Herschel investigation of cores and filamentary structures in L1251 located in the Cepheus flare

Abstract

Molecular clouds are the prime locations of star formation. These clouds contain filamentary structures and cores which are crucial in the formation of young stars. In this work, we aim to quantify the physical properties of structural characteristics within the molecular cloud L1251 to better understand the initial conditions for star formation. We applied the getsf algorithm to identify cores and filaments within the molecular cloud L1251 using the Herschel multi-band dust continuum image, enabling us to measure their respective physical properties. Additionally, we utilized an enhanced differential term algorithm to produce high-resolution temperature maps and column density maps with a resolution of ${13.5}^{\prime \prime }$. We identified 122 cores in the region. Of those, 23 are protostellar cores, 13 are robust prestellar cores, 32 are candidate prestellar cores (including 13 robust prestellar cores and 19 strictly candidate prestellar cores), and 67 are unbound starless cores. getsf also found 147 filament structures in the region. Statistical analysis of the physical properties (mass (M), temperature (T), size and core brightness (hereafter, we are using the word luminosity (L)) for the core brightness) of obtained cores shows a negative correlation between core mass and temperature and a positive correlation between (M/L) and (M/T). Analysis of the filaments gives a median width of 0.14 pc and no correlation between width and length. Out of those 122 cores, 92 are present in filaments ($\sim$ 75.4%) and the remaining were outside them. Out of the cores present in filaments, 57 ($\sim$ 62%) cores are present in supercritical filaments ($M_{\text{line}}>16M_{\odot }/\text{pc}$).

Introduction

Stars are found to be forming in the dense parts of the interstellar medium (ISM) called molecular clouds. Inside these clouds, star formation occurs in the large filamentary structures (André et al.2010; Molinari et al.2010). Using Herschel data, it has been found that these dusty structures are hierarchical in nature (Men’shchikov 2021; Pokhrel et al.2017) and contain smaller units such as clumps, cores, filaments and hub-filament structures. Investigating the statistics of these structures can provide clues on the formation and evolution of star-forming regions.

Herschel dust maps at sub-millimeter wavelengths have played a crucial role in showing the ubiquity of filamentary networks present in the cold ISM and their importance in star formation in molecular clouds. Surveys of multiple molecular clouds reveal a striking property of these filaments in that they tend to have a characteristic width of 0.1 pc (Arzoumanian et al.2011). The origin of this characteristic width is not yet clear, with multiple explanations claiming the width is a result of the thermodynamics of ISM gas (Larson 2005), to the magneto-sonic scale below which the turbulence becomes subsonic in diffuse, non-star-forming molecular gases (Padoan et al.2001). It has also been observed that the majority of prestellar cores form in ‘supercritical’ filaments where the mass per unit length exceeds the critical line mass. This is calculated by modeling the filaments as nearly isothermal long cylinders at a temperature of 10 K and is found to be $M_{\text{line}}>16$$M_{\odot }/\text{pc}$ (Inutsuka & Miyama 1997). These supercritical filaments are prone to gravitational instability, and as a result, they can undergo fragmentation (Inutsuka & Miyama 1997). This fragmentation process leads to the formation of denser regions called cores within the filaments (Zhang et al.2020).

Recent research into filaments has revealed more details regarding their structure and how it may impact core formation inside them. These include the existence of velocity gradients roughly perpendicular to the axis of the filament. These indicate the accretion of gas from the molecular cloud into the filaments (Shimajiri et al.2019; Chen et al.2020). This accretion flows along the filament in the radial direction as shown in Hacar et al. (2022) and Zhang et al. (2020). The relation of mass growth of cores within these filaments with this accretion was proposed by Gehman et al. (1996) and has been shown in works like Banerjee et al. (2006) and Gómez et al. (2018).

There is a strong link present between the formation of prestellar cores and filaments to star formation in these dense molecular clouds. These Herschel survey results are in remarkable agreement with theory and numerical simulations consistently showing that the ISM is highly filamentary on all scales and star formation is intimately connected with self-gravitating filaments (Hennebelle 2013). These simulations now successfully include turbulence, gravity, various cooling processes, magnetohydrodynamics and radiation from massive stars. These numerical advances have provided us with new and valuable insights into the physics of filaments and star formation, including the formation, fragmentation, and further evolution of filaments through accretion, and the central role of filaments in the rapid gathering of gas into cluster forming, dense regions (Könyves et al.2015; Li et al.2023).

