts as well as train and test the deep learning model. Results accuracy is assessed against ground based measurements from several Baseline Station Radiation Network (BSRN) pyranometers located across Europe. Cloud top height data obtained from an operational version of the SAF-NWC/GEO software are used to correct parallax and shadow effects in addition to training the deep learning model.
Our results demonstrate that the improved CMV based model yields competitive accuracy scores while unlocking the benefits of high-frequency data, particularly for predicting ramp events and rapid fluctuations. The ConvLSTM approach captured sophisticated spatio-temporal patterns inherent to high-resolution imagery on numerous case studies. These findings suggest that, while finer data introduces more complexity, dedicated correction and non-linear modeling are key to improving solar forecasts.
How to cite: Duchemin, V., Chea, N., Cros, S., and Badosa, J.: Challenges for improving solar nowcasting using Meteosat Third-Generation satellite imagery, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-690, https://doi.org/10.5194/ems2026-690, 2026.
The integration of solar photovoltaics at scale requires high-frequency irradiance data to quantify minute-scale variability for grid integration, storage sizing, and ramp-rate management. However, satellite-based solar irradiance estimates face a fundamental scale mismatch. Ground-based pyranometers provide accurate high-frequency measurements but are spatially sparse and observe point-scale transients that differ from the area-averaged irradiance a utility-scale plant experiences. Satellite products such as the Copernicus Atmosphere Monitoring Service (CAMS) offer broad spatial coverage but lack temporal resolution (15 min), smoothing away the high-frequency fluctuations driven by cloud passage. Because many distinct sub-pixel cloud configurations can produce the same coarse-resolution mean, recovering minute-scale temporal trajectories from satellite inputs is inherently ill-posed and requires probabilistic treatment.
To address this ambiguity, we present a physics-informed generative framework. We derive a temporal degradation operator grounded in Taylor's frozen turbulence hypothesis, which recasts the spatial averaging of a satellite footprint as a temporal Gaussian convolution — a point spread function (PSF) representing the spatial response of the sensor. To invert this operator, we train a conditional denoising diffusion probabilistic model (DDPM) that generates ensembles of minute-resolution irradiance trajectories conditioned on coarse satellite inputs. A spectral regularization term encourages the generated sequences to follow the power-law energy cascade observed in cloud-driven irradiance fluctuations, preserving realistic high-frequency structure while suppressing spurious artifacts.
The framework is evaluated through a sim-to-real domain transfer experiment: the model is trained on synthetic pairs built from BSRN measurements using the physical forward operator and then applied directly to operational CAMS data, isolating and quantifying the representativity gap between idealized and real-world degradation. The added value of the proposed approach is evaluated for selected operational applications. Results show that the generated sequences reproduce high-frequency variability consistent with ground observations, including the capture of significant ramp events, recovery of curtailed irradiance peaks relevant to inverter clipping estimation, and improved characterization of storage requirements relative to temporally smoothed operational inputs.
How to cite: Gomez, J., Saint-Drenan, Y.-M., Yehia Eissa, Y., and Blanc, P.: A Physics-Informed Diffusion Model for Surface Solar Irradiance Temporal Downscaling, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-752, https://doi.org/10.5194/ems2026-752, 2026.
Achieving Germany's national climate targets requires efforts to reduce emissions in all sectors. The German Ministry for Transport (BMV) established the ‘Network of Research’ (BMV-Forschungsnetzwerk), a network of German government agencies focused on the future-oriented transformation of transport in Germany. The focus of the main topic “Renewable energies” is the assessment of renewable energy potential along the transportation infrastructure. In this main topic, Germany’s national meteorological service DWD (Deutscher Wetterdienst), the Federal Highway and Transport Research Institute (BASt) and the German Centre for Rail Traffic Research at the Federal Railway Authority (DZSF/EBA) work closely together.
