Temporally compounding energy droughts in European electricity systems with hydropower

Temporally compounding energy droughts in European electricity systems with hydropower | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Temporally compounding energy droughts in European electricity systems with hydropower Lieke van der Most, Karin van der Wiel, Winnie Gerbens-Leenes, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-3796061/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 27 Sep, 2024 Read the published version in Nature Energy → Version 1 posted You are reading this latest preprint version Abstract As Europe's renewable energy capacities expand, electricity systems face increased risks of energy droughts—periods of low production coinciding with high demand. We evaluate characteristics of electricity variability due to weather variations by calculating 1600 years of daily production and demand. Focusing on six European countries—chosen for their energy mix including hydropower—we find that energy droughts result from processes that cause (temporally) compounding impacts in the energy and meteorological system. These can turn what might have been short-term droughts into prolonged, cumulative energy crises. For instance, low reservoir inflows in spring quadruple the chance of prolonged energy droughts: reduced snowpack and rainfall lower hydro-availability but also dry-out subsoils, increasing the chance of heatwaves and therewith extending the energy problems into summer. We identify and quantify three compounding energy/climate conditions and the associated characteristics and risks of multiyear energy droughts, crucial for future energy system design and policies. Physical sciences/Energy science and technology/Energy modelling Physical sciences/Energy science and technology/Renewable energy Earth and environmental sciences/Climate sciences Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Introduction As Europe’s capacity of renewable electricity sources grows, the region's electricity system is becoming increasingly dependent on variations in weather and climate 1,2 . Electric heating and cooling demand are largely defined by air temperatures and the production of renewables depends on the availability of resources (wind speed, irradiance, discharge). With this increased sensitivity to climate variations arises the risk of encountering weather conditions outside of the system’s design scope, undermining energy security. Furthermore, due to non-linear relationships between climate factors and the electricity system, extreme impact events in the form of high demand and low renewable electricity production (hereafter referred to as energy droughts) might occur under compounding meteorological conditions that in itself are not considered extreme 3 , and vice versa. These so-called compounding events 4 have recently had significant impact in electricity systems in a number of European regions. During the autumn of 2021, Europe experienced record-low gas storage levels, a situation influenced by several factors. Besides market prices and geopolitical causes, other contributing factors were a preceding cold winter in Northern Europe followed by another colder-than-usual season in spring, compounded by very low annual average wind speeds in parts of northwestern and central Europe 5 . In September 2021, this prolonged period of calm weather conditions even led to record-high electricity prices in the UK energy market 6 . Furthermore, in the beginning of 2022, Spain and Italy saw a dramatic 40% decline in hydroelectricity generation 7,8 . This sharp drop was caused by an exceptionally hot summer and one of the driest years for water resources since record-keeping began 9 . Consequently, some hydropower plants even had to cease operations in the midst of a cooling demand peak 10 . Hydropower is Europe’s main source of renewable electricity to date—with an installed capacity of 188 GW it makes up for 36% of the renewable capacity, and supplies 17% of the total EU electricity demand 11 . In seasons of high water availability, like spring and early summer, water is accumulated in reservoirs and released during periods of peak demand, such as winter. This seasonal and dispatchable nature of hydropower potentially makes an electricity system vulnerable to prolonged and temporally compounding energy droughts. However, previous studies that modelled extreme energy drought events in Europe’s renewable electricity system have focused on wind power and solar PV generation 12–18 or only included run of river hydropower production 19 , overlooking the critical role of reservoir storage and its operational dynamics. One of the reasons hydropower reservoir production is often excluded from studies into the weather sensitivity of electricity systems is that it is characterized by complex operating procedures 20 . Energy system models can be employed to simulate this, but they are typically computationally intensive, leading to reliance on shorter datasets or estimates of annual energy potential 21 . Recent work applied novel approaches to allow for the handling of an ensemble of single weather years 22 . However, hydropower not only experiences significant variations from year to year 23 —for instance, changes in the North Atlantic Oscillation (NAO) index can explain up to 30% of the fluctuations in Norway's hydropower production 24,25 —but might also be susceptible to sustained climate phenomena such as multi-year meteorological or hydrological droughts 23,26,27 . To accurately represent this variability, extended time series data is necessary. The climate and energy modelling community suggest identifying meteorological conditions that have extreme impacts in the electricity system and integrating them into energy system simulations to tackle the problem of handling such large datasets 21 . Here we identify meteorological conditions that (temporally) compound to have extreme impacts on electricity systems including hydropower using climate model data and a daily renewable electricity production and demand modelling framework 28 . We employ a large ensemble approach, using 160 sets of 10-year climate model simulations (1600 years) 29 , each set representing a different possible sequence of weather under present-day climate conditions. We present case-studies for the six European countries with the highest installed hydropower capacities (Norway, France, Italy, Spain, Switzerland and Sweden) and evaluate daily residual loads —demand minus renewable energy generation—for the full 1600 years of climate data to: 1) identify and quantify different types of energy droughts in the electricity system, 2) analyze temporal compounding in these events, and 3) quantify risk ratios of meteorological conditions leading to extreme impact events. High residual load events in electricity systems with hydropower We classify high residual load events using two metrics. First, we introduce the concept of 'Energy Drought Windows (EDWs)' 13 , which are fixed-duration periods for which we select the highest residual loads from the ensemble. Secondly, we examine 'Persistent Energy Droughts (PEDs)' 19,30 . These drought events occur when the daily residual load remains consistently above a predefined threshold for consecutive days. Additionally, we differentiate between summer, winter, and season independent events by defining PED thresholds relative to the respective seasonal residual loads. We find that electricity systems with relatively large hydropower reservoirs experience prolonged PEDs (Fig. 1 a). In Norway, Sweden, and Switzerland, PEDs can last more than 150 days. These countries use their reservoir storage to release stored water strategically during brief periods of extreme meteorological conditions characterized by heightened demand and/or reduced renewable electricity production, mitigating short term energy shortages. However, once the reservoir depletion exceeds a critical threshold and hydropower resources are exhausted, these systems may face very long energy droughts. In the present-day European electricity systems, energy droughts predominantly occur during the winter months (Fig. 1 b-g). These can be attributed to factors such as strong heating demand and reduced wind power production (a result of the greater variability in wind speeds during the winter compared to the summer, which enhances the chance of periods with low wind 13 ), and diminished reservoir storage levels. Regions in more northerly latitudes generally experience these events later in the year. This can be explained by two factors: the coldest periods occur later, and relatively large reservoir storages tend to deplete later in the year. All-year PEDs driven by increased cooling demands during hot summer days only occur in Southern European countries like Spain and Italy (Figs. 1 d and 1 g) 31 . These PEDs can be intensified by reduced water availability affecting hydropower production. Since electricity demand response to high temperatures is small in Northern European countries, summer events typically transpire during early or late summer when temperatures are unseasonably low, leading to increased heating requirements. The barplots in Fig. 1 indicate that the mean residual load during these summer events in Northern countries is not significantly higher than the average yearly residual load, meaning that, in an absolute sense, summer PEDs are not extreme. Compounding conditions during energy droughts For the analysis of the meteorological conditions during events we examine energy drought windows of 30 days. Figure 2 shows the meteorological conditions during these EDWs and the reservoir levels leading up to them, using Norway, France and Italy as illustrative examples (see Supplementary Figs. 1 and 2 for other countries). We find that prolonged high-pressure systems or ridges and atmospheric blocking are drivers of extreme low electricity production and demand over Europe during winter 13,31 . In addition to low windspeed and low temperatures, these high-pressure systems are characterized by little to no cloud formation, potentially driving meteorological droughts 32 . This effect becomes even more pronounced during summer energy droughts, when high-pressure systems over Southern Europe form 'heat domes' that trap warm air in the region, causing extremely hot weather conditions. Subsidence associated with these systems prevent cloud formation, leading to clear skies, increased solar radiation and subsequently higher temperatures, which in turn aggravate hydrological drought conditions by strong evaporation of surface water and soil moisture 32 . Thus, during periods of extremely high demand and minimal wind production, a concurrent reduction of water inflow into reservoirs often occurs (see Supplementary Fig. 5). This causes a decline in hydropower levels that can be exacerbated by the response in hydropower dispatch. In an attempt to offset the increased demand and decreased production of wind energy, there is a rise in hydropower outflow (for example in Italy, see Fig. 2 ). This dual challenge—of reduced water supply occurring during heightened electricity demand—underscores the importance of managing hydropower resources effectively. Due to the unique hydropower dispatch characteristics in different countries, the (severity of) meteorological conditions leading to energy droughts vary significantly. In France, the meteorological anomalies during energy