L1251 is a molecular cloud present in Cepheus Flare at a distance of $\sim 340\pm 7$ pc, extending from $l\sim {100}^{\circ }$–${116}^{\circ }$, $b\sim 9^{\circ }$–${25}^{\circ }$ (Right Ascension $\sim {19}^{\text{h}}$–${23}^{\text{h}}$, Declination $\sim {68}^{\circ }$–${80}^{\circ }$) (Sharma et al.2022) as shown in Figure 1. It has an elongated structure with a cometary morphology, which is believed to have been formed due to its interaction with a supernova bubble (Grenier et al.1989; Sharma et al.2022). The structure of the cloud, shown clearly in Figure 2, alongside the elongated structures mentioned above is similar to a body flying at hypersonic speed across an ambient medium (Balázs et al.2004). The position angle of the elongated cloud is found to be ${107}^{\circ }$.

[See PDF for image]

Fig. 1

L1251 highlighted in the 857 GHz Plank map of the Cepheus flare and another molecular cloud L1228.

[See PDF for image]

Fig. 2

Herschel SPIRE 350 $\mathrm{\mu }$m image. The elongated structures are highlighted in the picture.

This paper aims is to study the structure of this cloud by identifying the cores and filaments present and studying their physical properties. Our method in extracting cores and filaments is very similar to the work done in Zhang et al. (2022), where they studied the Perseus molecular cloud. We used multi-wavelength Herschel data (2.1) to create high-resolution (13.5${}^{\prime \prime }$) column density and temperature maps using an improved difference-term algorithm. We then extracted cores and filaments using getsf along with their physical properties. Understanding the formation of prestellar cores is crucial for unraveling the early stages of star formation. By analyzing the properties of cores and filaments in L1251 using the new tool getsf, we may gain insights into the role of these structures in facilitating star formation. This can contribute to broader discussions on the influence of cloud morphology on star formation efficiency. We also get to compare the results of different software used to detect prestellar cores in the region. This work is similar to existing work, particularly with Di Francesco et al. (2020) and Pattle et al. (2016), but while the former study employs other algorithms getsources (Men’shchikov et al.2012) and getfilaments (Men’shchikov 2013) for core and filament extraction, respectively, using Herschel images of Cepheus Flare in 70, 160, 250, 350 and 500 $\mathrm{\mu }$m with pixel sizes of the images set as 3.2${}^{\prime \prime }$, 3.2${}^{\prime \prime }$, 6${}^{\prime \prime }$, 10${}^{\prime \prime }$ and 14${}^{\prime \prime }$, respectively, we use getsf (Men’shchikov 2021), which handles core and filament extraction simultaneously. By comparing our results with theirs, we can see the difference in the results of these algorithms, especially since getsf does not take any input from the user. Meanwhile, the latter study uses SCUBA-2 observations in 450 and 850 $\mathrm{\mu }$m wavelengths with maps on to 6${}^{\prime \prime }$ pixels and CSAR algorithm (Kirk et al.2013); while we use Herschel dust maps at 160, 250, 350 and 500 $\mathrm{\mu }$m. We also focus exclusively on L1251 and find correlations between the mass, luminosity, size and temperature of detected cores.

In Section 2, we discuss the data we have used in our analysis and the algorithm we use for identifying cores and filaments. We then discuss the results including the cores and filaments we extracted, their properties, and their correlations in Section 3. We also do a small comparison with the findings of Di Francesco et al. (2020). Finally, Section 4 concludes our work and lists our findings.