The transport infrastructure offers significant potential for renewable energy production. To optimize energy management, it is crucial to analyze the variability of renewable energy in Germany. In a previous study, weather situations which can possibly influence the energy production of wind, solar and water power plants were collected. Additionally, events with possible influence on the electricity transmission were taken into account. Afterwards, a simple assessment of the frequency of occurrence was carried out.
The current analysis focuses on solar energy production units since those are the most commonly used to generate renewable energy along transport infrastructure. The goal is to assess how frequent the occurrence of events like thunderstorms with hail, storms or dust effect the energy production.
Different data sets from DWD will be used to quantify the occurrence of those events for Germany overall as well as for exemplary locations with solar power plants near the transport infrastructure. Further, the occurrence of influential or destructive events will be quantified. The ratio between general occurrence and influential occurrence will then be used in a subsequent study considering climate change and the shifted occurrences of previously analyzed events.
How to cite: Dost, V., Drücke, J., and Deutschländer, T.: Assessing the impact of extraordinary weather events on solar energy production units in Germany, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-151, https://doi.org/10.5194/ems2026-151, 2026.
The rapid growth of solar energy rises challenges due to its inherent intermittency, significantly driven by cloud dynamics. While approaches based on satellite images and cloud motion vectors (CMV) are considered as a reference for intra-day forecasting, they do not account for the vertical structure of the atmosphere. Such approaches treat clouds as a single two-dimensional layer, leading to significant errors during events where multiple cloud layers move at different altitudes and speeds. Numerical Weather Prediction models, on the other hand, account for the three-dimensional structure of clouds, but their spatial and temporal resolution limits their performance.
In parallel, recent research has leveraged Deep Learning (DL) architectures for forecasting solar irradiance from satellite images. For instance, Convolutional Neural Networks, including the UNet-based approach [1], have been used to forecast satellite images. While such approaches yield promising results and can theoretically capture cloud formation and dissipation mechanics, their accuracy remains constrained when relying solely on satellite images without supplementary meteorological data, notably the vertical structure of clouds.
In this work, we propose a layer segmentation approach to improve the physical consistency of cloud motion estimation. Motivated by the distinct behaviors of driving parameters across different altitudes, we reconstruct three atmospheric layers from consecutive satellite observations (kc). This methodology follows the direction of recent studies [2], in which cloud phases were separated and overlapping CMV were calculated using cloud properties and infrared temperatures.
Our approach leverage the apparent motion of clouds on image sequences for estimating the layered cloud motion vectors. The proposed method relies on a UNet-based architecture designed for the joint task of layer semgentation and layered cloud motion vector estimation. A sequence of four consecutive satellite images along with meteorological data, including kc forecasts and wind vectors at different pressure levels, is taken as input. From these variables, three distinct kc maps, each corresponding to a specific atmospheric layer (low, medium, and high altitude), are generated. The training process is supervised by comparing the reconstructed global kc against the ground truth from the subsequent satellite observation. This approach ensures that while the model learns the complex dynamics of individual layers, it remain constrained by satellite observation.
REFERENCES
[1] Nils Straub, Steffen Karalus, Wiebke Herzberg, and Elke Lorenz. Satellite-based solar irradiance forecasting: Replacing cloud motion vectors by deep learning. Solar RRL, 8(24):2400475, 2024. doi: https://doi.org/10.1002/solr. 202400475.
[2] Cuiping Liu, Wei Han, Feng Zhang, Jiaqi Jin, Qiong Wu, Wenwen Li, and Chloe Yuchao Gao. Deriving overlapped cloud motion vectors based on geostationary satellite and its application on monitoring typhoon mulan. Geophysical Research Letters, 52(13):e2025GL116397, 2025. doi: https://doi.org/10.1029/2025GL116397.
How to cite: Banes, C., Saint-Drenan, Y.-M., and Voyant, C.: A Supervised Approach for Estimating Layered Cloud Motion Vector and Clearsky Indices from Satellite Images, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-153, https://doi.org/10.5194/ems2026-153, 2026.
