Weather Maps as Tokens: Transformers for Renewable Energy Forecasting

TL;DR

This paper introduces a transformer-based method that treats weather maps as tokens to improve renewable energy forecasting by capturing spatial and temporal weather patterns, significantly reducing forecast errors.

Contribution

It presents a novel approach that encodes weather maps as tokens for transformers, enhancing the integration of spatial and temporal weather data for energy prediction.

Findings

Reduced RMSE by about 60% for wind energy

Reduced RMSE by about 20% for solar energy

Outperforms current operational forecasts

Abstract

Accurate renewable energy forecasting is essential to reduce dependence on fossil fuels and enabling grid decarbonization. However, current approaches fail to effectively integrate the rich spatial context of weather patterns with their temporal evolution. This work introduces a novel approach that treats weather maps as tokens in transformer sequences to predict renewable energy. Hourly weather maps are encoded as spatial tokens using a lightweight convolutional neural network, and then processed by a transformer to capture temporal dynamics across a 45-hour forecast horizon. Despite disadvantages in input initialization, evaluation against ENTSO-E operational forecasts shows a reduction in RMSE of about 60% and 20% for wind and solar respectively. A live dashboard showing daily forecasts is available at: https://www.sardiniaforecast.ifabfoundation.it.