DNV SolarFarmer
Bankable 3D-aware PV yield simulations, from automated layout to sub-hourly energy outputs, all in one streamlined tool.
Overview
Solar PV Power Forecast Models
Accurate and flexible PV power models, for the realities of operational solar assets
.avif)
_Frame.png)
Complete, end-to-end operational PV forecasting, for any asset
Solcast offers three solar energy forecasting models - Rooftop PV, Advanced PV and Premium PV - that model PV power for assets of any scale or complexity. With a single API call you can model historical, live and forecast PV power, either in your own notebooks, inside energy platforms, or in your own energy software tool.
Choose the PV forecast power model that you want to work on top of the irradiance and weather data you get through the Solcast API. The Rooftop PV forecast power model is included with Solcast Live and Forecast data.
Premium PV Power Forecast Model
A site-trained model, built and run for you by DNV
Trained on your plant's own generation history, so the model learns how your plant behaves. DNV prepares the data, configures the training and reviews the result before it goes live. Probabilistic scenarios (P90–P10, P75–P25)* let you size each bid to the real range.
*Max plan.
Advanced PV Power Forecast Model
A physics model you configure and tune yourself
Physics-based, built on pvlib with Solcast extensions. Up to 20 parameters let you tell the model how your plant behaves and match it to your own expected-power calculations. SDK tools support self-service tuning.
Rooftop PV Power Forecast Model
Predict PV power at one site or millions, with or without specs and measurement data
A simple, proven 6-parameter empirical model, which can operate on as little as the location and the installed capacity. Specify tilt, azimuth and a loss factor when available, to refine the accuracy. Use our SDK to train the model on your own measurement data. In commercial use at millions of solar assets globally.
Bring investment-grade solar energy forecasting & data into your workflow
Leave your details and one of our team will be in touch.
Oops! Something went wrong while submitting the form. Please refresh the page and try again.
Explore our Markets & Portfolios Energy Data Models for grid-level generation, portfolio-scale probabilistic forecasts and load insights
Explore other DNV Green Data products
.svg)
Frequently asked questions about PV forecasts
What outputs do the models provide?
All four models return continuous, gap-free time-series in kW or MW: estimated actuals (-7 days to “now”) and forecasts (“now” to +14 days) at 5-, 15- or 60-minute resolution.
How do I decide which model to use?
Choose by data availability and accuracy target: Advanced PV for high-spec utility sites, Rooftop PV for low-spec sub-MW fleets, Premium PV Power Model (PVML) when you have SCADA and need ultra-low error.
What geographic coverage is included?
Global. The models leverage Solcast’s 1–2 km satellite data and NWP grids, so you can request forecasts for any coordinate worldwide without local sensors.
Can the models handle snow, soiling or curtailment?
Yes. Rooftop PV Power and Advanced PV Power include dynamic snow-cover and soiling factors and Premium PV Power ingests SCADA availability/curtailment tags to refine training.
How do I access the data?
Via RESTful JSON or CSV API endpoints. You can also use the Solcast Toolkit for quick testing, bulk downloads and site management.
Are academic or research licences available?
Yes—discounted licences are offered to accredited researchers and universities.
What support do Solcast and DNV provide?
Solcast offers onboarding guides, sample code, email support and an optional Accuracy-Assist tuning service. DNV provides detailed documentation, training webinars and next-business-day email support..