Data, observation and techniques

Herschel data

Herschel PACS images in 70 and 160 $\mathrm{\mu }$m (Poglitsch et al.2010) and Herschel SPIRE images in 250, 350 and 500 $\mathrm{\mu }$m (Griffin et al.2010) were used in our analysis (Figure 2). The SPIRE/PACS parallel-mode was used to make scan maps with a speed of 60${}^{\prime \prime }$ s${}^{-1}$. The beam sizes for Herschel observations in 70, 160, 250, 350 and 500 $\mathrm{\mu }$m are 8.4, 13.5, 18.2, 24.9 and 36.3${}^{″}$, respectively. The data was downloaded from ESA Herschel Science Archive.1 The observation id’s for the images used are 1342188654, 1342188655, 1342189663 and 1342189664.

We assume that each band of Herschel satisfies optically thin dust emission (Planck Collaboration et al.2014):

1

$$\begin{array}{c}\hfill I_{\nu }={\kappa }_0{(\nu /{\nu }_0)}^{\beta }r\mu m_{\text{H}}{N}_{{\mathrm{H}}_2}{B}_{\nu }(T),\end{array}$$where $I_{\nu }$ is the specific intensity, ${\kappa }_0$ is the emission cross-section at a reference frequency ${\nu }_0$, r is the dust-to-gas mass ratio, $\mu$ is the mean molecular weight, $m_{\text{H}}$ is the mass of a hydrogen atom, $N_{{\mathrm{H}}_2}$ is the gas surface density, $B_{\nu }(T)$ is the Planck function for dust at temperature, T.

A power-law form was assumed for the dust opacity (Roy et al.2014), ${\kappa }_{\lambda }=0.1\times {(\lambda /300\mathrm{\mu }\text{m})}^{-\beta }{\text{cm}}^2{\text{g}}^{-1}$. $\beta$ could vary with frequency, grain size, and grain temperature, but using a fixed value of $\beta =2$, the accuracy of surface density will not deviate by more than 50% (Roy et al.2014). We use Planck data as a reference benchmark and convolve each band of Herschel to the Planck resolution. The offset in flux density of each band of Herschel can be obtained by linear comparison (Molinari et al.2010). They are 42.9, 11.3, 3.9 and 0.8 MJy sr${}^{-1}$ at 160, 250, 350 and 500 $\mathrm{\mu }$m, respectively.

We also used these multi-wavelength dust maps to obtain high-resolution (13.5${}^{″}$) column density and temperature maps of the region using pixel-by-pixel SED fitting of the data to a modified blackbody function. Instead of convolving all images to the lowest resolution and using that directly for fitting, we used getsf, which works using the method described in Men’shchikov (2021), which employs a difference term algorithm to increase the resolution from 36.3${}^{″}$ as we would have obtained from the former method to 13.5${}^{″}$. The obtained density and temperature maps via hires are shown in Figures 3 and 4, respectively.

We now use the derived column density maps to obtain the total mass (M) of L1251. We use the formula:

2

$$\begin{array}{c}\hfill M=N({\text{H}}_2)\mu m_{\text{H}}A,\end{array}$$where $N({\text{H}}_2)$ is the derived column density value from the column density map Figure 3, $\mu$ is the mean weight of molecular gas taken to be 2.8 assuming that the gas is 70% molecular hydrogen by mass (Ward-Thompson et al.2010), $m_{\text{H}}$ is mass of hydrogen atom, and A is projected area. The boundary of the map was chosen by an observational mask where the usable part of the map that contained the cloud was selected. This selection was made empirically. The remaining part was considered background and was excluded from measuring mass. While making the map, getsf separates the background image. This way, the mass of the background component is subtracted from the mass of the cloud. Using this, we find the mass to be $\approx 2000$$M_{\odot }$.

[See PDF for image]

Fig. 3

13.5${}^{″}$ resolution column density map of L1251 made using getsf.

[See PDF for image]

Fig. 4

13.5${}^{″}$ resolution temperature map of L1251 made using getsf.

[See PDF for image]

Fig. 5

Positions of 122 cores identified via getsf overlaid on 250 $\mathrm{\mu }$m dust map. 19 strictly candidate prestellar cores are shown in black, 23 protostellar cores are shown in green, 13 robust prestellar cores are shown in red, and 67 unbound starless cores are shown in purple.

[See PDF for image]

Fig. 6

Histograms of core properties. (a) Histogram of masses of detected cores. (b) Histogram of radii of detected cores. (c) Histogram of temperatures of detected cores. (d) Histogram of luminosities of detected cores and (e) Histogram of column density at locations of cores. The graph is fitted with a power-law of index $\sim$7.5.