drought events are more extreme than in the other countries, as renewable production highly depends on non-dispatchable sources. Conversely, Switzerland, Sweden and Norway have a relatively large reservoir storage capacity that can produce electricity in periods of high demand. Figure 2 shows that energy droughts in Norway are linked to highly negative hydropower reservoir level anomalies, with precise meteorological conditions during the EDW, such as exceptionally low temperatures and wind speeds, playing a secondary role. The depleted reservoir levels are a result of multiple years of imbalance between electricity demand and inflow, often caused by complex and temporally compounding processes. Temporally compounding conditions leading to energy droughts We distinguish three types of temporally compounding conditions (TCCs) connected to these low reservoir levels and concurrent energy droughts: TCC I) countries with reservoirs that are filled in the spring face vulnerability to summer energy droughts if these are preceded by dry spring conditions, resulting in low reservoir levels and a dry subsoil in the subsequent summer, TCC II) low precipitation or runoff during reservoir filling season compounded with persistent high-pressure systems in autumn and winter leading to winter energy droughts, and TCC III) in certain regions (North-Europe) there is an increased risk of low runoff in spring after a dry and extremely cold winter, increasing the chance of multi-year energy droughts. Under TCC I, the risks of a summer EDW in regions with considerable cooling demands, particularly Southern Europe, can be up to four times greater than under normal conditions (Fig. 3 e). This elevated risk occurs when a notable decline in reservoir inflow and an increase in demand during spring leaves the reservoir levels anomalously low in July and August (Fig. 3 a and 3 b, Italy serving as example). Low reservoir inflows occur when the runoff is low (Fig. 3 d), which can be attributed to two factors. Firstly, in mountainous regions like the Alpine area, an unusually warm winter with reduced snowfall results in a diminished snowpack by the end of winter. This reduction significantly impacts hydropower generation later that year, as indicated in Supplementary Fig. 6. For instance, in the Po River basin of Italy, snow accumulation typically peaks between December and February. As spring arrives, rising temperatures gradually initiate the snowmelt process, releasing water into the Po River and its tributaries. Diminished snowmelt during this period leads to reduced inflow into hydropower reservoirs during spring, affecting hydropower generation in the subsequent months. The Po River drought in the year 2022 serves as a pertinent example of this phenomenon 33 . Secondly, during the spring season, a meteorological drought characterized by reduced precipitation due to persistent high-pressure systems also contributes to decreased inflow. Similar to the patterns observed during the summer EDW, this drought coincides with higher-than-normal temperatures (Fig. 3 c). The electricity system faces increased vulnerability to an energy drought when not only supply is low (depleted reservoirs) but concurrently demand is high (cooling demand during exceptional heat). Dry winter and spring conditions not only leave reservoirs depleted but also result in dry subsoil. Note that the risk of exceptionally hot summers is heightened in these cases, due to this dry subsoil and land-atmosphere feedback mechanisms 34,35 . So, although by design, the energy drought window lasts 30 days, Fig. 3 b demonstrates how the compounding of these conditions can lead to extended periods of high residual loads; On average, the periods around an EDW encompass 104 consecutive days of positive residual load anomalies. Besides the elevated demand during a warmer than normal spring, at the end of a summer, temperatures may return to normal, but the reservoirs have been depleted, resulting in prolonged periods of high residual loads while entering the subsequent winter. Under TCC II (Fig. 4 ), the combination of low filling of reservoirs in spring and subsequent persisting high-pressure systems in the following winter, set the stage for winter energy droughts. The risk of a winter energy drought following such conditions is 2.6 times higher in Switzerland and up to 4.4 times higher in the other countries (Fig. 4 e). Figure 4 shows that for Switzerland, like in TCC I, increased temperatures and strongly reduced runoff in May lead to reduced reservoir levels. As winter approaches, prolonged high-pressure systems can reduce precipitation and thereby oppose the replenishment of these reservoirs. Together with increased heating demands associated with these weather systems (Fig. 4 b) residual loads can get exceptionally high for long periods of time leading up to winter events. Importantly, the inflow into the reservoirs remains low during subsequent filling seasons after the occurrence of winter EDWs, which works against the recovery and robustness of the entire system. In TCC III (see Fig. 5 , Sweden serving as example), winter energy droughts are followed by lower-than-usual reservoir inflows during the subsequent spring season. Precipitation, evaporation, and snowmelt after winter EDWs show little to no significant deviation during the filling season when compared to years without such events, but the runoff (Fig. 5 d) and consequently reservoir inflow (Fig. 5 b) is very low. This phenomenon is predominantly observed in Northern European countries, particularly when winter EDWs are preceded by dry springs and summers (as detailed in TCC II), rendering the electricity system vulnerable, but also leaving the soil dry. This soil moisture drought is exacerbated during the winter EDWs, during which episodes of extreme cold coincide with reduced runoff, as can be observed in Fig. 2 . This dry subsoil (Fig. 5 c) is the main cause of continued low runoff in spring. Precipitation following an energy drought initially infiltrates the soil to replenish the moisture deficit before eventually moving via runoff into streamflow and ultimately flowing into the hydropower reservoirs. This particular phenomenon nearly doubles the risk of experiencing extremely low inflow into the reservoirs, opposing the chance for the electricity system to rebound. Figure 5 a shows that up to 3 years after the occurrence of a winter EDW the mean hydropower reservoirs levels remain anomalously low. Prolonged, cumulative droughts In four out of the six studied countries we found that as a result of these compounding events, energy droughts are rarely isolated events, but rather a consequence of feedback mechanisms that tend to prolong periods of energy scarcity (Fig. 6 a) and can turn what might have been a short-term drought into a prolonged, cumulative energy crisis. The sequence of the three temporally compounding conditions creates a cyclical pattern, feeding into each other. Typically, low reservoir inflows in the spring season (AMJ) are caused by a thinner snowpack and reduced precipitation. These conditions also lead to relatively dry subsoils, which in turn increase the likelihood of extreme heat events during the summer months. The combination of diminished inflow and increased demand for cooling can lead to an energy shortage in the summer (JA), TCC I. If the low inflows persist, reduced reservoir levels by winter may trigger an energy drought (DJFM), TCC II. Furthermore, in Northern Europe, there is a notable risk of decreased spring runoff (MJ) due to anomalously high residual loads. This is attributed to the low soil moisture conditions originating from previous periods, TCC III, further enhancing the feedback loop. At least partially due to this feedback loop, the sets of 10-year climate model simulations that contain one or more energy droughts tend to experience a progressive increase in the residual load anomaly over time (Figs. 6 b and 6 c), suggesting that decadal climate variability plays a role in determining the occurrence of prolonged energy droughts in Europe. Discussion and conclusions This study highlights the challenge posed by high electricity demands coinciding with periods of reduced renewable electricity production. Within this context, we have identified three distinct types of temporally compounding conditions, providing a framework for early warning systems and the development of mitigation strategies. However, the results suggest that these three TCCs may be an intrinsic part of a cyclical pattern that plays a role in decadal variability in residual load anomalies. These analyses demonstrate that relying solely on multiple individual weather years to assess system robustness is inadequate to capture the increased risk and intricate nature associated with multi-year energy droughts. This is particularly relevant for hydroelectrical systems, for which, due to the temporally compounding nature of drought events, evaluating consecutive multi-year data becomes essential for a comprehensive understanding and planning for extreme scenarios. For Northern European countries, managing their hydropower reservoirs during winter energy droughts is essential as the risks of periods of low reservoir inflow increase after energy droughts, rendering the system's recovery more challenging. Based on the observation that meteorological conditions leading to electricity system stress often align with periods of anomalously high solar radiation, Southern European countries with large hydropower capacities could potentially mitigate summer energy droughts by the investment in additional solar panels. We have focused on the three main types of renewable energy production in Europe. An additional vulnerability of the electricity system is that during periods of reduced discharge (leading to low reservoir inflows) there might be concurrent shortfall in freshwater used for cooling by thermal power plants. This water scarcity aggravates the effects of meteorological conditions on the energy system, but this effect is not incorporated here. We based this analysis on a single global climate model that does not incorporate glacier dynamics and melt. Although this model has been assessed against ERA5 for the different regions, it has not been bias corrected. Future research should evaluate the consistency of our findings across multiple climate models. Additional next research phases would be to examine how climate change and the energy transition affect long-lasting energy droughts. Increased renewable energy capacities and more intense/frequent weather extremes may lead to even greater challenges in future energy management. The interplay between a changing climate and electricity system adaptations needs analyses to develop strategies mitigating the impacts of extended energy shortages, crucial for shaping sustainable energy policies in an era of climate uncertainty. Methods Meteorological data We use the daily outputs from the KNMI Large ENsemble TIme Slice (KNMI-LENTIS) 29 , a large climate ensemble composed of 160 sets of 10-year climate model simulations of present-day climate (the time slice 2000–2009). The KNMI-LENTIS ensemble was executed with EC-Earth3, a fully coupled, state-of-the-art global Global Climate Model that includes atmospheric, oceanic, land, and sea-ice components 36 . This time slice ensemble methodology is especially suitable for investigating extreme events. The 1600 years of data allows for the