GenONet: A Generative operator Network for High-Resolution Precipitation Nowcasting

arXiv cs.LG Papers
precipitation-nowcasting deep-learning generative-adversarial-network operator-network physics-informed weather-prediction

Summary

GenONet is a novel deep learning architecture that uses Deep Operator Networks within a Generative Adversarial Network framework for high-resolution precipitation nowcasting, producing sharp and physically consistent forecasts.

arXiv:2609.00544v1 Announce Type: new Abstract: High-resolution precipitation nowcasting is critical for reducing the impacts of severe weather but remains difficult because of rapid storm evolution. Deep learning models have shown great promise for this task, but their predictive skill often deteriorates over longer forecast horizons. This leads to increasingly blurry forecasts that fail to capture the complex, non-linear evolution of storm systems. In order to address these limitations, we introduce Spatio-Temporal U-DeepONet (GenONet), a novel architecture for long-range precipitation forecasting up to 3 hours, specifically designed to produce sharp and physically consistent results. GenONet's architecture pioneers the use of a Deep Operator Network (DeepONet) as a generator within a Generative Adversarial Network (GAN) framework for this task. The DeepONet learns the continuous-time dynamics of precipitation, ensuring stability over long forecast horizons. Adversial training against a spatio-temporal discriminator compels the model to produce sharp, coherent forecasts, while a physics-informed loss regularizer, derived from the Moisture Conservation Equation, improves physical plausibility in our ablation setting. Quantitative evaluations show that our model achieves consistently higher scores on most of the metrics, especially for highintensity events and at longer lead times. Qualitatively, GenONet produces structurally coherent forecasts that maintain their integrity, whereas baseline models degrade into indistinct patterns. Finally, an ablation study confirms the benefit of this physics-informed loss, highlighting the strength of combining operator learning with adversarial training.
Original Article
View Cached Full Text

Cached at: 09/02/26, 06:16 AM 

# GenONet: A Generative operator Network for High-Resolution Precipitation Nowcasting
Source: [https://arxiv.org/abs/2609.00544](https://arxiv.org/abs/2609.00544)
[View PDF](https://arxiv.org/pdf/2609.00544)

> Abstract:High\-resolution precipitation nowcasting is critical for reducing the impacts of severe weather but remains difficult because of rapid storm evolution\. Deep learning models have shown great promise for this task, but their predictive skill often deteriorates over longer forecast horizons\. This leads to increasingly blurry forecasts that fail to capture the complex, non\-linear evolution of storm systems\. In order to address these limitations, we introduce Spatio\-Temporal U\-DeepONet \(GenONet\), a novel architecture for long\-range precipitation forecasting up to 3 hours, specifically designed to produce sharp and physically consistent results\. GenONet's architecture pioneers the use of a Deep Operator Network \(DeepONet\) as a generator within a Generative Adversarial Network \(GAN\) framework for this task\. The DeepONet learns the continuous\-time dynamics of precipitation, ensuring stability over long forecast horizons\. Adversial training against a spatio\-temporal discriminator compels the model to produce sharp, coherent forecasts, while a physics\-informed loss regularizer, derived from the Moisture Conservation Equation, improves physical plausibility in our ablation setting\. Quantitative evaluations show that our model achieves consistently higher scores on most of the metrics, especially for highintensity events and at longer lead times\. Qualitatively, GenONet produces structurally coherent forecasts that maintain their integrity, whereas baseline models degrade into indistinct patterns\. Finally, an ablation study confirms the benefit of this physics\-informed loss, highlighting the strength of combining operator learning with adversarial training\.

## Submission history

From: Mohammad Kian Golkar \[[view email](https://arxiv.org/show-email/2ef61f46/2609.00544)\] **\[v1\]**Tue, 1 Sep 2026 01:30:33 UTC \(2,838 KB\)

Similar Articles

KernelOnet: An Interpretable Neural Operator Based on Kernel Functions

arXiv cs.LG

KernelOnet proposes an interpretable neural operator framework that explicitly incorporates kernel functions in place of trunk networks, offering data-driven, physics-informed, and hybrid kernel strategies for PDE solving with higher accuracy and fewer parameters than DeepONet.

google-deepmind/weathernext

GitHub Trending (daily)

Google DeepMind has open-sourced WeatherNext 2, its most advanced global weather and cyclone forecasting model, along with prior models GraphCast and GenCast, providing code and pretrained weights.