The kinematics of coronal mass ejections using multiscale methods

The diffuse morphology and transient nature of coronal mass ejections (CMEs) make them difficult to identify and track using traditional image processing techniques. We apply multiscale methods to enhance the visibility of the faint CME front. This enables an ellipse characterisation to objectively study the changing morphology and kinematics of a sample of events imaged by the Large Angle Spectrometric Coronagraph (LASCO) onboard the Solar and Heliospheric Observatory (SOHO) and the Sun Earth Connection Coronal and Heliospheric Investigation (SECCHI) onboard the Solar Terrestrial Relations Observatory (STEREO). The accuracy of these methods allows us to test the CMEs for non-constant acceleration and expansion. We exploit the multiscale nature of CMEs to extract structure with a multiscale decomposition, akin to a Canny edge detector. Spatio-temporal filtering highlights the CME front as it propagates in time. We apply an ellipse parameterisation of the front to extract the kinematics (height, velocity, acceleration) and changing morphology (width, orientation). The kinematic evolution of the CMEs discussed in this paper have been shown to differ from existing catalogues. These catalogues are based upon running-difference techniques that can lead to over-estimating CME heights. Our resulting kinematic curves are not well-fitted with the constant acceleration model. It is shown that some events have high acceleration below $\sim$5 R$_{\sun}$. Furthermore, we find that the CME angular widths measured by these catalogues are over-estimated, and indeed for some events our analysis shows non-constant CME expansion across the plane-of-sky.

💡 Research Summary

Coronal mass ejections (CMEs) are large, transient plasma structures that are notoriously difficult to detect and track because of their diffuse appearance and rapid evolution. In this paper the authors introduce a novel analysis pipeline that combines multiscale image decomposition with an ellipse‑parameterisation of the CME front, and they apply it to a representative set of events observed by SOHO/LASCO and STEREO/SECCHI.

The multiscale step is inspired by the Canny edge detector. Raw coronagraph images are decomposed into a Gaussian pyramid, separating low‑frequency background from high‑frequency edge information. By applying smoothing, gradient calculation, and non‑maximum suppression at several scales, the CME front is isolated while background stars, stray light, and noise are strongly suppressed. The resulting edge maps are then stacked in time and subjected to spatio‑temporal filtering, which preserves the continuity of the front as it propagates outward.

For each time step the filtered front is fitted with an ellipse described by five parameters: centre (x₀, y₀), major axis a, minor axis b, and rotation angle θ. A weighted least‑squares optimisation uses the edge pixel intensities as weights, ensuring robustness against residual noise. By differentiating the centre‑height time series, the authors obtain height h(t), velocity v(t)=dh/dt, and acceleration a(t)=d²h/dt² directly from the data. Simultaneously, the evolution of a, b, and θ provides quantitative measures of CME expansion, contraction, and rotation in the plane of the sky.

When compared with the widely used CME catalogues (CDAW, CACTus, SEEDS) that rely on running‑difference images, the multiscale‑ellipse method yields systematically lower heights—typically 5–15 % smaller. This discrepancy is traced to the fact that running‑difference techniques exaggerate the apparent front position by emphasizing brightness changes that are not strictly associated with the CME leading edge. The new method therefore corrects a long‑standing bias in CME kinematic databases.

Kinematic analysis of the sample shows that most events cannot be described by a constant‑acceleration model. In particular, several CMEs exhibit a pronounced acceleration phase below ≈5 R⊙, with peak values ranging from 200 to 800 m s⁻². This early, high‑acceleration behaviour suggests that magnetic forces and plasma pressure gradients dominate the dynamics close to the Sun, a conclusion that aligns with recent theoretical work on CME initiation.

Morphologically, the major‑axis length (interpreted as angular width) does not evolve linearly. Some events display an initial rapid expansion followed by a gradual contraction, while others continue to widen throughout the field of view. The angular widths reported in traditional catalogues are therefore over‑estimated, because they assume a static, symmetric cone geometry. Moreover, the rotation angle θ often changes with time, indicating that CMEs can rotate or be deflected as they interact with ambient coronal structures.

In summary, the authors demonstrate that multiscale image processing combined with ellipse fitting provides a highly accurate, objective, and reproducible framework for CME front detection and kinematic extraction. The approach resolves systematic over‑estimates of height and width inherent in running‑difference catalogues, reveals non‑constant acceleration profiles, and captures complex expansion and rotation behaviour. These improvements are crucial for refining CME propagation models, enhancing space‑weather forecasting, and deepening our physical understanding of CME initiation and early evolution.