Direct measurement of ISM proper motion with image registration

Abstract

To date, quantification of the on-sky motion for interstellar clouds have relied on proxies such as young stellar objects (YSO) and masers. We present the first direct measurement of an interstellar cloud proper motion using the VISTA Star Formation Atlas (VISIONS) multi-epoch infrared images of the Corona Australis star-forming region. Proper motions are extracted by tracking the morphology of extended structures in the cloud complex based on image registration techniques implemented in SimpleITK. Our determined values ($μ_{α^*} \sim +15$ mas/yr, $μ_δ \sim -30$ mas/yr) are in good agreement with those obtained for YSOs and young stellar clusters in the region. This study demonstrates the potential of image registration for directly mapping the kinematics of nearby molecular clouds, opening a new window into the study of cloud dynamics.

Figures (5)

• Figure 1: Negative images of the investigated CrA ISM regions. Based on the reference epoch $H$-band image from VISIONS (CrA_wide_1_4_3_A). These images are reprojected and displayed in world coordinates -- the image registration was applied in pixel coordinates. No background removal procedure was used prior to registration.
• Figure 2: Visualisation of the gas proper motion for CrA ISM VISIONS 2. We computed the density peaks along the ISM structure (skeletons) across four epochs between 2017 and 2021 (including CrA_wide_1_4_3_C and CrA_wide_1_4_3_F), and overlaid them on the image from the first epoch. The red arrow shows the direction of motion derived using the methods described in Sect. \ref{['section:3']}.
• Figure 3: Negative images of the investigated Herbig-Haro objects and the YSO CHLT 15. Based on the reference epoch $H$-band image from VISIONS (CrA_wide_1_4_3_A). These images are reprojected and displayed in world coordinates -- the image registration was applied in pixel coordinates. A simple background removal procedure was used prior to registration.
• Figure 4: Extreme example of a failed registration for a case without background subtraction. This situation can occur even if the background model is identical in both images.
• Figure 5: Comparison of extracted proper motions with true values from simulations of extreme offsets (see Appendix \ref{['section:B']}). The simulation illustrates the difference between including and disregarding a background removal procedure with the presented method.