Creating an Open Source Global Solar Forecast and Dashboard
Event Recap: FOSDEM 2026
TL;DR
At FOSDEM 2026, Alex Udaltsova and Sol Cotton of Open Climate Fix presented an open source global solar forecast and dashboard. The session explained how Open Climate Fix approaches renewable energy forecasting, introduced Open Quartz as a lightweight open source solar forecasting library, and demonstrated a global dashboard for exploring 48-hour solar generation forecasts by country.
Why Solar Forecasting Matters
Cotton opened by explaining that Open Climate Fix is a nonprofit open source startup focused on renewable energy forecasting, including solar, wind, and cloud forecasting.
He described solar forecasting as one way to help grid operators better understand expected solar generation. According to Cotton, uncertainty in solar generation can make grid balancing more difficult, while better forecasts can help operators plan around variable solar output.
PVNet and the UK Grid
Cotton described PVNet, a solar forecasting model developed by Open Climate Fix and provided to the UK National Grid.
He said the forecast is used in the control room to help operators understand expected solar generation in the country. He also noted that it is difficult to measure the exact impact because many decisions are involved in grid operations.
Making Forecasting More Open and Accessible
Cotton explained that Open Climate Fix wanted to make solar forecasting available to more people and easier for others to contribute to.
While PVNet is a larger service, the team developed Open Quartz as a lighter-weight open source alternative. Udaltsova described Open Quartz as a smaller machine learning model wrapped in a Python library that runs on open weather forecasts and can be run locally.
From Open Quartz to a Global Forecast
Udaltsova explained that Open Climate Fix began with an open UK service, but users started applying Open Quartz in other countries.
The global forecast builds on that work. It uses country center points and capacity information to produce 48-hour solar power forecasts by country.
Demonstrating the Dashboard
Udaltsova demonstrated a dashboard built on top of the global forecast.
The dashboard shows forecast solar generation for the world, country-level forecasts, capacity information, and utilization normalized by each country’s capacity. Udaltsova also noted that open source contributors helped add country capacity data and improve the map visualization.
Current Limitations
Udaltsova emphasized that the project is still a work in progress.
Open Quartz was trained on UK solar generation data, so it may not perform as well in other geographies. She noted that the team has benchmarking for the UK, but not yet for other regions.
How Contributors Can Help
Udaltsova invited contributors to share generation data, help benchmark forecasts, train or swap in different models, test the tool, and suggest missing features.
The speakers also discussed cloud forecasting, which uses satellite imagery to help understand where clouds are and where they may move in the future, since cloud cover affects solar generation.
AI Disclosure
This post used artificial intelligence tools for research, structural assistance, or grammatical refinement. The final content was reviewed, edited, and validated by human contributors to LF Energy to ensure accuracy and alignment with our community standards. We remain committed to transparency in the use of generative technologies within the open source ecosystem.