[See PDF for image]

Fig. 7

(a) No correlation between mass and luminosity of cores. (b) No correlation between luminosity and temperature of cores. (c) Correlation between mass and temperature of cores and (d) Correlation between (M/L) and (M/T) of cores.

getsf

We use getsf, a method for extracting sources and filaments in astronomical images using separation of their structural components (Men’shchikov 2021), to find, measure, and extract cores and filamentary structures in L1251. getsf identifies structures in astronomical images by separating their structural components. The basic steps involved in using getsf and its image processing are described below:

1. First, all the different images need to be resampled to images with the same pixel size and number. This is done via the built-in script—resample. An observational mask may also be needed to reduce computational time.

2. These resampled images are used to obtain column density and temperature maps using the built-in script—hires.

3. Then, getsf uses structural decomposition into single-scale images to isolate core and filamentary structures from their background. Noise due to local fluctuations is removed via image flattening.

4. These cleaned single-scale images are then combined over all the wavelengths given, cores and filaments are identified, and their physical properties are measured and cataloged using scripts fmeasure and smeasure.

5. The location and properties of detected structures are obtained in an extraction catalog.

getsf is a nearly fully autonomous algorithm with the only user input needed for hires being the offset in flux density (in MJy/sr) for the image, correcting the image absolute calibration, and the maximum size of the structure (in arcsecs) for detecting cores and filaments. More details on the working of getsf are in Appendix A.

Results and analysis

Core selection and classification

The obtained cores via getsf are selected by the criteria described in Appendix B. This selection is based on benchmark tests given in Men’shchikov (2021) and is based on properties like their size, detection significance, intensities, etc. One hundred twenty-two cores were identified by this method. Properties like their mass and bolometric luminosity can be found by pixel-by-pixel SED fitting the integrated flux measured at each wavelength by getsf to a modified blackbody function. The core size is defined as the mean deconvolved FWHM diameter at the resolution of 13.5${}^{″}$ of an equivalent elliptical Gaussian source: $R=\sqrt{\text{AFWHM}\times \text{BFWHM}-13.5^2}$, where AFWHM is the size of the major axis at half maxima and BFWHM is the size of the minor axis at half maxima of the detected core. Using these parameters, we classify the cores into different categories.

The classification is done by comparing the cores to a Bonnor–Ebert sphere (Bonnor 1956; Ebert 1955). The critical mass of a Bonnor–Ebert sphere – the maximum mass that can support itself against gravitational attraction is given by:

3

$$\begin{array}{c}\hfill M_{\text{crit}}=\frac{2.4R{c}_{\text{s}}^2}{G},\end{array}$$from Bonnor (1956) and Ebert (1955), where R is the radius of the sphere and $c_{\text{s}}$ is the isothermal speed of sound, which is calculated using the SED temperature of each core separately, and G is the gravitational constant.

1. The protostars in the region will have infrared emissions. The 70 $\mathrm{\mu }$m dust map is better suited to find regions of high temperature caused by these emissions. Hence, our first step is to do a separate getsf run with just the 70 $\mathrm{\mu }$m image to locate the protostars in the cloud. These cores are then checked via the criteria in Appendix B. All those who satisfy the criteria are termed protostellar cores. As an additional measure, we have also cross-checked these locations with the SIMBAD database to see if there are any known protostars in the vicinity of identified protostellar cores and also to remove any galaxies erroneously detected. The details of this are given in Appendix C.

2. If the ratio ${\alpha }_{\text{BE}}=M_{\text{crit}}/M_{\text{core}}\le 2$, the core is deemed self-gravitating. This is analogous to the criteria used to select self-gravitating objects based on the viral mass ratio (Bertoldi & McKee 1992). Cores meeting this criterion are called robust prestellar cores.

