
Solar irradiance data
Accurate, bankable solar insights from archive to real-time monitoring
- 20+ years of history
- 50% uncertainty reduction
- Any global location
Bias-corrected irradiance records spanning over 20 years, relied on by project developers, operators, and financiers worldwide.
:focal(1886x2340:1887x2341))
Finance-grade solar resource data from satellite observations
The Vaisala Xweather Solar Model 3 delivers industry-leading accuracy for historical evaluation and real-time site performance monitoring. The first truly global, finance-grade solar irradiation dataset powered entirely by satellite intelligence and refined through the proven Heliosat-V methodology.
years of historical irradiance data
greater GHI accuracy vs. ERA5
native satellite resolution
mean bias error
Solar Model 3: A new standard for satellite-based GHI
Solar Model 3 is Xweather’s proprietary model that transforms satellite imagery into ultra-accurate Global Horizontal Irradiance (GHI) data. Based on more than two decades of refinement, it sets a new benchmark for solar resource accuracy, 38% greater compared to ERA5.
Our cloud-opacity modeling delivers exceptionally low bias and minimal deviation, making it suited for developers, operators, and financiers who demand reliability at every stage of the solar lifecycle.
- Bias-corrected records
- Finance-grade quality
- API delivery
)
)
Bankable resource assessments
Data quality meets the requirements of project finance lenders and investment committees.
Reduced resource uncertainty
Bias-correction reduces estimation error by more than 50% compared to uncorrected reanalysis sources.
Complete irradiance picture
GHI, DNI, DHI, and albedo delivered from a single source, eliminating multi-vendor data reconciliation.
Operational performance monitoring
Ongoing real-time data feeds support operational performance tracking against expected generation profiles.
Solar Model 3 processes satellite imagery through a multi-stage pipeline to produce finance-grade irradiance data.
15-minute satellite imagery is processed to extract cloud opacity across the target region. Coverage spans multiple geostationary satellite footprints for near-global reach.
The McClear clear-sky model establishes baseline irradiance conditions for each location, accounting for atmospheric composition, altitude, and solar geometry.
The Heliosat-V cloud-corrected model applies atmospheric and surface corrections to convert cloud opacity into irradiance estimates at 3 km resolution.
Adaptive calibration refines output for each satellite sensor's spectral profile. Data is delivered via synchronous API in JSON, GeoJSON, CSV, and TSV formats.
Solar resource data across the project lifecycle
From initial site selection through operational monitoring, solar irradiance data supports every stage of a solar project's development, financing, and performance management.
Site assessment
The results from Solar Model 3 and ground-based measurements show a mean bias error (MBE) of 0.86%, a bias standard deviation of 2.72%, an hourly mean absolute error (MAE) of 9.84% and an hourly root mean square error (RMSE) of 15.06%.
A 1 % reduction in GHI uncertainty can increase allowable debt sizing by 0.5–1 %, representing $500 K–$1 M in additional financeable capital on a $100 M project.
)
Operational performance monitoring
Ongoing irradiance data feeds enable operators to continuously compare actual generation with expected performance.
When output deviates from modeled expectations, operators can distinguish between weather-driven shortfalls and equipment or degradation issues.
Real-time and historical data together support warranty claims, investor reporting, and operational optimization throughout the asset's life.
)
Solar Model 3 capabilities
Solar irradiance archive data
Renewable energy solar irradiance archive data delivers over 20 years of high-quality solar irradiance time-series data, providing unparalleled insight for long-term planning and performance analysis. Validated against more than 250 sites worldwide and produced at hourly resolution, our archive delivers 38% greater accuracy compared to ERA5, setting a new standard for reliable historical irradiance data.
Solar Model 3 methodology and validation
This validation paper documents the scientific methodology behind Solar Model 3 and evaluates its performance against high-quality ground-based measurements. Results confirm consistently low bias and strong agreement with observed Global Horizontal Irradiance (GHI) across regions, climates, and satellite domains.
)
Bias-correction for historical solar datasets
With just 3 months of on-site observation data, we can reduce the uncertainty in the historical satellite data by 25%. With 4 months of data, we reduce the uncertainty by 42%. Accuracy improves with more data, but we find the best return on investment comes from 12 months of observations, which will reduce your uncertainty by 57%.
Satellite data provides a readily available historical archive of solar resource data for most locations. Single-source satellite data is consistent and reliable, but it comes with ±5% uncertainty on average.
On-site observations from high-quality pyranometers reveal the ground-truth conditions. However, a limited sample of short-term observations might not reflect typical long-term conditions.
Xweather combines on-site observations, long-term satellite data, and NWP weather data in a multi-linear statistical correction called Model Output Statistics (MOS) to produce a site-specific, bias-corrected dataset.
)
Irradiance components and meteorological variables
Solar Model 3 delivers a complete set of irradiance components alongside supporting meteorological variables. All parameters are available as time series from 2004 to present.
Irradiance
Global horizontal irradiance (GHI), direct normal irradiance (DNI), diffuse horizontal irradiance (DHI), and horizontal albedo. The core parameters for resource assessment, yield modeling, and performance monitoring.
Clear-sky reference
McClear clear-sky irradiance provides the theoretical maximum for each location. Comparing measured or modeled values against clear-sky enables cloud impact quantification and anomaly detection.
Meteorological variables
Supporting variables include air temperature, wind speed, and humidity. These complement irradiance data for energy yield modeling, thermal performance analysis, and site characterization.
Related resources


Request a sample dataset
See Solar Model 3 data for your locations. Our team can walk through the methodology, coverage, and integration options.
- Trusted by the global solar finance community
- Available for any location on Earth
- API and online tool access
Frequently asked questions
Solar Model 3 provides global horizontal irradiance (GHI), direct normal irradiance (DNI), diffuse horizontal irradiance (DHI), and albedo. Supporting meteorological variables include temperature, wind speed, and humidity. Data is available as 15-minute, hourly, daily, and monthly time series, with a historical archive from 2004 to present.