DEMI: Dynamic High-Resolution Imaging of the Earth's Upper Atmosphere

Imaging submesoscale atmospheric dynamics-features on spatial scales of kilometers to tens of kilometers-is essential for understanding key processes in the upper atmosphere, yet it remains a persistent challenge due to the implicit stationary assumption in conventional radar imaging techniques. Emerging radar systems such as MAARSY-3-D and EISCAT-3-D, along with advancements in multiple-input multiple-output (MIMO) radar technology, promise unprecedented spatial resolution. Yet, traditional methods still suffer from motion-induced blurring, which critically degrades features at these scales. To address this limitation, we propose a dynamic ensemble-based method for imaging (DEMI), a computational imaging approach that treats the atmosphere as a dynamic system and employs the ensemble Kalman filter (EnKF) to sequentially reconstruct time-varying reflectivity fields. Unlike static techniques, conventionally applied in radar imaging, such as Capon or maximum entropy (MaxEnt), DEMI explicitly accounts for target motion, significantly reducing motion blur. This dynamic modeling framework provides clearer, more accurate depictions of submesoscale structures while remaining computationally efficient through adaptive filtering. Results from both simulations and experimental data demonstrate that DEMI substantially improves spatial resolution and image quality, opening new opportunities to study previously unresolved submesoscale dynamics and their broader implications in the upper atmosphere.