The daily stability of solar irradiance critically influences photovoltaic (PV) power generation, yet existing metrics for assessing such stability often fail to establish a robust and consistent correlation with PV output. Conventional metrics typically rely on simple statistical summaries of solar irradiance time series, limiting their generalizability across diverse climatic regimes and temporal scales. To address this gap, we introduce a novel metric—the Solar Instability Index (SII)—formulated by applying the Wasserstein distance to quantify the deviation of intra-day solar irradiance patterns from the idealized diurnal cycle. By leveraging the optimal transport framework, SII provides a distribution-aware measure of irradiance variability that is physically interpretable. At our case station, SII exhibits strong correlations with atmospheric moisture content and available solar energy resources, suggesting its close linkage to synoptic weather events that drive solar resource variability. Through two case studies, we evaluate the effectiveness of SII alongside two existing metrics commonly used in solar resource assessment. The results demonstrate that SII effectively captures low-frequency variations in solar irradiance without relying on arbitrarily assigned thresholds or window sizes, thereby achieving a robust correlation with PV power output compared to the established metrics. Furthermore, SII maintains consistent performance under varying sky conditions, highlighting its adaptability across different operational contexts. As such, SII offers a valuable and practical tool for assessing the potential stability of daily PV power generation. It holds particular promise for applications such as site selection for PV power plants, and for solar irradiance forecasting aimed at optimizing PV output utilization and grid integration.
How to cite: Tian, Q.: A Novel Metric for Quantifying Solar Irradiance Stability: Mapping Solar Irradiance Variability to Photovoltaic Power Generation, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-190, https://doi.org/10.5194/ems2026-190, 2026.
Climate data is essential for simulating, planning, and operating energy systems, particularly with the increasing penetration of renewable energy sources and the need for climate resilience. The countries around the Mediterranean Sea - a global climate change hotspot - face a critical challenge. The energy demand is expected to increase, creating tensions on energy security and environmental impacts.
In the frame of the ReEnergy-MED project supported by the Copernicus Joint Services, high-quality current data (both reanalysis and in-situ) along with downscaled future climate projections have been used for solar energy production simulations for two demonstration cases: Ouarzazate in South-Central Morocco and Evros in Northern Greece. The ERA5-Land reanalysis datasets served as a basis for the present climate analysis. For long-term projections - near (2031-2060) and distant future (2071-2100) - three CMIP6 GCM models were employed under three SSP-RCP emission scenarios. The long-term climate projections were statistically downscaled using the ERA5-Land as a reference dataset, in order to provide finer-resolution climate information. The generated energy production simulations, conducted with the Global Solar Energy Estimator (GSEE), were validated against in-situ observations provided by the energy producers collaborating on this project. Additionally, to enhance the collaborative development approach, two climate indices relevant to solar panel efficiency, as recommended by the energy producers, were computed for the long-term periods. Utilising appropriate climate data ensures that the energy simulations accurately reflect changing climate conditions and, in turn, the integration of the projected long-term changes in the solar energy potential will contribute to national and regional planning and a sustainable future.
Funding: This work has received funding from the European Center for Medium Range Weather Forecasts under framework agreement ECMWF/COPERNICUS/2024/CJS_155b_NOA-A for the provision of demonstration cases to support renewable energy transition across the Mediterranean
How to cite: Gratsea, M., Fountoulakis, I., Papadimitriou, N., Varotsos, K., Benbrahim, S., Papageorgiou, A., and Giannakopoulos, C.: Long-term solar energy production simulations and climate indices for two demonstration cases in Morocco and Greece - results from the ReEnergy-MED project, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-278, https://doi.org/10.5194/ems2026-278, 2026.
Mesoscale atmospheric models are widely used to provide long term time series of winds at turbine heights. Mesoscale models are known to poorly estimate diurnal cycles in wind speed. For example, the NEWA model dataset on average overestimates the magnitude of the diurnal cycle of the wind speed by 1.4 m/s. Diurnal errors in modelling wind speed can significantly affect estimates of annual energy production, even if the overall mean wind speed is correctly modelled.