examination of 160 extreme events that occur once in 10 years or less. Moreover, the sequence of 10 continuous weather years in each simulation makes it possible to study temporally compounding and multi-year events. We did not bias correct the LENTIS data prior to application in the energy modelling framework. In the context of this multivariate research, bias correction of climate data may unintentionally distort the physical relationships between different variables and the seasons, potentially leading to inaccurate interpretations of climate interactions and dynamics. However, to demonstrate the reliability and variance of the LENTIS data in combination with the energy modelling framework, we compared the climatological mean and variance of meteorological variables for each country against the 1991–2020 ERA5 reanalysis data 37 (Supplementary Fig. 7 and Supplementary Fig. 8). The discharge peak of Norway, Sweden and Switzerland occurs slightly earlier in the LENTIS data compared to the ERA5 data. However, this affect is mitigated by the dispatch optimization model, which strategically retains water in reservoirs to meet periods of high demand. This approach ensures that, regardless of inflow variations, water release occurs at comparable times annually for both LENTIS and ERA5. Electricity production and demand modelling framework The electricity modelling framework includes gridded production modules for offshore wind, onshore wind, solar photovoltaics (PV), run-of-river, and hydropower reservoir inflow. It also contains a hydropower dispatch module and an electricity demand module, both operating at the European country level 38 . For wind and PV solar power modules, the inputs include 10 m wind speed [ \(m {s}^{-1}\) ], near-surface temperature [°C], near-surface maximum daily temperature [°C], and solar irradiance [ \(kW {m}^{-2}\) ]. We calculate wind power by extrapolating 10 m wind speeds to hub height in both onshore and offshore scenarios using the power law equation. We then estimate the power generation through a power curve, neglecting downtime and efficiency losses. We base PV solar power on solar radiation and temperature-dependent cell efficiency, which we determine using near-surface temperature, maximum daily temperature, and wind speed. We assume that all PV modules are horizontally oriented and functional during all daylight hours. Demand estimation incorporates population-weighted temperature data and historical demand measurement data to create demand fits. These fits account for differences between weekday and weekend demand, though they do not consider cultural influences such as holidays. We calculate the generation of run-of-river hydropower based on discharge values obtained from runoff [ \(m {timestep}^{-1}\) ], which we process through a river routing scheme. In hydropower reservoir systems, we also determine reservoir inflow using this discharge data, and manage outflow with a 4-week moving horizon dispatch optimization, avoiding the benefit of perfect foresight. A detailed description of all the modules within the modeling framework, along with a validation of how it performs on the selection of energy droughts with the use of ERA5 data, are available in Ref 28 . Installed capacities We conducted the study using gridded files that reflect Europe’s 2020 installed renewable capacities. The locations of solar photovoltaic and onshore wind farms were determined based on the 2020 data from OpenStreetMap 39 , EMODnet provides vector data of installed offshore wind capacities 40 , and we sourced the locations of run-of-river and reservoir hydropower plants from the Joint Research Centre (JRC) Hydropower database 41 . These datasets were converted into gridded capacity files according to the methods described in Ref 28 . Selection of drought events Our analysis targets energy drought events that have a return period of 10 years or longer. To this end, we selected the most significant 160 events from our 1600-year database. Persistent Energy Droughts (PEDs) are defined as periods where the residual load exceeds the 97th percentile threshold for several consecutive days. A PED concludes when the load falls below this threshold for more than three consecutive days. The PEDs are ranked in descending order of duration, focusing on the 160 longest events to identify the most extreme cases. Differentiation between summer, winter, and all-year events is made, with each category benchmarked against their specific 97th percentile threshold. To evaluate the effects of meteorological compounding conditions in these events, we analyzed same-length events in the form of Energy Drought Windows (EDWs) that span 30 days, or are confined to a specific month. This selected duration aligns with the approximate average length of PEDs as observed across the countries included in our analysis. The 160 EDWs with the highest average residual loads (and without overlap) are selected to identify the most impactful energy droughts. Risk ratios We quantify and compare relative risks of energy droughts under distinct temporally compounding meteorological conditions using calculated risk ratios (RRs): the ratio of the probability of an energy drought occurring after certain meteorological conditions (e.g., low inflow into hydropower reservoirs in spring) to the climatological probability. We assessed the statistical significance and precision of the estimated RR by computing its 95% confidence interval (CI) according to \(CI= \text{exp}(\text{log}RR\pm Z x SE(\text{log}RR\left)\right)\) , where RR is the risk ratio, SE is the standard error of the log-transformed risk ratio, and Z represents the standard normal deviate (1.96 for a 95% CI). Risk ratios for which the 95% CI does not include 1 are considered statistically significant. Standardized Soil Moisture Index (SSMI) We analyze soil moisture variability after winter energy droughts with the Standardized Soil Moisture Index (SSMI) 42 . The SSMI uses the z-score methodology to quantify soil moisture deviations in terms of standard deviations from the mean and is calculated using the formula: \(SSMI=\left(mrso- \mu \right)/std\) , where \(mrso\) represents the soil moisture value for a given day, \(\mu\) is the mean soil moisture over the 1600-year period on that day of year, and \(std\) is the standard deviation of soil moisture during that day of year. Declarations Data availability All analysis data are included within the article and its Supplementary Information files. The simulated energy generation and demand data supporting the findings of this study will be made available after the review. References Yalew, S. G. et al. Impacts of climate change on energy systems in global and regional scenarios. Nat. Energy 5 , 794–802 (2020). Staffell, I. & Pfenninger, S. The increasing impact of weather on electricity supply and demand. Energy 145 , 65–78 (2018). Van Der Wiel, K., Selten, F. M., Bintanja, R., Blackport, R. & Screen, J. A. Ensemble climate-impact modelling: extreme impacts from moderate meteorological conditions. Environ. Res. Lett. 15 , (2020). Zscheischler, J. et al. A typology of compound weather and climate events. Nature Reviews Earth and Environment vol. 1 (2020). Copernicus Climate Change Service (C3S). Climate bulletins. https://climate.copernicus.eu/climate-bulletins (2021). Morison, R. & Shiryaevskaya, A. U.K. Power Surges to Record 400 Pounds as Wind Fails to Blow. Bloomberg (2021). Fernández, L. Hydroelectricity generation Spain 2010-2022. Statista https://www.statista.com/statistics/1006362/hydroelectricity-generation-in-spain/ (2023). Ingram, E. Hydroelectric generation in Italy decreased 37.7% in 2022. Hydro review (2023). Copernicus Climate Change Service (C3S). European State of the Climate 2022 Unprecedented extreme heat and widespread drought mark European climate in 2022. Copernicus https://climate.copernicus.eu/copernicus-european-state-climate-2022-unprecedented-extreme-heat-and-widespread-drought-mark (2023). Gualtieri, T. Drought Forces One of Spain’s Largest Hydro Plants to Halt. Bloomberg (2022). International Energy Agency (IEA). Hydropower Data Explorer. IEA (2021). Perera, A. T. D., Nik, V. M., Chen, D., Scartezzini, J. L. & Hong, T. Quantifying the impacts of climate change and extreme climate events on energy systems. Nat. Energy 5 , 150–159 (2020). van der Wiel, K. et al. Meteorological conditions leading to extreme low variable renewable energy production and extreme high energy shortfall. Renew. Sustain. Energy Rev. 111 , 261–275 (2019). Otero, N., Martius, O., Allen, S., Bloomfield, H. & Schaefli, B. Characterizing renewable energy compound events across Europe using a logistic regression-based approach. Meteorol. Appl. 29 , (2022). Jurasz, J., Mikulik, J., Dabek, P. B., Guezgouz, M. & Kaźmierczak, B. Complementarity and ‘resource droughts’ of solar and wind energy in poland: An era5-based analysis. Energies 14 , (2021). Kay, G. et al. Variability in North Sea wind energy and the potential for prolonged winter wind drought. Atmos. Sci. Lett. 24 , (2023). Otero, N., Martius, O., Allen, S., Bloomfield, H. & Schaefli, B. A copula-based assessment of renewable energy droughts across Europe. Renew. Energy 201 , (2022). Jahns, C., Osinski, P. & Weber, C. A statistical approach to modeling the variability between years in renewable infeed on energy system level. Energy 263 , (2023). Raynaud, D., Hingray, B., François, B. & Creutin, J. D. Energy droughts from variable renewable energy sources in European climates. Renew. Energy 125 , 578–589 (2018). Bloomfield, H. C. et al. Quantifying the sensitivity of european power systems to energy scenarios and climate change projections. Renew. Energy 164 , 1062–1075 (2020). Craig, M. T. et al. Overcoming the disconnect between energy system and climate modeling. Joule 1405–1417 (2022) doi:10.1016/j.joule.2022.05.010. Grochowicz, A., van Greevenbroek, K., Benth, F. E. & Zeyringer, M. Intersecting near-optimal spaces: European power systems with more resilience to weather variability. Energy Econ. 118 , (2023). Van Vliet, M. T. H., Sheffield, J., Wiberg, D. & Wood, E. F. Impacts of recent drought and warm years on water resources and electricity supply worldwide. Environ. Res. Lett. 11 , (2016). Tsanis, I. & Tapoglou, E. Winter North Atlantic Oscillation impact on European precipitation and drought under climate change. Theor. Appl. Climatol. 135 , 323–330 (2019). Ely, C. R., Brayshaw, D. J., Methven, J., Cox, J. & Pearce, O. Implications of the North Atlantic Oscillation for a UK-Norway Renewable power system. Energy Policy 62 , (2013). Rakovec, O. et al. The 2018–2020 Multi-Year Drought Sets a New Benchmark in Europe. Earth’s Futur. 