Back to opportunities | Category Index
These are the intended use-case/justification for one or multiple variable groups. Opportunities are linked to relevant experiment groups. Identifying opportunities helps to provide a structure to map variables against requirements. Each opportunity description will convey why this combination of variables and experiments is important and how they contribute to impact.
| Attribute | Value |
|---|---|
| description | This opportunity includes variable groups providing information needed for impacts and adaptation planning in the energy sector. Included are variable groups related to the minimum and extended inputs required to support core activities in renewables assessment and energy system modelling. An emphasis is placed on core sets of high-frequency (sub-daily) outputs as these are widely viewed as essential for high-quality onward impact modelling of energy system infrastructure, supported by addition outputs grouped by sector component or application. This opportunity requests data throughout the entire simulation period. However, if high temporal resolution data is difficult to provide for the entire period, this opportunity can be supported with high resolution data from the following time subsets: hist72 and histext + whole timeseries (2021-2100): first preference, hist43 and histext + whole timeseries (2021-2100): second preference, hist43 and histext + scenario30mid + scenario20: third preference, hist43 and histext + scenario20mid + scenario20: fourth preference, hist20 and histext + scenario20mid + scenario20: fifth preference. histext is included because the I&A community would like historical data that are as close to the present as possible. |
| desirable_ensemble_size | 10 |
| expected_impacts | This opportunity is therefore to transform the 'state of the art' in energy-climate modelling. By better linking with the modelling needs of the energy community, this opportunity will support improved climate impact risk assessment and adaptation in a sector which is critical to achieving international decarbonisation goals. As there are a very wide range of potential uses of the data, broad outcomes are presented first followed by a brief explanation of the selected variable groups. - Support for future decarbonisation in energy systems at regional, national and international scale (e.g power system operations, planning and integration of renewables, energy storage and transmission requirements, assessing heating/cooling load). - Climate adaptation and resilience assessment (e.g., nuclear site risk assessment, thermal power plant availability, siting of offshore wind power, risk of ‘energy droughts’, network adequacy assessment). - Energy demand in built environment (e.g., use of heat-pumps, air conditioning, building comfort, health impacts in heat/cold-waves) - Extreme event assessment (e.g., damage to infrastructure, risk to offshore wind) - Supporting climate services to the energy sector _impacts\_climserv\_hourly\_core_ - minimum essential variables for detailed modelling of energy system planning, operations and risk assessment for high-levels of renewables (wind and solar) and, to some extent, demand from the built environment. _impacts\_climserv\_hourly\_expanded_ - combined with _impacts\_climserv\_hourly\__core, these enable more advanced models of power system components to be used. _Impacts\_energy\_hydropower_ - allows minimal models of hydropower availability to be constructed (usually alongside _impacts\_climserv\_hourly\_core_). This is vital for countries/regions where run-of-river or storage hydropower meet a significant fraction of their energy needs (examples include Norway, China, Brazil, Uruguay, Kenya, USA). _ _ _Impacts\_energy\_building\_demand _– combined with_ impacts\_climserv\_hourly\_core _enables enhanced modelling of demand for heating / cooling as well as electrified supply (via air, water, or ground-source heat pumps) and the energy consumption of electric vehicles. _Impact\_enegy\_damaging\_wind _and _impact\_energy\_damaging\_other_ provide simple time-aggregated metrics to estimate the risk of damage to key energy system components. _Impacts\_energy\_river\_temperature_ - thermal power plants (including nuclear) provide vital services to electricity systems around the world and are likely to continue to do so throughout the process of decarbonisation. This data is important for modelling the availability of cooling water for these plants to maintain operation and assess future climate risk to power sector. |
| justification_of_resources | Decarbonising energy is central to climate mitigation. Global investment in clean energy reached US$2 trillion in 2024 (IEA World Energy Investment 2024) and this new infrastructure has multi-decadal life-spans. It is therefore essential that good decisions are made. However, many of the changes involved are increasing the exposure of the energy system to weather and climate (e.g. more renewables, electrification of heating). Anticipating and understanding climate risks to energy is therefore vital to for creating a resilient, clean and affordable future. Energy systems have highly complex dependencies on weather due to their highly interconnected nature. Supply (generation) must be balanced with load (demand) across complex geographical networks in near-real time. High-frequency multi-variate time-series are essential for estimating renewable resource variations (e.g., solar, wind, temperature) across an extended spatial domain. This information cannot be readily reconstructed from coarser-frequency climate information (see FAQs 3-5 & 7 below for further detail). A critical minimum set of a handful of surface / near-surface climate variables are identified as being crucial to many power system modelling applications, with hourly-output commonly seen as the minimum frequency required to support detailed assessment of renewables integration (though 3-hourly can also be used in some applications, it introduces significant challenges and compromises for many energy modelling groups – see FAQs 1, 2 & 7 below). A wider set of additional variables – many of which can be provided at lower frequency – enable a broader and more comprehensive view to be taken across a broader range of technologies (e.g., thermal power plants, building energy demand, hydropower). Core outputs are anticipated in terms of characterising rare/extreme events for present and future energy systems, understanding of long-term climate variability (years, decades), and assessing the impacts of projected climate change scenarios. All experiments listed in “impact_climserv_core_expt” requested to