Previous WRF model wind speed sensitivity studies have usually focused on overall performance metrics such as mean bias and correlation, while rarely targeting diurnal cycles of the wind speed explicitly. This work compares the performance of different WRF model setups in predicting the diurnal cycle of wind speed. Our goal is to identify the WRF setup that minimises the diurnal error in modelled wind speeds. We test three surface-layer schemes, three radiation schemes, and 15 planetary boundary layer (PBL) schemes.
For each of the WRF setups, we perform a 1-year-long run over Sweden. We compare modelled diurnal and seasonal cycles of the wind speed against observed ones from EMD’s internal mast database containing data from 62 observation campaigns at wind turbine heights. In addition, different model outputs are compared to each other, to better investigate the effects of different parametrisations on the modelled wind speeds.
Our results show that the choice of the PBL and surface layer schemes has a significant impact on modelled wind speeds during the night, while having little effect on modelled winds during the day. On the other hand, the choice of the radiation scheme mostly affects modelled wind speeds during the day. Results show that the better-performing setup for modelling diurnal wind cycles at 100 m is a combination of the MYJ PBL scheme, the Eta similarity surface layer scheme, and the RRTMG radiation scheme. Nevertheless, the best setup still overestimates the magnitude of the diurnal cycle of wind speed by 0.4 m/s.
Our results show that the choice of WRF parametrisation can significantly reduce the diurnal bias in mesoscale models. However, further research is needed to resolve the diurnal bias completely.
How to cite: Pogumirskis, M., Sīle, T., Svenningsen, L., and Hahmann, A. N.: Investigating the optimal WRF model setup for modelling diurnal cycles of wind speed at turbine heights in Sweden, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-289, https://doi.org/10.5194/ems2026-289, 2026.
Wind energy has become a significant component of the French electricity mix over the past decade currently representing around 10 % of national electricity consumption. This part is expected to significantly increase in the context of the national energy transition strategy with a target of 45 GW for offshore wind energy by 2050. Achieving these objectives requires improved understanding of the spatio-remporal variability of wind resources, assessment and improvement of current wind forecast capabilities, assistance in managing supply-demand balance and evaluating the impact of climate change on future wind resources. From research to operational product development, Météo-France actively contributes to these efforts through several activities. This poster aims at presenting an overview of current research activities and operational product development to assist the wind energy sector.
For offshore risk assessment, the contribution of Météo-France to climatological studies and analyses of wind Doppler measurements collected over 5 marine areas between 2020 and 2026 will be presented. Météo-France also contributes to a better understanding of the impact of wind farms on local meteorology. To that end, the Large-Eddy Simulation model Meso-NH has been coupled with a wind turbine model based on the Rotative Actuator Disk approach. The ability of this coupled model to capture wind turbine wake deficits and their interactions with complex meteorological conditions and terrain will be illustrated. These developments could lead in the future to improved wind forecasts within the convective-scale model AROME and, in the end, wind power production forecasts. In that sense, the Météo-France physically-based wind power production forecast model will be presented. Its ability to produce realistic wind power production estimations and how this tool can be improved will be discussed. More particularly, taking advantage of a user-oriented classification of the AROME ensemble forecasts in order to define main and secondary scenarii of electricity production will be presented.
In order to assist the wind energy sector clients in managing supply-demand balance, sub-hour variability of wind will be investigated (in the AROME model compared to wind observations) and Météo-France product to monitor wind energy shortage will be presented.
Finally, Météo-France coordinates the Cerra4Impact project, funded by the Copernicus Climate Change Service (C3S), with the aim of developing climate change indicators for wind energy based on regional climate projections at the supranational scale. The objectives of the project and future plans will be presented.
How to cite: Martinet, P., Güell, M., Bommier, A., Boumendil, P., Caumont, O., Guerin, N., Masson, V., Poyet, A., Raynaud, L., and Roubelat, F.: An overview of on-going activities for the development of products and services for the wind energy sector at Météo-France, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-440, https://doi.org/10.5194/ems2026-440, 2026.