10 , (2022). van der Wiel, K., Batelaan, T. J. & Wanders, N. Large increases of multi-year droughts in north-western Europe in a warmer climate. Clim. Dyn. 60 , (2023). van der Most, L. et al. Extreme events in the European renewable power system: Validation of a modeling framework to estimate renewable electricity production and demand from meteorological data. Renew. Sustain. Energy Rev. 170 , 112987 (2022). Muntjewerf, L., Bintanja, R., Reerink, T. & van der Wiel, K. The KNMI Large Ensemble Time Slice (KNMI-LENTIS). Geosci. Model Dev. 16 , 4581–4597 (2023). Ohlendorf, N. & Schill, W. P. Frequency and duration of low-wind-power events in Germany. Environ. Res. Lett. 15 , (2020). Bloomfield, H. C., Suitters, C. C. & Drew, D. R. Meteorological Drivers of European Power System Stress. J. Renew. Energy 2020 , 1–12 (2020). Kautz, L. A. et al. Atmospheric blocking and weather extremes over the Euro-Atlantic sector - A review. Weather and Climate Dynamics vol. 3 (2022). Montanari, A. et al. Why the 2022 Po River drought is the worst in the past two centuries. Sci. Adv. 9 , (2023). Fischer, E. M., Seneviratne, S. I., Lüthi, D. & Schär, C. Contribution of land-atmosphere coupling to recent European summer heat waves. Geophys. Res. Lett. 34 , (2007). Miralles, D. G., Gentine, P., Seneviratne, S. I. & Teuling, A. J. Land–atmospheric feedbacks during droughts and heatwaves: state of the science and current challenges. Ann. N. Y. Acad. Sci. 1436 , 19–35 (2019). Döscher, R. et al. The EC-Earth3 Earth system model for the Coupled Model Intercomparison Project 6. Geosci. Model Dev. 15 , (2022). Hersbach, H. et al. The ERA5 global reanalysis. Q. J. R. Meteorol. Soc. 146 , 1999–2049 (2020). Eurostat. NRG-IND-REN Share of energy from renewable sources. Eurostat https://ec.europa.eu/eurostat/databrowser/product/view/NRG_IND_REN?lang=en (2021). Dunnett, S., Sorichetta, A., Taylor, G. & Eigenbrod, F. Harmonised global datasets of wind and solar farm locations and power. Sci. Data 7 , 1–12 (2020). EMODnet Human Activities project. Emodnet_HA_WindFarms_20200305 [data set]. European Commission Joint Research Centre (JRC). JRC Hydro-power database. (2019) doi:https://doi.org/10.5281/zenodo.3862722 [data set]. Hao, Z. & AghaKouchak, A. Multivariate Standardized Drought Index: A parametric multi-index model. Adv. Water Resour. 57 , (2013). Additional Declarations There is NO Competing Interest. Supplementary Files Supplementaryinformation.docx supplementary information Cite Share Download PDF Status: Published Journal Publication published 27 Sep, 2024 Read the published version in Nature Energy → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-3796061","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":266057763,"identity":"4dc91bdf-e0d9-4e65-b95c-20a7e0c97479","order_by":0,"name":"Lieke van der Most","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA20lEQVRIiWNgGAWjYBACPgjFDOPbQKgHeLSwoWpJSIPSJGg5TIwWHsPPBQzWcuYN3IkPf/44n29wfgHbAwJajKVnMKQbyxzg3WzMk3DbcsONB+wG+LWwJUjzMBxOnMHAu02aIeG2gcGNA2wSBLQk/wZqqQdpkfyRcI4YLczHQLYkSAC1SPAkHDAwON9AQAsz8zFrHoN0wxnMIL+kJRtI3mBsw6uFn72x+TZPhbW8BHvvxoc/bOwM+M4fPibxAY8WSIwYMCAlAInEBnwasFp8gFQdo2AUjIJRMMwBAIR4Pm52jt7kAAAAAElFTkSuQmCC","orcid":"https://orcid.org/0000-0003-0678-6887","institution":"University of Groningen","correspondingAuthor":true,"prefix":"","firstName":"Lieke","middleName":"van der","lastName":"Most","suffix":""},{"id":266057764,"identity":"8cedda21-ba2b-4a41-89d0-25a48e90c782","order_by":1,"name":"Karin van der Wiel","email":"","orcid":"","institution":"Royal Netherlands Meteorological Institute","correspondingAuthor":false,"prefix":"","firstName":"Karin","middleName":"van der","lastName":"Wiel","suffix":""},{"id":266057765,"identity":"2eb05db7-ca35-421d-904b-9b12e794fd31","order_by":2,"name":"Winnie Gerbens-Leenes","email":"","orcid":"","institution":"University of Groningen","correspondingAuthor":false,"prefix":"","firstName":"Winnie","middleName":"","lastName":"Gerbens-Leenes","suffix":""},{"id":266057766,"identity":"259dedb0-0962-4ec1-b595-55ce0ed511b0","order_by":3,"name":"R.M.J. (René) Benders","email":"","orcid":"","institution":"University of Groningen","correspondingAuthor":false,"prefix":"","firstName":"R.M.J.","middleName":"(René)","lastName":"Benders","suffix":""},{"id":266057767,"identity":"dd0aafc2-1709-4a80-823e-94a4a469f8c0","order_by":4,"name":"Richard Bintanja","email":"","orcid":"","institution":"University of Groningen","correspondingAuthor":false,"prefix":"","firstName":"Richard","middleName":"","lastName":"Bintanja","suffix":""}],"badges":[],"createdAt":"2023-12-23 11:10:10","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-3796061/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-3796061/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41560-024-01640-5","type":"published","date":"2024-09-27T04:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":49433810,"identity":"fa027ed3-7a7e-4010-82ff-9781774b652c","added_by":"auto","created_at":"2024-01-10 19:21:14","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":289827,"visible":true,"origin":"","legend":"\u003cp\u003ePersistent Energy Droughts (PED) with daily residual loads above the 97\u003csup\u003eth\u003c/sup\u003e percentile for the once in 10 years or less PEDs. With in a) a map of the mean length of the PEDs per country, and in b-g) the number of of PED-days for each day of the year. Blue shows winter PEDs (Oct-May) with the threshold of 97\u003csup\u003eth\u003c/sup\u003e percentile relative to the residual loads in the winter months. Yellow shows summer PEDs, like the winter PEDs but for the months Jun-Sep. In pink the once in 10 years PEDs independent of the season. The barplots on top of the subplots show the mean residual load during the summer, winter or all-year PEDs relative to the overall mean residual daily load (in black).\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-3796061/v1/720b4b107e60d6b6aea383df.png"},{"id":49433812,"identity":"d26c23ef-e683-4929-beac-347f5a701960","added_by":"auto","created_at":"2024-01-10 19:21:14","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":1360745,"visible":true,"origin":"","legend":"\u003cp\u003eRepresentation of the 30-day Energy Drought Windows (EDW) occurring once in 10 years or less. Winter EDWs for Norway (top row) and France (middle row) and summer EDWs for Italy (bottom). From left to right, the panels show anomalies of near-surface temperature, near-surface wind speed, incoming solar radiation, and total runoff during the 30-day EDW period. Red contour lines show the mean sea level pressures during the events. On the right side, reservoir levels in the days leading up to and during the EDWs. The grey areas show the 10\u003csup\u003eth\u003c/sup\u003e-90\u003csup\u003eth\u003c/sup\u003e and the 25\u003csup\u003eth\u003c/sup\u003e-75\u003csup\u003eth\u003c/sup\u003e percentiles of the hydropower reservoir levels, and the dark grey line the composite mean of all events. The blue section shows the winter events and in orange for Italy the summer event.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-3796061/v1/b807cc75586ccccbbb5b90ed.png"},{"id":49433811,"identity":"e057e213-a585-44d2-aa3e-0562eff83c98","added_by":"auto","created_at":"2024-01-10 19:21:14","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":423761,"visible":true,"origin":"","legend":"\u003cp\u003eTemporally compounding conditions I. The once in 10 years or less August Energy Drought Windows (EDWs) for Italy. With a) a composite timeseries of the reservoir level anomalies before, during and after the events with in red the spring period before the summer energy drought during which a negative reservoir anomaly builds up and in orange the summer energy drought event, b) a plot of the temporal development of the 7-day rolling mean anomalies of residual load, demand, and reservoir in- and outflow, and c) composite map of the near-surface temperature anomalies during the red colored period leading-up to the events with the black contour lines showing mean sea level pressures, d) like c but for total runoff, and e) the risk factor of a once in 10 years or less summer drought event, after low reservoir inflow in May in Norway (NOR), Switzerland (CHE), Sweden (SWE), Italy (ITA), Spain (ESP), and France (FRA). Risk ratios that are not statistically significant are shaded.\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-3796061/v1/e3983fcc498f6dd2ad5cb029.png"},{"id":49434295,"identity":"064ce978-f07b-422f-a867-23a9c62bcc46","added_by":"auto","created_at":"2024-01-10 19:29:14","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":442016,"visible":true,"origin":"","legend":"\u003cp\u003eTemporally compounding conditions II. The once in 10 years or less February Energy Drought Windows (EDWs) in Switzerland. Individual panels as in Figure 3.\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-3796061/v1/8ebf6f4099e3e9310b9ea42e.png"},{"id":49433814,"identity":"2da27442-f464-4136-b4b3-ca34c6273755","added_by":"auto","created_at":"2024-01-10 19:21:14","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":364073,"visible":true,"origin":"","legend":"\u003cp\u003eTemporally compounding event III. The once in 10 years or less February Energy Drought Windows (EDW) for Sweden, and the reservoir filling season afterwards. With b) the distribution of inflow during the red period for all 1600 years and for the years with a preceding energy drought in winter, c) the Standardized Soil Moisture Index (SSMI) at the start of May, and e) the risk factor of a once in 10 years or less low inflow in spring after high residual loads in February in Norway (NOR), Switzerland (CHE), Sweden (SWE), Italy (ITA), Spain (ESP), and France (FRA). Panel a) and d) as in Figure 3.\u003c/p\u003e","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-3796061/v1/cc5318a1ecc52727d54d2446.png"},{"id":49433815,"identity":"a61b40d9-26cc-43b4-80f8-6835ab870457","added_by":"auto","created_at":"2024-01-10 19:21:14","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":231517,"visible":true,"origin":"","legend":"\u003cp\u003eThe reinforcing cycle of multi-year energy droughts. With a) a schematic overview of the energy drought cycle. (red arrows implicate increased risk as a result of temporally compounding conditions, black arrows a causal link), b) the cumulative sum of residual load anomalies normalized over yearly demand values with in red the sets of the 160 climate model simulations that include energy drought events and in grey all 160 simulations, and c) violin plots of the normalized residual load anomalies during the last timestep of the simulations for four countries.\u003c/p\u003e","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-3796061/v1/393c8f6c08084968b972d06a.png"},{"id":65485022,"identity":"8a82fc91-81cc-4a9c-9869-35d12f0cb146","added_by":"auto","created_at":"2024-09-28 07:12:57","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3487025,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-3796061/v1/f04c15b8-2c29-4ba4-a102-a11d6d440111.pdf"},{"id":49434296,"identity":"aa9a6a2a-8790-45bd-9901-1d74eae12419","added_by":"auto","created_at":"2024-01-10 19:29:14","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":5988864,"visible":true,"origin":"","legend":"\u003cp\u003esupplementary information\u003c/p\u003e","description":"","filename":"Supplementaryinformation.docx","url":"https://assets-eu.researchsquare.com/files/rs-3796061/v1/fb60dea86cd304e85cc0e4df.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Temporally compounding energy droughts in European electricity systems with hydropower","fulltext":[{"header":"Introduction","content":"\u003cp\u003eAs Europe\u0026rsquo;s capacity of renewable electricity sources grows, the region's electricity system is becoming increasingly dependent on variations in weather and climate\u003csup\u003e1,2\u003c/sup\u003e. Electric heating and cooling demand are largely defined by air temperatures and the production of renewables depends on the availability of resources (wind speed, irradiance, discharge). With this increased sensitivity to climate variations arises the risk of encountering weather conditions outside of the system\u0026rsquo;s design scope, undermining energy security. Furthermore, due to non-linear relationships between climate factors and the electricity system, extreme impact events in the form of high demand and low renewable electricity production (hereafter referred to as energy droughts) might occur under compounding meteorological conditions that in itself are not considered extreme\u003csup\u003e3\u003c/sup\u003e, and vice versa.