DayCast is Reuniwatt’s product to obtain day-ahead solar forecasts. Our algorithms are based on meteorological models optimised for solar that are combined with artificial intelligence and statistical models. They forecast solar irradiance and PV production up to 10 days in advance.

reuniwatt numerical weather prediction solar irradiance

Global coverage and local accuracy

Global coverage and local accuracy

Our algorithms of day-ahead solar forecasts are available worldwide. Moreover, Reuniwatt uses regional meteorological models in order to ensure a high temporal and spatial resolution on the chosen site.

Diversification of supply sources

Diversification of supply sources

Reuniwatt’s day-ahead solar forecasts integrate cloudiness parameters coming from several sources, thus enabling to improve their reliability.

Improved cloud representation in weather models

Improved cloud representation in weather models

The classical weather forecasts do not prioritize the thorough investigation of cloud cover evolution. Reuniwatt conducts research towards a dedicated modeling of this major source of variation of the photovoltaic power generation.

Ask for more info



DayCast is a weather forecasting technique developed by Reuniwatt, and specifically optimised for the day-ahead and up to a whole week ahead cloudiness and solar irradiance.

DayCast combines several numerical meteorological models, permitting to improve the forecasts’ robustness. Artificial intelligence techniques train our models to better take local phenomena into account.

DayCast offers a global coverage and offers many parameters useful for the next day’s forecasts:

  • Simulation thanks to a weather model integrating regional specificities
  • External weather data integration so as to improve the model’s pertinence
  • Frequent updates of the simulations
  • Initialisation of high-resolution cloud resolving regional models using innovative data assimilation methods
reuniwatt prévisions solaires au lendemain centrale


Single-site, portfolio or territories forecasts

Self-learning module for a continuous performance improvement

Forecasting horizon up to 10 days ahead

Global irradiance and photovoltaic power generation forecasts

Flexible sending (FTP, email, API, …) and format (xml, json, csv…)

More info