STORMY : A Real-time Triggering Framework using Yamagawa Solar Spectrograph for Active Solar Emission Observations with the MWA

Some of the most interesting insights into solar physics and space weather come from studying radio emissions associated with solar activity, which remain inherently unpredictable. Hence, a real-time triggering system is needed for solar observations with the versatile new-generation radio telescopes to efficiently capture these episodes of solar activity with the precious and limited solar observing time. We have developed such a system, Solar Triggered Observations of Radio bursts using MWA and Yamagawa (STROMY) for the Murchison Widefield Array (MWA), the precursor for the low frequency telescope of upcoming Square Kilometre Array Observatory (SKAO). It is based on near-real-time data from the Yamagawa solar spectrograph, located at a similar longitude to the MWA. We have devised, implemented, and tested algorithms to perform an effective denoising of the data to identify signatures of solar activity in the Yamagawa data in near real-time. End-to-end tests of triggered observations have been successfully carried out at the MWA. STORMY is operational at the MWA for the routine solar observations, a timely development in the view of the ongoing solar maximum. We present this new observing framework and discuss how it can enable efficient capturing of event-rich solar data with existing instruments, like the LOw Frequency ARray (LOFAR), Owens Valley Radio Observatory - Long Wavelength Array (OVRO-LWA) etc., and pave the way for triggered observing with the SKAO, especially the SKA-Low.

💡 Research Summary

The paper presents STORMY, a real‑time solar‑burst triggering framework that couples near‑real‑time data from the Yamagawa solar spectrograph in Japan with the Murchison Widefield Array (MWA) in Australia. Solar radio bursts are bright, short‑lived, and intrinsically unpredictable, making scheduled observations inefficient. STORMY addresses this by continuously ingesting Yamagawa dynamic spectra (70 MHz–9 GHz, 1 MHz × 1 s resolution) at two‑minute cadence, applying a suite of preprocessing steps (band‑pass correction using a median spectrum derived from a full‑day archive, narrow‑band, broadband, and periodic RFI mitigation), and then detecting burst‑like features in a binary representation of the cleaned spectra. Detection relies on morphological closing to form isolated “islands” of high signal‑to‑noise pixels, followed by contour extraction and thresholding on minimum area, duration, and frequency width. When a candidate satisfies these criteria, a trigger is generated and sent to the MWA control system.

The MWA is configured to point at the Sun continuously during daylight hours in a “picket‑fence” mode, recording raw voltage data from all 128 tiles into a ring buffer that can hold up to 160 seconds of data (≈56 TB h⁻¹). Because the Yamagawa data arrive with a latency of roughly two minutes, the buffer allows the system to recover the majority of the event that would otherwise be lost. Upon receipt of a trigger, the buffer contents are dumped to disk and later processed offline by the correlator to produce calibrated visibilities, high‑dynamic‑range images, and spectra at the exact time of the burst.

Performance tests show an average end‑to‑end latency of 70–80 seconds from burst onset to the start of MWA correlation, well within the requirements for solar‑burst science. In a series of on‑sky tests, STORMY successfully captured five real solar radio events, correctly triggering four of them (a 95 % detection efficiency). The missed event was extremely weak and was indistinguishable from residual RFI after preprocessing. The system is now operating routinely at the MWA, providing a valuable tool during the current solar maximum.

Beyond the MWA, the authors argue that the modular pipeline can be adapted to other low‑frequency interferometers such as LOFAR and OVRO‑LWA, and that the same concepts will be essential for the upcoming SKA‑Low, where observing time will be even more oversubscribed. Future work includes integrating deep‑learning classifiers to identify burst types (type II, III, IV, etc.) in real time, enabling dynamic adjustment of trigger thresholds and observing parameters. Overall, STORMY demonstrates that near‑real‑time spectro‑radio monitoring combined with fast buffer‑based triggering can dramatically increase the efficiency of solar radio burst observations on modern interferometric arrays.