\u003c/p\u003e \u003cp\u003eThese so-called compounding events\u003csup\u003e4\u003c/sup\u003e have recently had significant impact in electricity systems in a number of European regions. During the autumn of 2021, Europe experienced record-low gas storage levels, a situation influenced by several factors. Besides market prices and geopolitical causes, other contributing factors were a preceding cold winter in Northern Europe followed by another colder-than-usual season in spring, compounded by very low annual average wind speeds in parts of northwestern and central Europe\u003csup\u003e5\u003c/sup\u003e. In September 2021, this prolonged period of calm weather conditions even led to record-high electricity prices in the UK energy market\u003csup\u003e6\u003c/sup\u003e. Furthermore, in the beginning of 2022, Spain and Italy saw a dramatic 40% decline in hydroelectricity generation\u003csup\u003e7,8\u003c/sup\u003e. This sharp drop was caused by an exceptionally hot summer and one of the driest years for water resources since record-keeping began\u003csup\u003e9\u003c/sup\u003e. Consequently, some hydropower plants even had to cease operations in the midst of a cooling demand peak\u003csup\u003e10\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eHydropower is Europe\u0026rsquo;s main source of renewable electricity to date\u0026mdash;with an installed capacity of 188 GW it makes up for 36% of the renewable capacity, and supplies 17% of the total EU electricity demand\u003csup\u003e11\u003c/sup\u003e. In seasons of high water availability, like spring and early summer, water is accumulated in reservoirs and released during periods of peak demand, such as winter. This seasonal and dispatchable nature of hydropower potentially makes an electricity system vulnerable to prolonged and temporally compounding energy droughts.\u003c/p\u003e \u003cp\u003eHowever, previous studies that modelled extreme energy drought events in Europe\u0026rsquo;s renewable electricity system have focused on wind power and solar PV generation\u003csup\u003e12\u0026ndash;18\u003c/sup\u003e or only included run of river hydropower production\u003csup\u003e19\u003c/sup\u003e, overlooking the critical role of reservoir storage and its operational dynamics. One of the reasons hydropower reservoir production is often excluded from studies into the weather sensitivity of electricity systems is that it is characterized by complex operating procedures\u003csup\u003e20\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eEnergy system models can be employed to simulate this, but they are typically computationally intensive, leading to reliance on shorter datasets or estimates of annual energy potential\u003csup\u003e21\u003c/sup\u003e. Recent work applied novel approaches to allow for the handling of an ensemble of single weather years\u003csup\u003e22\u003c/sup\u003e. However, hydropower not only experiences significant variations from year to year\u003csup\u003e23\u003c/sup\u003e\u0026mdash;for instance, changes in the North Atlantic Oscillation (NAO) index can explain up to 30% of the fluctuations in Norway's hydropower production\u003csup\u003e24,25\u003c/sup\u003e\u0026mdash;but might also be susceptible to sustained climate phenomena such as multi-year meteorological or hydrological droughts\u003csup\u003e23,26,27\u003c/sup\u003e. To accurately represent this variability, extended time series data is necessary. The climate and energy modelling community suggest identifying meteorological conditions that have extreme impacts in the electricity system and integrating them into energy system simulations to tackle the problem of handling such large datasets\u003csup\u003e21\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eHere we identify meteorological conditions that (temporally) compound to have extreme impacts on electricity systems including hydropower using climate model data and a daily renewable electricity production and demand modelling framework\u003csup\u003e28\u003c/sup\u003e. We employ a large ensemble approach, using 160 sets of 10-year climate model simulations (1600 years)\u003csup\u003e29\u003c/sup\u003e, each set representing a different possible sequence of weather under present-day climate conditions. We present case-studies for the six European countries with the highest installed hydropower capacities (Norway, France, Italy, Spain, Switzerland and Sweden) and evaluate daily residual loads \u0026mdash;demand minus renewable energy generation\u0026mdash;for the full 1600 years of climate data to: 1) identify and quantify different types of energy droughts in the electricity system, 2) analyze temporal compounding in these events, and 3) quantify risk ratios of meteorological conditions leading to extreme impact events.\u003c/p\u003e"},{"header":"High residual load events in electricity systems with hydropower","content":"\u003cp\u003eWe classify high residual load events using two metrics. First, we introduce the concept of 'Energy Drought Windows (EDWs)'\u003csup\u003e13\u003c/sup\u003e, which are fixed-duration periods for which we select the highest residual loads from the ensemble. Secondly, we examine 'Persistent Energy Droughts (PEDs)'\u003csup\u003e19,30\u003c/sup\u003e. These drought events occur when the daily residual load remains consistently above a predefined threshold for consecutive days. Additionally, we differentiate between summer, winter, and season independent events by defining PED thresholds relative to the respective seasonal residual loads.\u003c/p\u003e \u003cp\u003eWe find that electricity systems with relatively large hydropower reservoirs experience prolonged PEDs (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea). In Norway, Sweden, and Switzerland, PEDs can last more than 150 days. These countries use their reservoir storage to release stored water strategically during brief periods of extreme meteorological conditions characterized by heightened demand and/or reduced renewable electricity production, mitigating short term energy shortages. However, once the reservoir depletion exceeds a critical threshold and hydropower resources are exhausted, these systems may face very long energy droughts.\u003c/p\u003e \u003cp\u003eIn the present-day European electricity systems, energy droughts predominantly occur during the winter months (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb-g). These can be attributed to factors such as strong heating demand and reduced wind power production (a result of the greater variability in wind speeds during the winter compared to the summer, which enhances the chance of periods with low wind\u003csup\u003e13\u003c/sup\u003e), and diminished reservoir storage levels. Regions in more northerly latitudes generally experience these events later in the year. This can be explained by two factors: the coldest periods occur later, and relatively large reservoir storages tend to deplete later in the year.\u003c/p\u003e \u003cp\u003eAll-year PEDs driven by increased cooling demands during hot summer days only occur in Southern European countries like Spain and Italy (Figs.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ed and \u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eg)\u003csup\u003e31\u003c/sup\u003e. These PEDs can be intensified by reduced water availability affecting hydropower production. Since electricity demand response to high temperatures is small in Northern European countries, summer events typically transpire during early or late summer when temperatures are unseasonably low, leading to increased heating requirements. The barplots in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e indicate that the mean residual load during these summer events in Northern countries is not significantly higher than the average yearly residual load, meaning that, in an absolute sense, summer PEDs are not extreme.\u003c/p\u003e "},{"header":"Compounding conditions during energy droughts","content":"\u003cp\u003eFor the analysis of the meteorological conditions during events we examine energy drought windows of 30 days. Figure\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e shows the meteorological conditions during these EDWs and the reservoir levels leading up to them, using Norway, France and Italy as illustrative examples (see Supplementary Figs.\u0026nbsp;1 and 2 for other countries).\u003c/p\u003e\n\u003cp\u003eWe find that prolonged high-pressure systems or ridges and atmospheric blocking are drivers of extreme low electricity production and demand over Europe during winter\u003csup\u003e13,31\u003c/sup\u003e. In addition to low windspeed and low temperatures, these high-pressure systems are characterized by little to no cloud formation, potentially driving meteorological droughts\u003csup\u003e32\u003c/sup\u003e. This effect becomes even more pronounced during summer energy droughts, when high-pressure systems over Southern Europe form 'heat domes' that trap warm air in the region, causing extremely hot weather conditions. Subsidence associated with these systems prevent cloud formation, leading to clear skies, increased solar radiation and subsequently higher temperatures, which in turn aggravate hydrological drought conditions by strong evaporation of surface water and soil moisture\u003csup\u003e32\u003c/sup\u003e. Thus, during periods of extremely high demand and minimal wind production, a concurrent reduction of water inflow into reservoirs often occurs (see Supplementary Fig.\u0026nbsp;5). This causes a decline in hydropower levels that can be exacerbated by the response in hydropower dispatch. In an attempt to offset the increased demand and decreased production of wind energy, there is a rise in hydropower outflow (for example in Italy, see Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e). This dual challenge\u0026mdash;of reduced water supply occurring during heightened electricity demand\u0026mdash;underscores the importance of managing hydropower resources effectively.\u003c/p\u003e\n\u003cp\u003eDue to the unique hydropower dispatch characteristics in different countries, the (severity of) meteorological conditions leading to energy droughts vary significantly. In France, the meteorological anomalies during energy drought events are more extreme than in the other countries, as renewable production highly depends on non-dispatchable sources. Conversely, Switzerland, Sweden and Norway have a relatively large reservoir storage capacity that can produce electricity in periods of high demand. Figure\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e shows that energy droughts in Norway are linked to highly negative hydropower reservoir level anomalies, with precise meteorological conditions during the EDW, such as exceptionally low temperatures and wind speeds, playing a secondary role. The depleted reservoir levels are a result of multiple years of imbalance between electricity demand and inflow, often caused by complex and temporally compounding processes.