Reconstructing Synthetic SDO/AIA 193 A EUV Images from He I 10830 A Observations with Diffusion Model Translator

Researchers developed a diffusion model-based framework called CH-aware DMT that reconstructs synthetic SDO/AIA 193 Å EUV solar images from historical He I 10830 Å observations, enabling coronal analysis extending back decades before modern EUV imaging became available. The model achieves high fidelity on test data (CC=0.92 for full-disk morphology) and demonstrates physical plausibility when validated against SOHO, Yohkoh, and long-term solar activity proxies spanning 1974-2015.

This research addresses a significant gap in solar physics by enabling synthetic reconstruction of extreme ultraviolet coronal data from readily available historical observations. The diffusion model approach leverages the multi-decade availability of He I measurements, whose absorption properties correlate with coronal structure and magnetic topology, to bridge observational epochs separated by instrumental capabilities. The temporal co-alignment training strategy using SOLIS and AIA data from 2011-2015 provides robust physical grounding, while the month-based data split prevents temporal leakage.

The validation methodology demonstrates scientific rigor by cross-referencing reconstructed images against independent observational datasets spanning three decades and different instruments. Comparison against SOHO/EIT and Yohkoh/SXT observations confirms morphological preservation, while disk-integrated emission statistics align with established solar activity proxies like sunspot numbers and F10.7 radio flux. This multi-layered validation suggests the model captures genuine solar physics rather than learning spurious patterns.

The implications extend beyond academic solar physics into space weather forecasting and historical climate analysis. By enabling continuous coronal context over forty years rather than isolated snapshots, researchers can examine large-scale coronal evolution trends previously inaccessible. The framework's success demonstrates how modern machine learning techniques can retrospectively enhance historical scientific datasets. Future applications might include refining geomagnetic storm predictions and improving understanding of solar variability's terrestrial climate impacts, establishing this as a meaningful contribution to computational solar science and data reconstruction methodologies.

• →Diffusion model framework reconstructs synthetic EUV coronal images from historical He I observations with 0.92 correlation coefficient for full-disk morphology
• →Enables continuous solar coronal analysis spanning 1974-2015, bridging gap before modern EUV satellites became operational
• →Validation across multiple independent instruments (SOHO, Yohkoh) and solar activity metrics confirms physical plausibility of reconstructions
• →Model preserves coronal hole structures (0.84 CC) critical for understanding open magnetic field regions and space weather
• →Demonstrates machine learning's capability to retrospectively enhance historical scientific datasets for multi-decade trend analysis