provide data throughout entire simulation. Historic experiments (hist* and esm-hist) can use hist72 timeslice. Future scenarios to prioritize data output period to 2100, but extensions beyond this are welcomed where possible (as a lower priority but particularly to 2150 for hydropower applications). Frequently asked questions about this request: 1. Are all these different outputs really needed (is it a ‘credible minimum set’ of variables and experiments)? We have sought to avoid requesting variables which can be easily derived or estimated from each other, and limited the experiments requested to the most important ones for energy impact research. While many energy modellers would seek higher frequency and many more variables, the request seeks to balance their wishes against feasibility of production (see FAQ 7 below). If hourly data cannot be provided then 3-hourly or 6-hourly versions of the same variables would enable many groups to begin to work with the data but at a scientific cost (lower quality modelling of energy systems) and a community cost (fewer energy research groups will be able to work with the more complex data-processing tasks involved, see FAQs 2 & 8 below). The variables requested are a good approximation to the “minimum set” required to support this community and will go a long way to “Overcoming the disconnect between energy and climate modelling” https://doi.org/10.1016/j.joule.2022.05.010. 2. Is there a community who will use this data? There is an international modelling community ready to support the uptake of this output. Equivalent datasets derived from reanalyses are already used by many leading research groups internationally in academia, industry and policy relating to energy (e.g., https://doi.org/10.1016/j.rser.2021.111614). 3. Can’t the GCMs be downscaled statistically to provide higher frequency/resolution? Statistical downscaling of GCMs is not an alternative. Many energy-climate applications require high-frequency time-series which are ‘self-consistent’ (in space, time and across variables). This rules out many standard statistical downscaling methods. 4. Shouldn’t initiatives like CORDEX provide this high frequency/resolution data for impact models rather than CMIP? Dynamical downscaling of GCMs is not a like-for-like alternative to the wider CMIP ensemble. CORDEX, for example, lags CMIP in time (delaying energy-climate impact assessment) and only downscales a sub-set of GCMs (limiting multi-model risk assessment). It is also very expensive and does not provide universal coverage (in many regions of the world CMIP not CORDEX will be available as an extensive multi-model ensemble). 5. Could RCM emulators provide the required high-frequency data? RCM emulators for downscaling may be an alternative in the long-term but remain experimental and access to training data is highly unequal. While these tools are incredibly promising (e.g., Sup3rCCC), research and development of these techniques remains at an early stage. Deployment of RCM emulators pre-supposes the availability of suitable high-frequency and high-resolution training data (e.g., an extensive dynamical RCM downscaling of reanalysis) which is not available in many regions of the world where energy systems are changing most rapidly. 6. Can high-frequency GCM output be provided only for short timeslices or single ensemble members rather than the whole simulation period across an initial condition ensemble? Providing short timeslices or single ensemble members is not sufficient for robust impact assessment. Energy infrastructure has a long life-span (decades) and it is therefore relevant to understand how it is impacted by long-term climate variability and uncertainty. For example, a recent study showed that life-time output from wind-farms in Germany could vary by as much as 10% based on which 20-year sample of reanalysis is selected: this has profound implications for the optimal siting and composition of European renewable assets https://doi.org/10.5194/wes-4-515-2019. Multi-decadal (and ideally multi-realisation ensembles) samples are required for robust impact assessment and understanding the contributions of natural and anthropogenically-Data Request's Gridforced future climate scenarios, just as would be expected when assessing more “traditional” climate science metrics (e.g., heatwaves or storm-track/jet movements). 7. Does hourly grid-point output from GCMs actually contain information beyond simply interpolating 3-hourly output? There is evidence to suggest that hourly output CMIP-class GCMs contains information that cannot be recovered from coarser (3h/6h) output. Examination of previous CMIP ensembles (particularly HighResMIP via the EU H2020 PRIMAVERA project) confirms that there is “information” in hourly gridded GCM outputs (https://ui.adsabs.harvard.edu/abs/2018EGUGA..20.7784G/abstract, following similar methods previous applied to reanalyses here https://doi.org/10.1016/j.renene.2014.10.024). Notwithstanding this, the provision of hourly output is vital for creating a strong connection with energy modelling. It naturally “plugs in” to many onward energy models, substantially reducing the burden of technical expertise for users by avoiding the need for up-sampling in time. |
| lead_theme | Impacts & Adaptation |
| minimum_ensemble_size | 1 |
| name | Energy System Impacts |
| opportunity_id | 22 |
| technical_notes | If data volume is a concern, the following priorities can be taken into account: 1. Scenario extensions: these are slightly lower priority than the 'regular' scenarios (through to 2100). If the extensions are performed, daily data is the higher priority. 2. Time slices: hist72 and histext + whole timeseries (2021-2100): first preference, hist43 and histext + whole timeseries (2021-2100): second preference, hist43 and histext + scenario30mid + scenario20: third preference, hist43 and histext + scenario20mid + scenario20: fourth preference, hist20 and histext + scenario20mid + scenario20: fifth preference. histext is included because the I&A community would like historical data that are as close to the present as possible. |
| data_request_themes | Impacts & Adaptation, Atmosphere, Land & Land-Ice, Ocean & Sea-Ice |
| experiment_groups | impact_climserv_core_expt, scenarios_extensions-low-medium-high, highresmip2-ia, scenarios |
| mips | DCPP |
| time_subsets | hist20, hist43, hist72, histext, scenario20, scenario20mid, scenario30mid |
| variable_groups | impact_energy_building_demand, impact_energy_damaging_other, impact_energy_hydropower, Impact_energy_river_temperature, impacts_climserv_hourly_core, impacts_climserv_hourly_expanded, impacts_energy_damaging_wind |