\u003c/p\u003e"},{"header":"Temporally compounding conditions leading to energy droughts","content":"\u003cp\u003eWe distinguish three types of temporally compounding conditions (TCCs) connected to these low reservoir levels and concurrent energy droughts: TCC I) countries with reservoirs that are filled in the spring face vulnerability to summer energy droughts if these are preceded by dry spring conditions, resulting in low reservoir levels and a dry subsoil in the subsequent summer, TCC II) low precipitation or runoff during reservoir filling season compounded with persistent high-pressure systems in autumn and winter leading to winter energy droughts, and TCC III) in certain regions (North-Europe) there is an increased risk of low runoff in spring after a dry and extremely cold winter, increasing the chance of multi-year energy droughts.\u003c/p\u003e\n\u003cp\u003eUnder TCC I, the risks of a summer EDW in regions with considerable cooling demands, particularly Southern Europe, can be up to four times greater than under normal conditions (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003ee). This elevated risk occurs when a notable decline in reservoir inflow and an increase in demand during spring leaves the reservoir levels anomalously low in July and August (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003ea and \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eb, Italy serving as example). Low reservoir inflows occur when the runoff is low (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003ed), which can be attributed to two factors.\u003c/p\u003e\n\u003cp\u003eFirstly, in mountainous regions like the Alpine area, an unusually warm winter with reduced snowfall results in a diminished snowpack by the end of winter. This reduction significantly impacts hydropower generation later that year, as indicated in Supplementary Fig.\u0026nbsp;6. For instance, in the Po River basin of Italy, snow accumulation typically peaks between December and February. As spring arrives, rising temperatures gradually initiate the snowmelt process, releasing water into the Po River and its tributaries. Diminished snowmelt during this period leads to reduced inflow into hydropower reservoirs during spring, affecting hydropower generation in the subsequent months. The Po River drought in the year 2022 serves as a pertinent example of this phenomenon\u003csup\u003e33\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eSecondly, during the spring season, a meteorological drought characterized by reduced precipitation due to persistent high-pressure systems also contributes to decreased inflow. Similar to the patterns observed during the summer EDW, this drought coincides with higher-than-normal temperatures (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003ec).\u003c/p\u003e\n\u003cp\u003eThe electricity system faces increased vulnerability to an energy drought when not only supply is low (depleted reservoirs) but concurrently demand is high (cooling demand during exceptional heat). Dry winter and spring conditions not only leave reservoirs depleted but also result in dry subsoil. Note that the risk of exceptionally hot summers is heightened in these cases, due to this dry subsoil and land-atmosphere feedback mechanisms\u003csup\u003e34,35\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eSo, although by design, the energy drought window lasts 30 days, Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eb demonstrates how the compounding of these conditions can lead to extended periods of high residual loads; On average, the periods around an EDW encompass 104 consecutive days of positive residual load anomalies. Besides the elevated demand during a warmer than normal spring, at the end of a summer, temperatures may return to normal, but the reservoirs have been depleted, resulting in prolonged periods of high residual loads while entering the subsequent winter.\u003c/p\u003e\n\u003cp\u003eUnder TCC II (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003e), the combination of low filling of reservoirs in spring and subsequent persisting high-pressure systems in the following winter, set the stage for winter energy droughts. The risk of a winter energy drought following such conditions is 2.6 times higher in Switzerland and up to 4.4 times higher in the other countries (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003ee). Figure\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003e shows that for Switzerland, like in TCC I, increased temperatures and strongly reduced runoff in May lead to reduced reservoir levels. As winter approaches, prolonged high-pressure systems can reduce precipitation and thereby oppose the replenishment of these reservoirs. Together with increased heating demands associated with these weather systems (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003eb) residual loads can get exceptionally high for long periods of time leading up to winter events. Importantly, the inflow into the reservoirs remains low during subsequent filling seasons after the occurrence of winter EDWs, which works against the recovery and robustness of the entire system.\u003c/p\u003e\n\u003cp\u003eIn TCC III (see Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003e, Sweden serving as example), winter energy droughts are followed by lower-than-usual reservoir inflows during the subsequent spring season. Precipitation, evaporation, and snowmelt after winter EDWs show little to no significant deviation during the filling season when compared to years without such events, but the runoff (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003ed) and consequently reservoir inflow (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003eb) is very low. This phenomenon is predominantly observed in Northern European countries, particularly when winter EDWs are preceded by dry springs and summers (as detailed in TCC II), rendering the electricity system vulnerable, but also leaving the soil dry. This soil moisture drought is exacerbated during the winter EDWs, during which episodes of extreme cold coincide with reduced runoff, as can be observed in Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e.\u003c/p\u003e\n\u003cp\u003eThis dry subsoil (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003ec) is the main cause of continued low runoff in spring. Precipitation following an energy drought initially infiltrates the soil to replenish the moisture deficit before eventually moving via runoff into streamflow and ultimately flowing into the hydropower reservoirs. This particular phenomenon nearly doubles the risk of experiencing extremely low inflow into the reservoirs, opposing the chance for the electricity system to rebound. Figure\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003ea shows that up to 3 years after the occurrence of a winter EDW the mean hydropower reservoirs levels remain anomalously low.\u003c/p\u003e"},{"header":"Prolonged, cumulative droughts","content":"\u003cp\u003eIn four out of the six studied countries we found that as a result of these compounding events, energy droughts are rarely isolated events, but rather a consequence of feedback mechanisms that tend to prolong periods of energy scarcity (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ea) and can turn what might have been a short-term drought into a prolonged, cumulative energy crisis.\u003c/p\u003e \u003cp\u003eThe sequence of the three temporally compounding conditions creates a cyclical pattern, feeding into each other. Typically, low reservoir inflows in the spring season (AMJ) are caused by a thinner snowpack and reduced precipitation. These conditions also lead to relatively dry subsoils, which in turn increase the likelihood of extreme heat events during the summer months. The combination of diminished inflow and increased demand for cooling can lead to an energy shortage in the summer (JA), TCC I. If the low inflows persist, reduced reservoir levels by winter may trigger an energy drought (DJFM), TCC II. Furthermore, in Northern Europe, there is a notable risk of decreased spring runoff (MJ) due to anomalously high residual loads. This is attributed to the low soil moisture conditions originating from previous periods, TCC III, further enhancing the feedback loop.\u003c/p\u003e \u003cp\u003eAt least partially due to this feedback loop, the sets of 10-year climate model simulations that contain one or more energy droughts tend to experience a progressive increase in the residual load anomaly over time (Figs.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eb and \u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ec), suggesting that decadal climate variability plays a role in determining the occurrence of prolonged energy droughts in Europe.\u003c/p\u003e "},{"header":"Discussion and conclusions","content":"\u003cp\u003eThis study highlights the challenge posed by high electricity demands coinciding with periods of reduced renewable electricity production. Within this context, we have identified three distinct types of temporally compounding conditions, providing a framework for early warning systems and the development of mitigation strategies. However, the results suggest that these three TCCs may be an intrinsic part of a cyclical pattern that plays a role in decadal variability in residual load anomalies.\u003c/p\u003e \u003cp\u003eThese analyses demonstrate that relying solely on multiple individual weather years to assess system robustness is inadequate to capture the increased risk and intricate nature associated with multi-year energy droughts. This is particularly relevant for hydroelectrical systems, for which, due to the temporally compounding nature of drought events, evaluating consecutive multi-year data becomes essential for a comprehensive understanding and planning for extreme scenarios.\u003c/p\u003e \u003cp\u003eFor Northern European countries, managing their hydropower reservoirs during winter energy droughts is essential as the risks of periods of low reservoir inflow increase after energy droughts, rendering the system's recovery more challenging. Based on the observation that meteorological conditions leading to electricity system stress often align with periods of anomalously high solar radiation, Southern European countries with large hydropower capacities could potentially mitigate summer energy droughts by the investment in additional solar panels.\u003c/p\u003e \u003cp\u003eWe have focused on the three main types of renewable energy production in Europe. An additional vulnerability of the electricity system is that during periods of reduced discharge (leading to low reservoir inflows) there might be concurrent shortfall in freshwater used for cooling by thermal power plants. This water scarcity aggravates the effects of meteorological conditions on the energy system, but this effect is not incorporated here.\u003c/p\u003e \u003cp\u003eWe based this analysis on a single global climate model that does not incorporate glacier dynamics and melt. Although this model has been assessed against ERA5 for the different regions, it has not been bias corrected. Future research should evaluate the consistency of our findings across multiple climate models.\u003c/p\u003e \u003cp\u003eAdditional next research phases would be to examine how climate change and the energy transition affect long-lasting energy droughts. Increased renewable energy capacities and more intense/frequent weather extremes may lead to even greater challenges in future energy management. The interplay between a changing climate and electricity system adaptations needs analyses to develop strategies mitigating the impacts of extended energy shortages, crucial for shaping sustainable energy policies in an era of climate uncertainty.