Morphological Comparison of Three Algorithms for 13CO Clump Detection in the Milky Way Imaging Scroll Painting

Abstract

The objective evaluation of molecular clump detection algorithms is pivotal for unraveling the hierarchical structure of the interstellar medium and for revealing the physical processes that drive star formation. However, systematic comparisons of morphological differences and structural characterization capabilities among algorithms applied to real observational data are still lacking. We apply FellWalker, FacetClumps, and Dendrograms to ${}^{13}$CO ($J=1-0$) spectral data from three representative regions of the Milky Way Imaging Scroll Painting (MWISP): 15.25${}^{\circ }$ $\le l\le$ 18.25${}^{\circ }$, 0${}^{\circ }$ $\le b\le$ 1.5${}^{\circ }$; 100${}^{\circ }$ $\le l\le$ 110${}^{\circ }$, $\left|b\right|\le$ 5.25${}^{\circ }$; and 180${}^{\circ }$ $\le l\le$ 190${}^{\circ }$, $\left|b\right|\le$ 5.25${}^{\circ }$. Here, l and b are respectively galactic longitude and latitude. Using Rotated J-values(RJ)-plots to classify the morphologies of ${}^{13}$CO clumps, we find that the algorithm FacetClumps identifies a substantially higher fraction of centrally overdense and quasi-circular clumps in all three regions (all exceeding 85%), whereas FellWalker and Dendrograms detect a larger share of elongated structures, ranging from 20% to 40% of the detected clumps. A comparison based on C${}^{18}$O counterpart statistics in the M16 field shows that the centrally overdense and quasi-circular ${}^{13}$CO clumps found by FellWalker and FacetClumps contain C${}^{18}$O counterparts at rates of 77% and 91%, respectively. We further compute $H_2$ column densities for the ${}^{13}$CO clumps and examine their relation to compactness. We find that clumps with higher column densities tend to be more compact, and high-column-density ${}^{13}$CO clumps are more likely to have associated C${}^{18}$O counterparts.