SolarSeer: Ultrafast and accurate 24-hour solar irradiance forecasts outperforming numerical weather prediction across the USA

• URL: http://arxiv.org/abs/2508.03590v2
• Date: Tue, 02 Sep 2025 10:49:25 GMT
• Title: SolarSeer: Ultrafast and accurate 24-hour solar irradiance forecasts outperforming numerical weather prediction across the USA
• Authors: Mingliang Bai, Zuliang Fang, Shengyu Tao, Siqi Xiang, Jiang Bian, Yanfei Xiang, Pengcheng Zhao, Weixin Jin, Jonathan A. Weyn, Haiyu Dong, Bin Zhang, Hongyu Sun, Kit Thambiratnam, Qi Zhang, Hongbin Sun, Xuan Zhang, Qiuwei Wu,
• Abstract summary: SolarSeer is an end-to-end large artificial intelligence (AI) model for solar irradiance forecasting across the Contiguous United States (CONUS)<n>It is designed to directly map the historical satellite observations to future forecasts, eliminating the computational overhead of data assimilation and PDEs solving.<n>It significantly reduces the root mean squared error of solar irradiance forecasting by 27.28% in reanalysis data and 15.35% across 1,800 stations.
• Score: 18.52761376569946
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Accurate 24-hour solar irradiance forecasting is essential for the safe and economic operation of solar photovoltaic systems. Traditional numerical weather prediction (NWP) models represent the state-of-the-art in forecasting performance but rely on computationally costly data assimilation and solving complicated partial differential equations (PDEs) that simulate atmospheric physics. Here, we introduce SolarSeer, an end-to-end large artificial intelligence (AI) model for solar irradiance forecasting across the Contiguous United States (CONUS). SolarSeer is designed to directly map the historical satellite observations to future forecasts, eliminating the computational overhead of data assimilation and PDEs solving. This efficiency allows SolarSeer to operate over 1,500 times faster than traditional NWP, generating 24-hour cloud cover and solar irradiance forecasts for the CONUS at 5-kilometer resolution in under 3 seconds. Compared with the state-of-the-art NWP in the CONUS, i.e., High-Resolution Rapid Refresh (HRRR), SolarSeer significantly reduces the root mean squared error of solar irradiance forecasting by 27.28% in reanalysis data and 15.35% across 1,800 stations. SolarSeer also effectively captures solar irradiance fluctuations and significantly enhances the first-order irradiance difference forecasting accuracy. SolarSeer's ultrafast, accurate 24-hour solar irradiance forecasts provide strong support for the transition to sustainable, net-zero energy systems.