\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec9\" class=\"Section2\"\u003e\n \u003ch2\u003eMeteorological data\u003c/h2\u003e\n \u003cp\u003eWe use the daily outputs from the KNMI Large ENsemble TIme Slice (KNMI-LENTIS)\u003csup\u003e29\u003c/sup\u003e, a large climate ensemble composed of 160 sets of 10-year climate model simulations of present-day climate (the time slice 2000\u0026ndash;2009). The KNMI-LENTIS ensemble was executed with EC-Earth3, a fully coupled, state-of-the-art global Global Climate Model that includes atmospheric, oceanic, land, and sea-ice components\u003csup\u003e36\u003c/sup\u003e.\u003c/p\u003e\n \u003cp\u003eThis time slice ensemble methodology is especially suitable for investigating extreme events. The 1600 years of data allows for the examination of 160 extreme events that occur once in 10 years or less. Moreover, the sequence of 10 continuous weather years in each simulation makes it possible to study temporally compounding and multi-year events.\u003c/p\u003e\n \u003cp\u003eWe did not bias correct the LENTIS data prior to application in the energy modelling framework. In the context of this multivariate research, bias correction of climate data may unintentionally distort the physical relationships between different variables and the seasons, potentially leading to inaccurate interpretations of climate interactions and dynamics.\u003c/p\u003e\n \u003cp\u003eHowever, to demonstrate the reliability and variance of the LENTIS data in combination with the energy modelling framework, we compared the climatological mean and variance of meteorological variables for each country against the 1991\u0026ndash;2020 ERA5 reanalysis data\u003csup\u003e37\u003c/sup\u003e (Supplementary Fig.\u0026nbsp;7 and Supplementary Fig.\u0026nbsp;8). The discharge peak of Norway, Sweden and Switzerland occurs slightly earlier in the LENTIS data compared to the ERA5 data. However, this affect is mitigated by the dispatch optimization model, which strategically retains water in reservoirs to meet periods of high demand. This approach ensures that, regardless of inflow variations, water release occurs at comparable times annually for both LENTIS and ERA5.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e\n \u003ch2\u003eElectricity production and demand modelling framework\u003c/h2\u003e\n \u003cp\u003eThe electricity modelling framework includes gridded production modules for offshore wind, onshore wind, solar photovoltaics (PV), run-of-river, and hydropower reservoir inflow. It also contains a hydropower dispatch module and an electricity demand module, both operating at the European country level\u003csup\u003e38\u003c/sup\u003e.\u003c/p\u003e\n \u003cp\u003eFor wind and PV solar power modules, the inputs include 10 m wind speed [\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(m {s}^{-1}\\)\u003c/span\u003e\u003c/span\u003e], near-surface temperature [\u0026deg;C], near-surface maximum daily temperature [\u0026deg;C], and solar irradiance [\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(kW {m}^{-2}\\)\u003c/span\u003e\u003c/span\u003e]. We calculate wind power by extrapolating 10 m wind speeds to hub height in both onshore and offshore scenarios using the power law equation. We then estimate the power generation through a power curve, neglecting downtime and efficiency losses. We base PV solar power on solar radiation and temperature-dependent cell efficiency, which we determine using near-surface temperature, maximum daily temperature, and wind speed. We assume that all PV modules are horizontally oriented and functional during all daylight hours.\u003c/p\u003e\n \u003cp\u003eDemand estimation incorporates population-weighted temperature data and historical demand measurement data to create demand fits. These fits account for differences between weekday and weekend demand, though they do not consider cultural influences such as holidays.\u003c/p\u003e\n \u003cp\u003eWe calculate the generation of run-of-river hydropower based on discharge values obtained from runoff [\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(m {timestep}^{-1}\\)\u003c/span\u003e\u003c/span\u003e], which we process through a river routing scheme. In hydropower reservoir systems, we also determine reservoir inflow using this discharge data, and manage outflow with a 4-week moving horizon dispatch optimization, avoiding the benefit of perfect foresight.\u003c/p\u003e\n \u003cp\u003eA detailed description of all the modules within the modeling framework, along with a validation of how it performs on the selection of energy droughts with the use of ERA5 data, are available in Ref \u003csup\u003e28\u003c/sup\u003e.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\n \u003ch2\u003eInstalled capacities\u003c/h2\u003e\n \u003cp\u003eWe conducted the study using gridded files that reflect Europe\u0026rsquo;s 2020 installed renewable capacities. The locations of solar photovoltaic and onshore wind farms were determined based on the 2020 data from OpenStreetMap\u003csup\u003e39\u003c/sup\u003e, EMODnet provides vector data of installed offshore wind capacities\u003csup\u003e40\u003c/sup\u003e, and we sourced the locations of run-of-river and reservoir hydropower plants from the Joint Research Centre (JRC) Hydropower database\u003csup\u003e41\u003c/sup\u003e. These datasets were converted into gridded capacity files according to the methods described in Ref \u003csup\u003e28\u003c/sup\u003e.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e\n \u003ch2\u003eSelection of drought events\u003c/h2\u003e\n \u003cp\u003eOur analysis targets energy drought events that have a return period of 10 years or longer. To this end, we selected the most significant 160 events from our 1600-year database. Persistent Energy Droughts (PEDs) are defined as periods where the residual load exceeds the 97th percentile threshold for several consecutive days. A PED concludes when the load falls below this threshold for more than three consecutive days. The PEDs are ranked in descending order of duration, focusing on the 160 longest events to identify the most extreme cases. Differentiation between summer, winter, and all-year events is made, with each category benchmarked against their specific 97th percentile threshold.\u003c/p\u003e\n \u003cp\u003eTo evaluate the effects of meteorological compounding conditions in these events, we analyzed same-length events in the form of Energy Drought Windows (EDWs) that span 30 days, or are confined to a specific month. This selected duration aligns with the approximate average length of PEDs as observed across the countries included in our analysis. The 160 EDWs with the highest average residual loads (and without overlap) are selected to identify the most impactful energy droughts.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e\n \u003ch2\u003eRisk ratios\u003c/h2\u003e\n \u003cp\u003eWe quantify and compare relative risks of energy droughts under distinct temporally compounding meteorological conditions using calculated risk ratios (RRs): the ratio of the probability of an energy drought occurring after certain meteorological conditions (e.g., low inflow into hydropower reservoirs in spring) to the climatological probability. We assessed the statistical significance and precision of the estimated RR by computing its 95% confidence interval (CI) according to \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(CI= \\text{exp}(\\text{log}RR\\pm Z x SE(\\text{log}RR\\left)\\right)\\)\u003c/span\u003e\u003c/span\u003e, where RR is the risk ratio, SE is the standard error of the log-transformed risk ratio, and Z represents the standard normal deviate (1.96 for a 95% CI). Risk ratios for which the 95% CI does not include 1 are considered statistically significant.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e\n \u003ch2\u003eStandardized Soil Moisture Index (SSMI)\u003c/h2\u003e\n \u003cp\u003eWe analyze soil moisture variability after winter energy droughts with the Standardized Soil Moisture Index (SSMI)\u003csup\u003e42\u003c/sup\u003e. The SSMI uses the z-score methodology to quantify soil moisture deviations in terms of standard deviations from the mean and is calculated using the formula: \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(SSMI=\\left(mrso- \\mu \\right)/std\\)\u003c/span\u003e\u003c/span\u003e, where \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(mrso\\)\u003c/span\u003e\u003c/span\u003e represents the soil moisture value for a given day, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\mu\\)\u003c/span\u003e\u003c/span\u003e is the mean soil moisture over the 1600-year period on that day of year, and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(std\\)\u003c/span\u003e\u003c/span\u003e is the standard deviation of soil moisture during that day of year.\u003c/p\u003e\n\u003c/div\u003e"},{"header":"Declarations","content":"\u003ch2\u003eData availability\u003c/h2\u003e\n\u003cp\u003eAll analysis data are included within the article and its Supplementary Information files. The simulated energy generation and demand data supporting the findings of this study will be made available after the review.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eYalew, S. G. \u003cem\u003eet al.\u003c/em\u003e Impacts of climate change on energy systems in global and regional scenarios. \u003cem\u003eNat. Energy\u003c/em\u003e \u003cstrong\u003e5\u003c/strong\u003e, 794\u0026ndash;802 (2020).\u003c/li\u003e\n\u003cli\u003eStaffell, I. \u0026amp; Pfenninger, S. The increasing impact of weather on electricity supply and demand. \u003cem\u003eEnergy\u003c/em\u003e \u003cstrong\u003e145\u003c/strong\u003e, 65\u0026ndash;78 (2018).\u003c/li\u003e\n\u003cli\u003eVan Der Wiel, K., Selten, F. M., Bintanja, R., Blackport, R. \u0026amp; Screen, J. A. Ensemble climate-impact modelling: extreme impacts from moderate meteorological conditions. \u003cem\u003eEnviron. Res. Lett.\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, (2020).\u003c/li\u003e\n\u003cli\u003eZscheischler, J. \u003cem\u003eet al.\u003c/em\u003e A typology of compound weather and climate events. \u003cem\u003eNature Reviews Earth and Environment\u003c/em\u003e vol. 1 (2020).\u003c/li\u003e\n\u003cli\u003eCopernicus Climate Change Service (C3S). Climate bulletins. https://climate.copernicus.eu/climate-bulletins (2021).\u003c/li\u003e\n\u003cli\u003eMorison, R. \u0026amp; Shiryaevskaya, A. U.K. Power Surges to Record 400 Pounds as Wind Fails to Blow. \u003cem\u003eBloomberg\u003c/em\u003e (2021).\u003c/li\u003e\n\u003cli\u003eFern\u0026aacute;ndez, L. Hydroelectricity generation Spain 2010-2022. \u003cem\u003eStatista\u003c/em\u003e https://www.statista.com/statistics/1006362/hydroelectricity-generation-in-spain/ (2023).\u003c/li\u003e\n\u003cli\u003eIngram, E. Hydroelectric generation in Italy decreased 37.7% in 2022. \u003cem\u003eHydro review\u003c/em\u003e (2023).\u003c/li\u003e\n\u003cli\u003eCopernicus Climate Change Service (C3S). European State of the Climate 2022 Unprecedented extreme heat and widespread drought mark European climate in 2022. \u003cem\u003eCopernicus\u003c/em\u003e https://climate.copernicus.eu/copernicus-european-state-climate-2022-unprecedented-extreme-heat-and-widespread-drought-mark (2023).