In this document
You will learn how to interpret the data acquired from the Solargis Forecast service.
Overview
Solargis Forecast delivers high-precision predictions of solar resource, meteorological parameters, and PV power output for PV plants and energy management. By combining Solargis Cloud Motion Vector (CMV) and global Numerical Weather Prediction (NWP) models, Solargis Forecast provides actionable data from minutes ahead up to 14 days into the future. Understanding how to interpret these forecast data streams is essential for grid integration, operational planning, and performance optimization.
Note: You can read more details about the Solargis Forecast modeling approach here.
Key Forecast data groups and their characteristics
Solargis Forecast data is structured into two main groups, each with a distinct modeling approach, time horizon, and update frequency.
Nowcast (Short-term forecast, up to approximately 3 hours ahead)
Definition: Nowcast data is generated by the Solargis CMV model, which analyzes recent satellite images to predict cloud movement and resulting solar irradiance. This method provides highly accurate forecasts for the next 3 hours.
Every 5, 10, or 15 minutes, depending on your geographic location (Fig.1).
Nowcast excels at capturing rapid changes in solar conditions, making it ideal for grid operations and short-term scheduling.
Its accuracy diminishes beyond the 3-hour horizon, where cloud movement becomes less predictable.
In regions where Nowcast is unavailable, forecasts are based solely on NWP models.
Access Nowcast data at the update rate for your location; more frequent requests do not yield new values.
In Solargis Forecast, although data for the whole current day is delivered in one response, only the first ~3 hours from the current moment are updated at the Nowcast frequency. Beyond this, data is from NWP models and does not update as frequently.
Tip: Use Nowcast for immediate operational decisions and short-term energy management. For periods beyond 3 hours, rely on NWP forecast data.
Figure 1: Data update frequency map for world regions.
NWP Forecast (Mid- to long-term forecast, beyond 3 hours up to 14 days ahead)
Definition: For forecasts beyond approximately 3 hours, Solargis uses the following advanced NWP models:
These models simulate atmospheric conditions globally.
Every 6 hours for most regions
Every 1 hour for the contiguous USA
NWP models provide robust, consistent forecasts for longer time horizons, though they are less effective at short-term cloud tracking.
Access and use NWP forecast data at the update frequency for your region (every 1 or 6 hours).
The forecasted meteorological parameters are post-processed by Solargis, and PV power forecasts are calculated for each time horizon.
Tip: For day-ahead and longer-term planning (DAY+1 to DAY+14), use NWP-based forecast data and align your data access with the model update schedule for your region.
Performance essentials
Model integration: Solargis Forecast combines CMV-based Nowcast and NWP models to maximize accuracy for each forecast horizon.
Update cycles:
Nowcast: Every 5, 10, or 15 minutes (location-dependent)
NWP Forecast: Every 6 hours (global), every 1 hour (contiguous USA)
Data maturity: Forecasts are updated as new satellite and meteorological inputs are received, ensuring the lowest expected deviation in predictions.
Use cases:
Nowcast for real-time grid management, energy dispatch, and short-term trading.
NWP Forecast for day-ahead scheduling, market bidding, and maintenance planning.
PV power forecast: Calculated from forecasted solar and meteorological parameters for each time horizon, supporting site-specific and portfolio-level predictions.
Data reliability: Solargis Forecast does not require real production data for calibration, relying on advanced simulation and robust quality control.