3. An empirical condition ${\alpha }_{\text{BE}}\le 5\times {(\sqrt{(\text{AFWHM}\times \text{BFWHM})}/13.5^{″})}^{0.4}$ obtained by Monte Carlo simulations in Könyves et al. (2015) is used to select candidate prestellar cores. The simulated sources in Könyves et al. (2015) are all gravitationally bound, but the above-mentioned mass criteria cannot fully select them, while this empirical formula can screen out 95% of the gravitationally bound sources. We have modified the denominator in the formula to 13.5 to account for the resolution of our column density map.

4. The remaining cores are termed as unbound starless.

This methodology of the classification is described in Könyves et al. (2015). Out of 122 total cores, we found that 23 are protostellar, 13 are robust prestellar cores, 32 are candidate prestellar (including 13 robust prestellar cores and 19 strictly candidate prestellar cores), and 67 are unbound starless cores. All selected robust prestellar cores also belong to the broader criteria of candidate prestellar cores and hence are counted with them. The 19 strictly candidate cores are those that do not fit the criteria for robust cores. The obtained cores are marked on the 250 $\mathrm{\mu }$m dust map of L1251 in Figure 5.

Table 1. Number of cores in filaments of varying linear densities.

Linear density (D)	Unbound	Candidate	Robust
($M_{\odot }$/pc)	starless (%)	prestellar (%)	prestellar (%)	Protostellar (%)
Not in filament	25 (37.31)	0 (0)	0 (0)	5 (21.74)
$D<8$	18 (26.87)	0 (0)	0 (0)	1 (4.35)
$8<D<16$	8 (11.94)	7 (21.88)	2 (15.38)	1 (4.35)
$D>16$	16 (23.88)	25 (78.13)	11 (84.62)	16 (69.56)

Statistical analysis of core properties

Using the data in the getsf extraction catalog, the temperature and column density maps, and Equation (2), we obtained the mass, size, temperature, and brightness of cores. These properties are shown in the histograms (see Figure 6). We obtained 0.2 $M_{\odot }$ as the median mass, 46.3${}^{″}$ (corresponding to 0.08 pc) as the median radius, 13.9 K as the median temperature, and 0.05 $L_{\odot }$ as the median luminosity.

We then looked for correlations between these properties. We found no correlation in the way luminosity varied with mass or temperature (Figure 7a and 7b) but obtained a negative correlation ($R^2=54.23\%$) of core mass with core temperature (Figure 7c) and a positive correlation ($R^2=72.97\%$) between (M/L) and (M/T) (Figure 7d).

Similiar correlations are also observed by Zhang et al. (2022) and Marsh et al. (2014). The negative correlation of core mass with temperature is consistent with our expectations. For a prestellar core, the primary heating source is the Interstellar Radiation Field (ISRF). As the mass of the prestellar core increases, the amount of dust shielding the interior of the core from the radiation increases, and hence, the temperature decreases.

The positive correlation obtained between (M/L) and (M/T) of the cores can be explained by considering the relation of the bolometric luminosity of the cores with their temperature. This relation is shown explicitly in Zhang et al. (2022).

By making a histogram of the column densities of the positions of cores, we see a sharp jump at a density of $\sim {10}^{21}{\text{cm}}^{-2}$ after which cores start to form in the cloud. The maximum number of cores are formed just after this density and their number decreases with increasing densities. The histogram can be fitted by using a power-law of index $\sim$7.5 as shown in Figure 6(e).

[See PDF for image]

Fig. 8

Linear densities of the 147 detected filaments in the region.

Detected filaments

Filaments are complicated in their shapes and widths, often interconnected with each other and various nearby branches, and have variable intensity along their crests. The length of a filament is also not well defined, and neither is it always clear which branch of the complex 3D structure of the filamentary structure belongs to which filament. getsf identifies and analyzes filamentary structures by isolating a basic skeleton structure. The simplified skeletons enable an easy selection and better measurements of only the well-behaving, preferably isolated (not blended), and relatively straight parts of the filaments. It identified 147 filament structures in the region. Using the inbuilt script fmeasure, we can measure the filamentary structure along its skeleton and find its properties.

