Live and Forecast Data

Live solar irradiance data refers to real-time data of solar energy received per unit area at a specific location. Solcast live data covers a time frame from 7 days ago up to the present time, and is updated every 5 minutes. This data set is also referred to as “actuals” as it represents the modeled actual weather conditions at the location.

Solcast’s live solar irradiance data is generated using a combination of satellite observations and weather models. Solcast processes these data sources using advanced algorithm models that account for atmospheric conditions, such as cloud cover and aerosol levels. Real cloud movements are tracked and forecast at a resolution of 1-2km and 5 minutes, which is then downscaled to 90-meter resolution to produce accurate, real-time solar irradiance values.

Solcast’s live solar irradiance data is updated every 5-15 minutes, depending on the frequency of satellite observations available. Temporal resolutions available are 5-60 minutes. Data for each new period is added as soon as the time period passes, and is available through the Solcast API. This allows for real-time data access and timely information for accurate, immediate decision-making.

Solcast’s live irradiance data primarily relies on real-time satellite observations to provide real-time actuals up to the present moment. While forecast irradiance data combines satellite imagery and weather models to predict irradiance from now up to 14 days ahead. Live data is crucial for real-time grid management, operational adjustments, and performance monitoring, whereas forecast data is used for planning, energy production estimates, and scheduling maintenance activities in both utility-scale solar farms and distributed solar installations.

Solcast data has low bias of +0.05% with a standard deviation of ±2.05%, validated against high-quality site measurements. This means that estimates of total accumulated energy production are highly accurate.

Solcast has best-in-class accuracy on a wide variety of metrics including bias, RMSE (root-mean squared error), MAE (mean absolute error), MAPE (mean absolute percentage error), Pearson’s R value, and others. This accuracy has been independently validated against high-quality irradiance site measurements. Solcast achieves this accuracy with a combination of quality input data, processes using advanced algorithm models, calibrated data built for solar that outperforms more generic datasets.

‍

Solcast’s solar forecast irradiance data predicts future solar irradiance at a specific location, from +5 minutes to 14 days ahead. The data is generated using recent satellite imagery to predict the movement of clouds to produce a “nowcast” and global numerical weather prediction models. Check our irradiance methodology.

A solar nowcast is a short-term solar irradiance forecast, covering the next several hours, using real-time data. Solcast uses satellite cloud tracking models, using near real-time imagery from geostationary satellites, to track cloud movement and deliver high-resolution solar data every 5 to 15 minutes. Accurate solar nowcasts are essential for quick operational adjustments to optimize energy production and respond to changing weather conditions.

Day-ahead forecasting predicts solar irradiance for the upcoming day, usually 0 to 24 hours in advance. Solcast provides probabilistic forecasting, including P10, P50 and P90, for day-ahead forecasts. These help gauge uncertainties and provide a range of possible outcomes for better decision-making, preventing inefficient power management, regulatory violations, unnecessary storage system cycles and more. 

‍

Several factors affect the accuracy of forecasts, such as cloud impacts, aerosols, snow soiling, terrain shading and more. Unlike generic weather models, the Solcast API takes these factors into account. Another factor that affects accuracy is the quality of input data. Solcast takes advantage of the rise of meso-scale weather models and integrates them into the model, which has helped enhance forecast accuracy in Europe and North America.

Forecast accuracy is evaluated using metrics that compare predicted values with actual outcomes, typically including bias and RMSE. When looking for forecast data that meets your needs, an evaluation requires extensive data analysis and review. Check the assessment guide compiled from Solcast’s numerous commercial data evaluations, or assess the accuracy of Solcast data in your region using the interactive Live and Forecast accuracy map.