Implementation of a Near-real-time Recording and Reporting System of Solar Radio Bursts

Strong solar activities are often accompanied by a variety of radio bursts. These radio bursts not only serve as valuable diagnostics of coronal and heliospheric processes but also as potential tools in space weather monitoring and forecasting. However, space weather applications call for the capability for low-latency and high-sensitivity radio burst recording and reporting, which has remained lacking. In this work, we present the development of a near-real-time radio burst recording and reporting system with the Owens Valley Radio Observatory’s Long Wavelength Array. The system directly clips data from the real-time buffer and streams it as a live real-time radio dynamic spectrogram. The spectrograms are then fed to a deep learning–based burst identification module for type III radio bursts. The identifier is built on a You Only Look Once architecture, trained by synthetic type III radio bursts generated by using a physics-based model to achieve accurate and robust detection. This system enables continuous real-time radio spectrum streaming and the automatic reporting of type III radio bursts within ∼10 s of their occurrence.