Predictable Makes Profitable

Accurate wind and solar power forecasts for power generating, site planning, and trading

5%

Expected prediciton accuracy improvement by at least 5%

20%

20+% accuracy improvement achievable in certain areas

per farm

Automatic customization for “one farm, one model”

Landscape Matters

Local High-Resolution Digital Twin of Earth Surface

Landscape plays important role when it comes to modeling micro-climate. TerraQuanta has developed 0.5-meter-level digital twin for earth surface, making it possible to build better power forecasting models for sites in hilly and canyon areas.

Decode The Wind

Wake Effect Analysis

Wind farms' power outputs may be significantly influenced by turbulence or windspeed change caused by the turbines. For an individual turbine, the power loss caused by wake effect can be as much as 100%. TerraQuanta's forecasting model uses graph neural network(GCN) to calculate the impact of wake effect and minimize its influence on forecasting results.

Scale It Down

Combining Global And Local Models

Weather condition varies from farm to farm, that's why TerraQuanta treats power plants differently. Our AI system delivers forecasting results 4 times a day, on top of which farm-scale models are customized and optimized.

Defrost The Problem

Ice/Snow Coverage Monitoring for Solar Power Plants

Ice/snow coverage is a common issue for solar plants in high altitudes or latitudes, causing power loss of up to 100%. TerraQuanta uses satellite remote sensing to monitor ice and snow, and the monitoring results are analyzed by the forecasting model to further improve accuracy.

Go Flexible

Forecasting For The Next 4 Hours To Up To 15 Days

TerraQuanta's power forecasting service is capable of delivering results of different spatial scales and predict lengths, at various frequency. The forecasting pattern is customizable from 1min to 1hr interval.

Take A Look At our AI Model for Global Weather Forecasting!

Site-level power forecasting models heavily rely on the fundamental weather data provided by Simba, TerraQuanta's globe-scale AI model for meteorological forecasting.