Objectives of the service
ATLAScast delivers hyper-local weather intelligence that translates atmospheric variability into operationally useful forecasts for renewable energy generation and energy-intensive operations. Users face significant financial and operational risks due to forecast errors, rapid weather-driven power fluctuations, and limited visibility of site-specific effects such as cloud movement, terrain, shading, and exposure.
The service provides site-resolved, short-range weather forecasts tailored to individual assets, including solar farms, wind farms, battery energy storage systems, and data centres. By combining satellite-based observations, numerical weather prediction context, and site geometry, ATLAScast produces actionable outputs such as power ramp risks, derating probabilities, cooling load forecasts, and uncertainty bands to support operational decision-making.
The objective of the project is to develop, validate, and demonstrate ATLAScast with real operational users, starting with high-value, data-rich sites. The activity focuses on proving forecast accuracy improvements, quantifying operational and economic benefits, and delivering a scalable service through live data feeds and performance reporting.
Users and their needs
ATLAScast targets professional users responsible for energy generation, flexibility, and operational resilience.
Primary user communities (Phase 1)
- Utility-scale solar, wind, and hybrid renewable operators
- Battery energy storage system operators
- Energy suppliers managing imbalance risk
- Data centre operators and facilities managers
User needs
- Accurate, site-resolved short-range weather forecasts
- Early warning of power ramps and derating risks
- Improved scheduling of maintenance and flexibility
- Reduced imbalance exposure and operational costs
- Better forecasting of cooling demand under temperature and humidity extremes
Project challenges
- Delivering reliable forecasts at asset-scale resolution
- Translating weather data into operational decision variables
- Integrating forecasts with existing operational systems
Target countries
Initial users are expected in European countries, focusing on regions with high penetration of renewable energy and grid-constrained infrastructure.
Service/ system concept
The service provides highly accurate, site-specific weather and environmental forecasts to support planning, construction, and operation of renewable energy assets. Users receive real-time and short-term predictions on key parameters such as solar radiation, wind, temperature, humidity, and precipitation, delivered via a web platform or API. Key features include forecast visualisation, automated alerts, and integration into existing energy management systems, enabling users to optimise site selection, improve operational efficiency, reduce risk, and lower costs.
In simple terms, the system works by combining data from satellites, local sensors, and weather stations. Satellites provide a wide view of weather patterns, while ground sensors capture precise local conditions. This data is continuously transmitted and processed using artificial intelligence models, which learn patterns and generate accurate, location-specific forecasts.
At a high level, the system architecture consists of three layers:
- Data collection (satellites, IoT sensors, weather stations),
- Data processing and modelling (AI-driven analytics platform), and
- Delivery layer (web interface and API).
This integrated approach ensures forecasts are both broad in coverage and highly localised, enabling better, data-driven decision-making.
Space Added Value
ATLAScast uses space-based Earth observation data, including high-frequency satellite imagery, as a core input for detecting and forecasting atmospheric features such as cloud development and movement. These space assets provide wide-area, consistent coverage and temporal refresh rates that cannot be achieved through ground-based measurements alone.
The added value of space assets lies in their ability to resolve rapidly evolving, localised weather phenomena that drive power variability at renewable energy sites. When combined with numerical weather prediction models, site geometry, and artificial intelligence, satellite data enables forecasts that account for horizon effects, shading, surface roughness, and local exposure.
Compared to conventional kilometre-scale forecasts, this approach delivers asset-aware, probabilistic outputs directly aligned with operational needs, such as power ramp probabilities and uncertainty ranges. The integration of space and non-space data allows ATLAScast to outperform existing methods that rely on coarse models or single-source data, enabling more accurate, timely, and actionable decisions for energy operations.
Current Status
The focus of the project so far has been to focus on data collection for both ground truth data (installation of weather stations) and satellite data.
Dec 2025 - Project kick-off & partner engagement
Jan 2026 - Business case refinement & data acquisition
Feb 2026 - Data acquisition & AI model development
Prime Contractor(s)
