Reliable energy datasets form the backbone of the C3S Energy service, supporting the calibration and validation of the climate-driven energy conversion models. These datasets ensure that the resulting energy indicators accurately reflect real-world system behaviour across technologies and regions.
An extensive search was conducted to collect energy-related data with global coverage, focusing on power plant installations, electricity generation, and electricity demand.
These datasets are used to support the modelling of wind, solar photovoltaic, and hydropower indicators, as well as electricity and energy demand. Specifically, they are employed to:
Train statistical and machine learning models (e.g., for hydropower and demand)
Calibrate physical models (e.g., for wind and solar generation)
Validate model outputs against observed generation and demand patterns
The selection of data sources was based on criteria including spatial and temporal coverage, update frequency, and the granularity of information provided.
Key Data Sources
Table 1.1 summarises the main energy-related datasets used in the development of the C3S Energy service, including their geographic coverage, general description, and specific role in the modelling, training, or validation of the energy indicators.
Table 1.1: Summary of key energy datasets used in the C3S Energy service, including their description, geographic coverage, and role in model calibration, training, or validation for energy indicators.
| Source | Description | Coverage | Use in C3S Energy service |
|---|---|---|---|
| The Wind Power (commercial) | Commercial database of global wind power installations, including plant locations, turbine specifications, and power curves. | Global | Used to define turbine characteristics (hub height, rotor diameter, power curves) and plant locations for wind power conversion modelling. |
| Global Energy Monitor (GEM) | Open-source dataset tracking global wind, solar, and hydro power infrastructure, with plant-level detail. | Global | Used to map the location and capacity of hydropower plants for the Installed Capacity Weighted Precipitation (IWP) model. |
| ENTSO-E Transparency Platform | Official platform from the European network of Transmission System Operators (TSOs) providing electricity market and infrastructure data. | Pan-European (PECD) | Provides high-resolution generation, installed capacity, and reservoir data used to train and validate the European hydropower Random Forest model. In the electricity demand model, it supplies hourly and daily national load data, which serve as the target variable in statistical learning algorithms linking demand with climate drivers such as temperature, solar radiation, and wind speed. |
| EMBER | International research initiative that compiles and publishes monthly electricity generation and demand statistics across a broad set of countries. | >85 countries | Used to validate hydropower and electricity demand indicators outside Europe, especially in data-scarce regions. |
| International Energy Agency (IEA) | International organisation that provides comprehensive energy statistics and modelling tools, including specifications for renewable energy technologies. | >40 countries | Supplies technical specifications for a representative 15 MW offshore wind turbine (IEA 15MW_240_RWT) used in the wind power model. Also used to validate hydropower generation indicators. |
| International Renewable Energy Agency (IRENA) | Global agency compiling statistics on renewable energy capacity, generation, and policy frameworks. | Global | Supports validation of long-term trends in hydropower and solar PV indicators; used for cross-checking capacity data. |
| National Renewable Energy Laboratory (NREL) | U.S. research lab providing open-access technical data on renewable technologies, including turbine design studies. | Global | Supplies specifications for a future onshore wind turbine (Bespoke 6 MW) used in scenario-based wind power modelling. |
| National Statistics | Country-specific official energy databases, often providing high-resolution capacity and generation data. | Country-specific | Used to validate or refine solar PV outputs where available, especially when international datasets lack coverage or granularity. |
| Private PV Plant Metadata | Technical data (tilt, azimuth, mounting structure) extracted from large-scale PV plant installations. | Germany, France | Used to generalise installation configuration rules for the global application of the solar PV model. |
Data Availability and Limitations
While global datasets on wind and solar installations are generally available, consistent and long-term time series for electricity generation and demand remain scarce, especially outside Europe. This poses challenges for modelling hydropower and electricity demand, both of which require robust historical records for statistical training.
To overcome this limitation, the C3S Energy service adopts a dual-modelling approach tailored to data availability:
In Europe, where detailed and high-frequency data are available, advanced statistical and machine learning models are employed. For example, a Random Forest model is used to simulate hydropower generation, and a Generalised Additive Model (GAM) is applied to estimate electricity demand.
Globally (including Europe), where data are sparse or of lower resolution, simplified proxy methods are used. Hydropower is estimated using the Installed capacity Weighted Precipitation (IWP), while electricity demand is approximated based on Energy Degree Days (EDD), which reflect heating and cooling needs derived from temperature patterns.
This tiered approach ensures that energy indicators are produced consistently across all regions while adapting the modelling complexity to the quality of available data.
For further methodological details, see the dedicated pages: Hydropower Conversion Models and Electricity and Energy Demand Models.