Related papers

• Ultra-short-term solar power forecasting by deep learning and data reconstruction [60.200987006598524]
  We propose a deep-learning based ultra-short-term solar power prediction with data reconstruction.<n>We employ deep-learning models to capture long- and short-term dependencies towards the target prediction period.
  arXiv  Detail & Related papers  (2025-09-21T14:22:35Z)
• SolarCrossFormer: Improving day-ahead Solar Irradiance Forecasting by Integrating Satellite Imagery and Ground Sensors [0.8808021343665318]
  SolarCrossFormer is a novel deep learning model for day-ahead irradiance forecasting.<n>It combines satellite images and time series from a ground-based network of meteorological stations.<n>It generates probabilistic forecasts for any location in Switzerland with a 15-minute resolution for horizons up to 24 hours ahead.
  arXiv  Detail & Related papers  (2025-09-19T09:57:40Z)
• Data-driven solar forecasting enables near-optimal economic decisions [38.70705528456187]
  We present SunCastNet, a lightweight data-driven forecasting system.<n>It provides 0.05$circ$, 10-minute resolution predictions of surface solar radiation downwards.<n>It reduces operational regret by 76--93% compared to robust decision making.<n>It enables up to five of ten high-emitting industrial sectors per region to cross the commercial viability threshold of 12% Internal Rate of Return.
  arXiv  Detail & Related papers  (2025-09-08T17:38:05Z)
• Intra-day Solar and Power Forecast for Optimization of Intraday Market Participation [31.426172725953638]
  The prediction of solar irradiance enhances reliability in photovoltaic (PV) solar plant generation and grid integration.<n>This research employs Long Short-Term Memory (LSTM) and Bi-LSTM models to predict solar irradiance with a 6-hour horizon and 10-minute resolution.<n>The LSTM predictions were averaged to create an hourly resolution model, evaluated using Mean Absolute Error, Root-Mean-Square Error, Normalized Root-Mean-Square Error, and Mean Absolute Percentage Error metrics.
  arXiv  Detail & Related papers  (2025-01-16T14:12:03Z)
• Forecasting SEP Events During Solar Cycles 23 and 24 Using Interpretable Machine Learning [38.321248253111776]
  We employ a suite of machine learning strategies to evaluate the predictive potential of a new data product for a forecast of post-solar flare SEP events. Despite the augmented volume of data, the prediction accuracy reaches 0.7 +- 0.1, which aligns with but does not exceed these published benchmarks.
  arXiv  Detail & Related papers  (2024-03-04T23:12:17Z)
• Transformer-based nowcasting of radar composites from satellite images for severe weather [45.0983299269404]
  We present a Transformer-based model for nowcasting ground-based radar image sequences using satellite data up to two hours lead time. Trained on a dataset reflecting severe weather conditions, the model predicts radar fields occurring under different weather phenomena. The model can support precipitation nowcasting across large domains without an explicit need for radar towers.
  arXiv  Detail & Related papers  (2023-10-30T13:17:38Z)
• Improving day-ahead Solar Irradiance Time Series Forecasting by Leveraging Spatio-Temporal Context [46.72071291175356]
  Solar power harbors immense potential in mitigating climate change by substantially reducing CO$_2$ emissions. However, the inherent variability of solar irradiance poses a significant challenge for seamlessly integrating solar power into the electrical grid. In this paper, we put forth a deep learning architecture designed to harnesstemporal context using satellite data.
  arXiv  Detail & Related papers  (2023-06-01T19:54:39Z)
• Data-driven soiling detection in PV modules [58.6906336996604]
  We study the problem of estimating the soiling ratio in photo-voltaic (PV) modules. A key advantage of our algorithms is that they estimate soiling, without needing to train on labelled data. Our experimental evaluation shows that we significantly outperform current state-of-the-art methods for estimating soiling ratio.
  arXiv  Detail & Related papers  (2023-01-30T14:35:47Z)
• GraphCast: Learning skillful medium-range global weather forecasting [107.40054095223779]
  We introduce a machine learning-based method called "GraphCast", which can be trained directly from reanalysis data. It predicts hundreds of weather variables, over 10 days at 0.25 degree resolution globally, in under one minute. We show that GraphCast significantly outperforms the most accurate operational deterministic systems on 90% of 1380 verification targets.
  arXiv  Detail & Related papers  (2022-12-24T18:15:39Z)
• FourCastNet: Accelerating Global High-Resolution Weather Forecasting using Adaptive Fourier Neural Operators [44.20526496271154]
  Current physics-based numerical weather prediction (NWP) limits accuracy due to high computational cost and strict time-to-solution limits. We report that a data-driven deep learning Earth system emulator, FourCastNet, can predict global weather and generate medium-range forecasts five orders-of-magnitude faster than NWP.
  arXiv  Detail & Related papers  (2022-08-08T06:06:31Z)
• A Moment in the Sun: Solar Nowcasting from Multispectral Satellite Data using Self-Supervised Learning [4.844946519309793]
  We develop a general model for solar nowcasting from abundant and readily available multispectral satellite data using self-supervised learning. Our model estimates a location's future solar irradiance based on satellite observations. We evaluate our approach for different coverage areas and forecast horizons across 25 solar sites.
  arXiv  Detail & Related papers  (2021-12-28T03:13:44Z)
• Prediction of Solar Radiation Using Artificial Neural Network [0.0]
  This paper presents an algorithm that can be used to predict an hourly activity of solar radiation. The dataset consists of temperature of air, time, humidity, wind speed, atmospheric pressure, direction of wind and solar radiation data. Two models are created to efficiently create a system capable of interpreting patterns through supervised learning data.
  arXiv  Detail & Related papers  (2021-04-01T20:41:27Z)
• Improving solar wind forecasting using Data Assimilation [0.0]
  We use a variational DA scheme with a computationally efficient solar wind model and in situ observations from STEREO-A, STEREO-B and ACE. This scheme enables solar-wind observations far from the Sun to update and improve the inner boundary conditions of the solar wind model. We find that 27-day root mean-square error (RMSE) for STEREO-B corotation and DA forecasts are comparable and both are significantly lower than non-DA forecasts.
  arXiv  Detail & Related papers  (2020-12-11T14:05:12Z)
• Short term solar energy prediction by machine learning algorithms [0.47791962198275073]
  We report daily prediction of solar energy by exploiting the strength of machine learning techniques. Forecast models of base line regressors including linear, ridge, lasso, decision tree, random forest and artificial neural networks have been implemented. It has been observed that improved accuracy is achieved through random forest and ridge regressor for both grid sizes.
  arXiv  Detail & Related papers  (2020-10-25T17:56:03Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.