Reconstructing Synthetic SDO/AIA 193 A EUV Images from He I 10830 A Observations with Diffusion Model Translator
Researchers have developed a diffusion-based model, dubbed CH-aware DMT, to reconstruct synthetic EUV images of the sun from historical He I observations. This method aims to extend the availability of solar EUV imaging data into earlier periods before modern satellites like SDO were operational. The model was trained on co-aligned SOLIS He I and AIA 193 Å data and demonstrated strong performance in preserving EUV morphology and coronal hole structures. Its historical applicability was further validated by comparing reconstructions with data from SOHO/EIT, Yohkoh/SXT, and independent solar activity proxies, suggesting its utility for multi-decade analyses of coronal evolution. AI
IMPACT Enables multi-decade solar evolution studies by reconstructing historical EUV solar imagery.