Statistical analysis of filament properties

Using fmeasure to analyze filament properties on the 13.5${}^{″}$ column density map, we can find physical properties like their lengths, widths, linear densities (same as line mass), and masses. The working of fmeasure is described in Appendix A. Although the length of a filament is not objectively well defined, it is clear to understand that the longer a segment is, the more reliable it is as a structure to measure filament properties. Out of the 147 identified filaments, we selected the filaments for statistical analysis, which had a length >0.2 pc (at least twice the accepted filament width) and detection signal-to-noise ratio >2.5. This criteria leaves us with 70 filaments. Of them, 11 have a length/width ratio >3. For the selected filaments, we can see the variation in linear densities of the filaments in the region in Figure 8. From this picture, it is clear that the filaments near the center of the cloud have a higher average density than those near the boundaries (Figure 8), with clear regions of high filament densities in the middle and left part of the region. We note that the obtained median width of the identified filaments is $\sim$0.14 pc, in good agreement with the universal characteristic Herschel filament width of 0.1 pc Herschel filaments (Arzoumanian et al.2011). However, the universality of this characteristic width has been challenged in recent studies, which contend that filament widths are a function of the resolution of the data used to derive them (Panopoulou et al.2022).

[See PDF for image]

Fig. 9

The detected cores overlaid on detected filaments along with their linear densities.

[See PDF for image]

Fig. 10

Histograms of filament properties. (a) Histogram of lengths of detected filaments. (b) Histogram of widths of detected filaments. (c) Histogram of densities of detected filaments and (d) Histogram of masses of detected filaments.

We can now check the position of the cores and see how many of them are located in filaments and the linear density of the filaments they are present in. The methods used to measure linear densities by getsf are discussed in Appendix A. We can calculate the distance 1 pixel corresponds to as the distance to the cloud multiplied by resolution (in radians). Taking the resolution as ${13.5}^{″}$ and distance as 340 pc, we see that 1 pixel corresponds to $\approx 8.9\times {10}^{-9}$ pc. Therefore, 0.05 pc on both sides of the 1-pixel thick filament skeleton (for a total width of 0.1 pc) obtained via getsf corresponds to 5.6 pixels on each side. With this, we will claim a core to be present inside a filament if it is within 6 pixels of a filament. We can see from Figure 9 that the majority of cores are present in gravitationally supercritical filaments with linear densities >16 $M_{\odot }/\text{pc}$. Their distribution with linear density is given in Table 1. Other properties of the selected filaments are shown in the histograms in Figure 10.

Comparison with similar work

If we compare our results with those obtained in Di Francesco et al. (2020) who used getsources, an older version of the program getsf we used, we see that they obtain a total of 187 cores. Of them, 11 are protostellar, 53 are robust prestellar, 86 are candidate prestellar and 90 are unbound cores. We can see that using getsf on the same region gives us a lower number of cores but a higher number of identified protostellar cores identified via 70 $\mathrm{\mu }$m emission. The percentage of unbound cores obtained via getsf is 54.92% while it is 48.13% using getsources. They also found an extensive filamentary network in L1251 and observed 80–100% of cores to be in filaments, in agreement with our findings.

Conclusion

We studied the dust properties of molecular cloud L1251 in the Cepheus flare region using multi-band Herschel maps. We began with understanding the cloud structure and other physical properties by making temperature and column densities maps via the hires script of getsf. getsf was also utilized to locate and classify prestellar cores and filaments in the cloud. Using this, we identified 122 cores and 147 filaments. We then used getsf’s in-built scripts smeasure, fitfluxes and fmeasure to inspect the physical properties of cores, namely temperature, size, mass and luminosity, as well as filament masses, lengths, linear densities and widths.

Analysis of this data revealed that out of 122 cores, 92 are present in filaments (75.4%) and the remaining do not have an identified parent filament. Of the cores present in filaments, 57 (62%) are present in supercritical filaments ($M_{\text{line}}>16M_{\odot }/\text{pc}$). We also find a clear correlation between M vs T and (M/L) vs (M/T) of the detected cores.

The former correlation is explained by considering the shielding effect of dust grains on the core from interstellar radiation, while the latter correlation is a result of the relation of bolometric luminosity on temperature via Wein’s Law.

Table 2. SIMBAD objects corresponding to identified protostars.