To meet the greenhouse gas targets of 2050, European countries plan a large expansion of offshore wind energy in the North Sea. This expansion is expected to cover 11% of the basin's surface area. The environmental impacts of this expansion are not fully understood yet. However, it is known that wind turbines alter the local microclimate by creating wakes, regions of decreased wind speed and enhanced turbulence. Mesoscale models are a promising tool to investigate the feedback effects of wind turbines on the environment. As such, recent studies show that offshore wind farms affect environmental factors such as sea surface temperature (SST). However, these studies are not conclusive and show both warming and cooling of SST induced by wind farms.
This study seeks to clarify the environmental feedback effects on SST of already installed wind farms using observations, as observational evidence of these effects remains limited. The study is focused on North Sea wind farms with over 100 turbines and examines both buoy measurements and satellite images. Time series of at least four years from before and after the construction of wind farms are investigated using conventional statistical models. Since buoy measurements are limited to a few offshore wind farms, satellite images provide additional spatial information and coverage at farms where buoy information is absent. In a second part of the study, the observational relationships between offshore wind farms and sea surface temperature derived from buoy and satellite data are used to train machine learning models. These models are then applied to projected offshore wind farm configurations to estimate potential future impacts on SST.
Preliminary results from observations show a consistent warming of SST in the wakes of large offshore wind farms, generally on the order of 0.2–0.5°C. This is in line with previous numerical modelling studies. The feedback effect of wind farms on SST has a strong seasonal component, being more pronounced in winter. During spring and summer months, SST cooling is also observed. Furthermore, a strong correlation with atmospheric stability is found, along with interdependent site variations. In future studies, these insights are expected to inform parameterizations in (coupled wave–ocean-)atmosphere mesoscale models.
How to cite: Barbieri Dumont, G., van de Wiel, B., Meyer, A., Lanzilao, L., and Porchetta, S.: Observational evidence of offshore wind farm impacts on sea surface temperature in the North Sea, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-496, https://doi.org/10.5194/ems2026-496, 2026.
The rapid expansion of wind and solar power across Europe is reshaping the structure and operation of the electricity system, with growing emphasis on interconnection and the potential role of a European supergrid. In this context, periods of simultaneously low wind and solar availability, commonly referred to as renewable energy droughts, pose a critical challenge for system reliability and planning. Such events are often assessed at national or regional scales, but these approaches are constrained by geopolitical boundaries that may not reflect the true spatial structure and evolution of atmospherically driven renewable resource deficits.
This work presents a spatio-temporal characterization of renewable energy droughts across Europe using gridded fields, with the aim of identifying how these events emerge, persist, and propagate. The analysis is based on 70+ years of ERA5 reanalysis data over the North Atlantic–European sector and focuses on compound low-wind and low-solar conditions. Renewable energy droughts are identified using threshold-based methods applied consistently across the grid.
By adopting a gridded perspective, this study moves beyond country-based assessments to capture the physical footprint of renewable energy drought events as shaped by atmospheric variability. In particular, the analysis examines the displacement, extent, and persistence of these compound deficits, as well as their relation to large-scale atmospheric circulation. This provides a framework for assessing whether renewable energy droughts remain localized, propagate across regions, or organize into broader structures relevant to pan-European energy balancing.
Ultimately, this work aims to provide a physically grounded view of high-impact renewable energy droughts in Europe and of the atmospheric processes underlying their evolution. These insights are relevant for the design and resilience assessment of highly interconnected electricity systems, particularly in evaluating the extent to which a European supergrid may mitigate, or remain vulnerable to, widespread and persistent renewable energy droughts.
How to cite: Maimo Far, A. and Sweeney, C.: Spatio-temporal analysis of renewable energy droughts in Europe: a gridded perspective, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-511, https://doi.org/10.5194/ems2026-511, 2026.