\u003c/li\u003e\n\u003cli\u003eGualtieri, T. Drought Forces One of Spain\u0026rsquo;s Largest Hydro Plants to Halt. \u003cem\u003eBloomberg\u003c/em\u003e (2022).\u003c/li\u003e\n\u003cli\u003eInternational Energy Agency (IEA). Hydropower Data Explorer. \u003cem\u003eIEA\u003c/em\u003e (2021).\u003c/li\u003e\n\u003cli\u003ePerera, A. T. D., Nik, V. M., Chen, D., Scartezzini, J. L. \u0026amp; Hong, T. Quantifying the impacts of climate change and extreme climate events on energy systems. \u003cem\u003eNat. Energy\u003c/em\u003e \u003cstrong\u003e5\u003c/strong\u003e, 150\u0026ndash;159 (2020).\u003c/li\u003e\n\u003cli\u003evan der Wiel, K. \u003cem\u003eet al.\u003c/em\u003e Meteorological conditions leading to extreme low variable renewable energy production and extreme high energy shortfall. \u003cem\u003eRenew. Sustain. Energy Rev.\u003c/em\u003e \u003cstrong\u003e111\u003c/strong\u003e, 261\u0026ndash;275 (2019).\u003c/li\u003e\n\u003cli\u003eOtero, N., Martius, O., Allen, S., Bloomfield, H. \u0026amp; Schaefli, B. Characterizing renewable energy compound events across Europe using a logistic regression-based approach. \u003cem\u003eMeteorol. Appl.\u003c/em\u003e \u003cstrong\u003e29\u003c/strong\u003e, (2022).\u003c/li\u003e\n\u003cli\u003eJurasz, J., Mikulik, J., Dabek, P. B., Guezgouz, M. \u0026amp; Kaźmierczak, B. Complementarity and \u0026lsquo;resource droughts\u0026rsquo; of solar and wind energy in poland: An era5-based analysis. \u003cem\u003eEnergies\u003c/em\u003e \u003cstrong\u003e14\u003c/strong\u003e, (2021).\u003c/li\u003e\n\u003cli\u003eKay, G. \u003cem\u003eet al.\u003c/em\u003e Variability in North Sea wind energy and the potential for prolonged winter wind drought. \u003cem\u003eAtmos. Sci. Lett.\u003c/em\u003e \u003cstrong\u003e24\u003c/strong\u003e, (2023).\u003c/li\u003e\n\u003cli\u003eOtero, N., Martius, O., Allen, S., Bloomfield, H. \u0026amp; Schaefli, B. A copula-based assessment of renewable energy droughts across Europe. \u003cem\u003eRenew. Energy\u003c/em\u003e \u003cstrong\u003e201\u003c/strong\u003e, (2022).\u003c/li\u003e\n\u003cli\u003eJahns, C., Osinski, P. \u0026amp; Weber, C. A statistical approach to modeling the variability between years in renewable infeed on energy system level. \u003cem\u003eEnergy\u003c/em\u003e \u003cstrong\u003e263\u003c/strong\u003e, (2023).\u003c/li\u003e\n\u003cli\u003eRaynaud, D., Hingray, B., Fran\u0026ccedil;ois, B. \u0026amp; Creutin, J. D. Energy droughts from variable renewable energy sources in European climates. \u003cem\u003eRenew. Energy\u003c/em\u003e \u003cstrong\u003e125\u003c/strong\u003e, 578\u0026ndash;589 (2018).\u003c/li\u003e\n\u003cli\u003eBloomfield, H. C. \u003cem\u003eet al.\u003c/em\u003e Quantifying the sensitivity of european power systems to energy scenarios and climate change projections. \u003cem\u003eRenew. Energy\u003c/em\u003e \u003cstrong\u003e164\u003c/strong\u003e, 1062\u0026ndash;1075 (2020).\u003c/li\u003e\n\u003cli\u003eCraig, M. T. \u003cem\u003eet al.\u003c/em\u003e Overcoming the disconnect between energy system and climate modeling. \u003cem\u003eJoule\u003c/em\u003e 1405\u0026ndash;1417 (2022) doi:10.1016/j.joule.2022.05.010.\u003c/li\u003e\n\u003cli\u003eGrochowicz, A., van Greevenbroek, K., Benth, F. E. \u0026amp; Zeyringer, M. Intersecting near-optimal spaces: European power systems with more resilience to weather variability. \u003cem\u003eEnergy Econ.\u003c/em\u003e \u003cstrong\u003e118\u003c/strong\u003e, (2023).\u003c/li\u003e\n\u003cli\u003eVan Vliet, M. T. H., Sheffield, J., Wiberg, D. \u0026amp; Wood, E. F. Impacts of recent drought and warm years on water resources and electricity supply worldwide. \u003cem\u003eEnviron. Res. Lett.\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, (2016).\u003c/li\u003e\n\u003cli\u003eTsanis, I. \u0026amp; Tapoglou, E. Winter North Atlantic Oscillation impact on European precipitation and drought under climate change. \u003cem\u003eTheor. Appl. Climatol.\u003c/em\u003e \u003cstrong\u003e135\u003c/strong\u003e, 323\u0026ndash;330 (2019).\u003c/li\u003e\n\u003cli\u003eEly, C. R., Brayshaw, D. J., Methven, J., Cox, J. \u0026amp; Pearce, O. Implications of the North Atlantic Oscillation for a UK-Norway Renewable power system. \u003cem\u003eEnergy Policy\u003c/em\u003e \u003cstrong\u003e62\u003c/strong\u003e, (2013).\u003c/li\u003e\n\u003cli\u003eRakovec, O. \u003cem\u003eet al.\u003c/em\u003e The 2018\u0026ndash;2020 Multi-Year Drought Sets a New Benchmark in Europe. \u003cem\u003eEarth\u0026rsquo;s Futur.\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, (2022).\u003c/li\u003e\n\u003cli\u003evan der Wiel, K., Batelaan, T. J. \u0026amp; Wanders, N. Large increases of multi-year droughts in north-western Europe in a warmer climate. \u003cem\u003eClim. Dyn.\u003c/em\u003e \u003cstrong\u003e60\u003c/strong\u003e, (2023).\u003c/li\u003e\n\u003cli\u003evan der Most, L. \u003cem\u003eet al.\u003c/em\u003e Extreme events in the European renewable power system: Validation of a modeling framework to estimate renewable electricity production and demand from meteorological data. \u003cem\u003eRenew. Sustain. Energy Rev.\u003c/em\u003e \u003cstrong\u003e170\u003c/strong\u003e, 112987 (2022).\u003c/li\u003e\n\u003cli\u003eMuntjewerf, L., Bintanja, R., Reerink, T. \u0026amp; van der Wiel, K. The KNMI Large Ensemble Time Slice (KNMI-LENTIS). \u003cem\u003eGeosci. Model Dev.\u003c/em\u003e \u003cstrong\u003e16\u003c/strong\u003e, 4581\u0026ndash;4597 (2023).\u003c/li\u003e\n\u003cli\u003eOhlendorf, N. \u0026amp; Schill, W. P. Frequency and duration of low-wind-power events in Germany. \u003cem\u003eEnviron. Res. Lett.\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, (2020).\u003c/li\u003e\n\u003cli\u003eBloomfield, H. C., Suitters, C. C. \u0026amp; Drew, D. R. Meteorological Drivers of European Power System Stress. \u003cem\u003eJ. Renew. Energy\u003c/em\u003e \u003cstrong\u003e2020\u003c/strong\u003e, 1\u0026ndash;12 (2020).\u003c/li\u003e\n\u003cli\u003eKautz, L. A. \u003cem\u003eet al.\u003c/em\u003e Atmospheric blocking and weather extremes over the Euro-Atlantic sector - A review. \u003cem\u003eWeather and Climate Dynamics\u003c/em\u003e vol. 3 (2022).\u003c/li\u003e\n\u003cli\u003eMontanari, A. \u003cem\u003eet al.\u003c/em\u003e Why the 2022 Po River drought is the worst in the past two centuries. \u003cem\u003eSci. Adv.\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, (2023).\u003c/li\u003e\n\u003cli\u003eFischer, E. M., Seneviratne, S. I., L\u0026uuml;thi, D. \u0026amp; Sch\u0026auml;r, C. Contribution of land-atmosphere coupling to recent European summer heat waves. \u003cem\u003eGeophys. Res. Lett.\u003c/em\u003e \u003cstrong\u003e34\u003c/strong\u003e, (2007).\u003c/li\u003e\n\u003cli\u003eMiralles, D. G., Gentine, P., Seneviratne, S. I. \u0026amp; Teuling, A. J. Land\u0026ndash;atmospheric feedbacks during droughts and heatwaves: state of the science and current challenges. \u003cem\u003eAnn. N. Y. Acad. Sci.\u003c/em\u003e \u003cstrong\u003e1436\u003c/strong\u003e, 19\u0026ndash;35 (2019).\u003c/li\u003e\n\u003cli\u003eD\u0026ouml;scher, R. \u003cem\u003eet al.\u003c/em\u003e The EC-Earth3 Earth system model for the Coupled Model Intercomparison Project 6. \u003cem\u003eGeosci. Model Dev.\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, (2022).\u003c/li\u003e\n\u003cli\u003eHersbach, H. \u003cem\u003eet al.\u003c/em\u003e The ERA5 global reanalysis. \u003cem\u003eQ. J. R. Meteorol. Soc.\u003c/em\u003e \u003cstrong\u003e146\u003c/strong\u003e, 1999\u0026ndash;2049 (2020).\u003c/li\u003e\n\u003cli\u003eEurostat. NRG-IND-REN Share of energy from renewable sources. \u003cem\u003eEurostat\u003c/em\u003e https://ec.europa.eu/eurostat/databrowser/product/view/NRG_IND_REN?lang=en (2021).\u003c/li\u003e\n\u003cli\u003eDunnett, S., Sorichetta, A., Taylor, G. \u0026amp; Eigenbrod, F. Harmonised global datasets of wind and solar farm locations and power. \u003cem\u003eSci. Data\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, 1\u0026ndash;12 (2020).\u003c/li\u003e\n\u003cli\u003eEMODnet Human Activities project. Emodnet_HA_WindFarms_20200305 [data set].\u003c/li\u003e\n\u003cli\u003eEuropean Commission Joint Research Centre (JRC). JRC Hydro-power database. (2019) doi:https://doi.org/10.5281/zenodo.3862722 [data set].\u003c/li\u003e\n\u003cli\u003eHao, Z. \u0026amp; AghaKouchak, A. Multivariate Standardized Drought Index: A parametric multi-index model. \u003cem\u003eAdv. Water Resour.\u003c/em\u003e \u003cstrong\u003e57\u003c/strong\u003e, (2013).\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-3796061/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-3796061/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"As Europe's renewable energy capacities expand, electricity systems face increased risks of energy droughts—periods of low production coinciding with high demand. We evaluate characteristics of electricity variability due to weather variations by calculating 1600 years of daily production and demand. Focusing on six European countries—chosen for their energy mix including hydropower—we find that energy droughts result from processes that cause (temporally) compounding impacts in the energy and meteorological system. These can turn what might have been short-term droughts into prolonged, cumulative energy crises. For instance, low reservoir inflows in spring quadruple the chance of prolonged energy droughts: reduced snowpack and rainfall lower hydro-availability but also dry-out subsoils, increasing the chance of heatwaves and therewith extending the energy problems into summer. We identify and quantify three compounding energy/climate conditions and the associated characteristics and risks of multiyear energy droughts, crucial for future energy system design and policies.","manuscriptTitle":"Temporally compounding energy droughts in European electricity systems with hydropower","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-01-10 19:21:09","doi":"10.21203/rs.3.rs-3796061/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-energy","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"nenergy","sideBox":"Learn more about [Nature Energy](http://www.nature.com/nenergy/)","snPcode":"","submissionUrl":"","title":"Nature Energy","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Research","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"66a78100-7f14-4cf4-9b18-12f66381fd60","owner":[],"postedDate":"January 10th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":28029763,"name":"Physical sciences/Energy science and technology/Energy modelling"},{"id":28029764,"name":"Physical sciences/Energy science and technology/Renewable energy"},{"id":28029765,"name":"Earth and environmental sciences/Climate sciences"}],"tags":[],"updatedAt":"2024-09-28T07:12:50+00:00","versionOfRecord":{"articleIdentity":"rs-3796061","link":"https://doi.org/10.1038/s41560-024-01640-5","journal":{"identity":"nature-energy","isVorOnly":false,"title":"Nature Energy"},"publishedOn":"2024-09-27 04:00:00","publishedOnDateReadable":"September 27th, 2024"},"versionCreatedAt":"2024-01-10 19:21:09","video":"","vorDoi":"10.1038/s41560-024-01640-5","vorDoiUrl":"https://doi.org/10.1038/s41560-024-01640-5","workflowStages":[]},"version":"v1","identity":"rs-3796061","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-3796061","identity":"rs-3796061","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}