Solar power forecasting software for asset operators

Site-level solar generation forecasts with sub-hourly precision — trained on your asset's specific configuration, irradiance history, and local weather patterns.

Book a demo

What is solar power forecasting?

Solar power forecasting predicts how much electricity a specific solar asset will generate — from day-ahead bidding to real-time dispatch and curtailment management. Unlike grid-level estimates, asset-level forecasting models your exact site conditions, including panel configuration, shading, and local weather.

Amperon’s sub-hourly ML models combine multiple weather sources, satellite irradiance data, and historical generation patterns to continuously adapt to rapid output swings and changing weather conditions.

Amperon's solar forecasting products

Asset solar short term forecast

Horizon:

Short-term (day-ahead+)

Update frequency:

xxx

Delivery:

API, snowflake, file

Use cases:

Day-ahead bidding, generation scheduling, portfolio reporting

Short-term solar generation forecasts for individual assets across day-ahead and multi-day horizons. Models are trained on each site’s historical generation, physical configuration, and local weather patterns to support bidding, scheduling, and portfolio reporting.

X.X%

MAPE for DA generation forecasts

Asset solar sub-hourly

Horizon:

Intra-day

Resolution:

Sub-hourly

Updates:

Continuous

Delivery:

API, snowflake, file

Use cases:

Curtailment management, intra-day market participation, real-time dispatch

Intra-day solar generation forecasts with sub-hourly resolution for real-time dispatch, curtailment, and market participation. Continuously updated using real-time satellite imagery and irradiance data to capture rapid output swings caused by changing cloud conditions.

X

Minute resolution with Y-minute update cycles

Reduce solar curtailment with better forecasts

Every curtailed MWh is lost revenue. Sub-hourly forecasting lets you see it coming and act before it hits.

See curtailment windows before they hit

Sub-hourly precision gives you hours of lead time to adjust bids and ramp schedules.

Capture revenue others leave on the table

X% fewer unexpected curtailment events.

Connect to the full market picture

Asset forecasts feed into Amperon's net demand signal for complete grid-level context.

X.X%

DA generation forecast accuracy (MAPE)

X+

Solar assets currentlyforecasted

X wks

Average time from contract to first forecast delivery

How it works

View integration documentation

Data ingestion

Public ISO demand data combined with 25+ years of historical weather data and real-time weather feeds. Multiple weather model sources blended for optimal accuracy.

ML Modeling

Ensemble machine learning models retrained and updated hourly as new demand and weather data arrives. Each ISO has dedicated model architectures tuned to that market's characteristics.

Output delivery

Short-term and long-term portfolio forecasts delivered via API or Snowflake. Automated day-ahead scheduling pushed directly to the ISO. Threshold alerts flag data anomalies before they impact bids.

Continuous learning loop

Models self-improve with every new data point. Accuracy compounds over time as the training dataset grows and the model adapts to shifting demand patterns, EV adoption, and electrification trends.