Germany’s offshore wind capacity is expected to expand rapidly from todays about 10 GW to 70 GW by 2045. Achieving this ambitious target requires robust, high‑quality meteorological and oceanographic data to characterize site conditions and thus to reduce uncertainties in planning and operation. Under the Offshore Wind Energy Act, new offshore wind sites are tendered by the Federal Network Agency, in cooperation with the Federal Maritime and Hydrographic Agency (BSH) and with support from the German Meteorological Service (DWD).
This contribution presents an overview of the meteorological and oceanographic datasets and recent advancements supporting Germany’s offshore wind development. Each site is typically characterized by one‑year floating LiDAR and oceanographic campaigns, providing wind profiles up to a height of about 250 m, wave heights and other sea state parameters, vertical profiles of sea water currents, and a large variety of further parameters. These in-situ observations are complemented by long-term reanalysis and hindcast model datasets, notably the new ICON-DREAM-EU reanalysis developed by DWD. The comparison and evaluation of these observational and model data allows for robust assessments of the long-term meteorological and oceanographic conditions and forms the basis for the comprehensive reports generated for each site. All data, reports, and derived products are publicly accessible via BSH’s PINTA portal – https://pinta.bsh.de.
The primary stakeholders of these data and reports are offshore wind developers, who rely on accurate wind and wave statistics for their energy yield calculations, design and risk assessment impacting the overall business case. In addition, the data provide a valuable reference for the evaluation and advancement of atmospheric and oceanographic models.
Recent analyses highlight the added value of these products: (i) a reanalysis‑based identification of weather windows, that is periods when atmospheric and oceanographic conditions remain below operational limits for offshore vessels, supporting logistics and operations; (ii) generation of a multi‑year, wake‑corrected reference time series (virtual met mast) from the one‑year LiDAR campaigns, with plans to incorporate future measurements; and (iii) the development of integrated meteorological and oceanographic measurement campaigns supporting site pre‑investigations in increasingly remote areas. Selected examples of these developments will be presented.
How to cite: Michaelis, J., Hansen, A., Hansen, F., Möller, T., Spangehl, T., Külheim, F., Kelbch, A., Hüttl-Kabus, S., Brast, M., Hahn, J., Outzen, O., and Andersson, A.: Comprehensive Meteorological and Oceanographic Data for Germany's Offshore Wind Expansion - Recent Developments and Applications, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-636, https://doi.org/10.5194/ems2026-636, 2026.
The transition from fossil fuels to renewable energy sources in Europe is accelerating under climate policy commitments (Kapica et al., 2024), including the Paris Agreement (2015) and European Green deal (2019). This shift increasingly relies on weather-dependent generation from wind and solar power, while climate change is simultaneously intensifying summer heatwaves and associated cooling demand (Filahi et al., 2024). As a result, electricity systems are becoming more exposed to weather-driven variability, highlighting the need for more research in regions historically characterised by winter peak loads.
Despite the growing importance of climate–energy interactions, progress in understanding these dynamics is still limited by a disconnect between climate and energy modelling communities. Climate modellers often do not provide outputs that are directly usable for energy system applications, while energy system models tend to overlook climate related uncertainty (Craig et al., 2022). This mismatch makes it difficult to properly assess system reliability and economic stress under future climate conditions.
Some recent studies have started to bridge this gap by linking large-scale weather patterns to energy system impacts. For example, targeted circulation types have been developed to better capture weather sensitivity in electricity systems (Bloomfield et al., 2020), while other studies connect weather regimes directly to metrics such as Energy Not Served (ENS), reflecting system reliability from a grid operator perspective (Biewald et al., 2025; Wuijts et al., 2023). However, these approaches still struggle to fully account for climate uncertainty and economic signals such as price variability.
This paper proposes a data-driven framework to better connect climate variability with energy system impacts, focusing on indicators such as energy shortfall, ENS and high-price events. Data-driven methods are explored to identify climate drivers behind these energy impacts, allowing a more system-relevant characterization of climate risks. These drivers can then be used as input for AI-based forecasting approaches, improving the prediction of system stress under uncertain future conditions. Bridging this gap is essential for ensuring reliable, affordable, and resilient low-carbon energy systems in a changing climate.
How to cite: Bakels, E., Bloemendaal, N., Jaeger, W., Coumou, D., and Ward, P.: Data-driven approaches to link climate drivers to energy impacts, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-640, https://doi.org/10.5194/ems2026-640, 2026.
The amount of installed wind power capacity has grown in Finland over the last 10 years. According to statistics from Renewables Finland, there were 2,002 wind turbines installed with a total capacity of 9,433 MW by the end of 2025. Electricity statistics from Finnish Energy show that wind power amounted to 26% of the total Finnish electricity production during 2025, which makes wind power an important energy source. Since wind power production by nature is volatile due to natural variations of wind, and has a big share of total electricity production, it is important to have both accurate and reliable forecasts for wind power.
The wind power forecast model at the Finnish Meteorological Institute is based on forecast data from the MetCoOp ensemble prediction system (MEPS). MEPS is based on the limited area model Harmonie-AROME and covers Northern Europe with a horizontal resolution of 2.5 km. Every hour a five-member ensemble is run with a maximum forecast lead time of 66h. The output of MEPS is used as the input to the wind power model. The wind power model further considers the locations of the wind turbines and information about their heights and power curves when computing an aggregated wind power forecast. The wind power forecast model is computed for all ensemble members and for all forecast lead times. After this, a 30-member lagged ensemble is formed by collecting the forecast output from five members over a six-hour time window.
In this research, we study ensemble wind power forecasts for aggregated wind power in Finland and how to calibrate and post-process wind power forecasts with the aim of having accurate and reliable probabilistic forecasts. We use statistical post-processing tools, such as ensemble model output statistics (EMOS), and study how to adapt it to a lagged 30-member ensemble. As observations we use statistics on wind power production provided by Finnish Energy and Finland’s transmission system operator Fingrid. A total of two years of data is used in this study. We will present the results from our on-going study on probabilistic wind power forecasts for Finland.
How to cite: Ekblom, M., Atlaskin, E., Laine, M., and Lindfors, A.: Calibration of MetCoOp-based ensemble forecasts for aggregated wind power in Finland, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-350, https://doi.org/10.5194/ems2026-350, 2026.
With the increasing share of renewables in the European grid, predicting time series for wind speed and direction is becoming vital for accurate wind resource assessment. During periods of strong winds, energy prices are low, meaning one must account for the fact that the energy produced is less "valuable." Traditionally, Annual Energy Production (AEP) is calculated using Weibull distributions and a power curve; however, this approach ignores the impact of varying energy prices.
To further complicate matters, the onshore wind resource is spatially highly heterogeneous. Numerical models capable of predicting flow at microscale resolution are computationally expensive. Therefore, we introduce a method to statistically downscale mesoscale time series to the microscale (50 m grid spacing).
The method is based on wind-speed-independent roughness and orographic speedup factors, and a statistical representation of stability effects. The stability model accounts for both the mean and variance of the wind distribution as a function of height and surface roughness. By assuming that both mesoscale and microscale wind speed distributions are Weibull distributed, the mesoscale time series are transformed. We test several approaches for this transformation using mesoscale time series from the New European Wind Atlas (3 km resolution) and downscale them to 50 m resolution. For this downscaling we use high-resolution landcover (CORINE) and elevation data (Copernicus DEM).
Finally, we validate this approach using a database of meteorological masts and wind lidars across Europe. These measurements demonstrate that microscale variability is significant, and accounting for it can substantially improve wind resource estimations derived from mesoscale models. Future improvements to this workflow are also discussed.
How to cite: Floors, R. and Larsén, X. G.: Statistical downscaling of mesoscale wind time series for microscale resource assessment, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-717, https://doi.org/10.5194/ems2026-717, 2026.
Supporters & sponsors