No.	RA (J2000)	DEC (J2000)	SIMBAD identifier
1	338.8502558	75.2851396	IRAS $22343+7501$
2	337.4295695	75.2242914	[KBB2008] L1251A C2
3	337.8064044	75.2133953	[PWK2017] 62
4	337.844001	75.2057197	[PWK2017] 63
5	339.8243047	75.1591029	[PWK2017] 65
6	339.8120191	5.1800094	[PWK2017] 72
7	339.0847361	75.3707067	[PWK2017] 96
8	338.884616	75.2782663	2MASS $\text{J}22352664+7516369$
9	339.80096	75.2017797	2MASS $\text{J}22391329+7512161$
10	337.4995097	75.2344617	IRAS $22290+7458$
11	340.2499785	75.0652508	IRAS $22398+7448$
12	339.8728992	75.1822671	JCMTSE $\text{J}223931.5+751055$
13	337.6335849	75.2360765	NAME LDN 1251 A IRS 3
14	339.6975257	75.1921614	SSTgbs $\text{J}2238469+7511337$

Our results regarding core and filament properties are consistent with other studies for Herschel filaments of star-forming regions showing a characteristic filament width of $\sim$0.1 pc and most prestellar cores existing within dense filaments.We also see a stark column density threshold above which prestellar cores are found (André 2017).

Acknowledgements

We would like to thank Alexander Men’shchikov for his repeated assistance in downloading and using getsf, which made most of this work possible. The Herschel data was obtained from the ESA’s Herschel Science Archive (HSA). DD and AS thank the IIA VSP program for providing internship opportunities and supporting this work. The Basic Science Research Program supports C.W.L. through the National Research Foundation of Korea (NRF), funded by the Ministry of Education, Science and Technology (NRF-2019R1A2C1010851) and by the Korea Astronomy and Space Science Institute grant funded by the Korea government (MSIT; Project No. 2023-1-84000).

1

http://archives.esac.esa.int/hsa/whsa/.

References

André, P. Comptes Rendus Geosci.; 2017; 349, 187.2017CRGeo.349.187A [DOI: https://dx.doi.org/10.1016/j.crte.2017.07.002]

André, P; Men’shchikov, A; Bontemps, S et al. Astron. Astrophys.; 2010; 518, L102.2010A&A..518L.102A [DOI: https://dx.doi.org/10.1051/0004-6361/201014666]

Arzoumanian, D; André, P; Didelon, P et al. Astron. Astrophys.; 2011; 529, L6.2011A&A..529L..6A [DOI: https://dx.doi.org/10.1051/0004-6361/201116596]

Balázs, LG; Ábrahám, P; Kun, M; Kelemen, J; Tóth, LV. Astron. Astrophys.; 2004; 425, 133.2004A&A..425.133B [DOI: https://dx.doi.org/10.1051/0004-6361:20047059]

Banerjee, R; Pudritz, RE; Anderson, DW. Mon. Not. R. Astron. Soc.; 2006; 373, 1091.2006MNRAS.373.1091B [DOI: https://dx.doi.org/10.1111/j.1365-2966.2006.11089.x]

Bertoldi F., McKee C. F. 1992, Astrophys. J., 395, 140

Bonnor W. B. 1956, Mon. Not. R. Astron. Soc., 116, 351

Chen M.C.-Y., Di Francesco J., Rosolowsky E. et al. 2020, Astrophys. J., 891, 84

Di Francesco, J; Keown, J; Fallscheer, C et al. Astrophys. J.; 2020; 904, 172.2020ApJ..904.172D [DOI: https://dx.doi.org/10.3847/1538-4357/abc016]

Ebert, R. Zeitschrift fuer Astrophysik; 1955; 37, 217.1955ZA...37.217E

Gehman, CS; Adams, FC; Fatuzzo, M; Watkins, R. Astrophys. J.; 1996; 457, 718.1996ApJ..457.718G [DOI: https://dx.doi.org/10.1086/176766]

Gómez, GC; Vázquez-Semadeni, E; Zamora-Avilés, M. Mon. Not. R. Astron. Soc.; 2018; 480, 2939.2018MNRAS.480.2939G [DOI: https://dx.doi.org/10.1093/mnras/sty2193]

Grenier I. A., Lebrun F., Arnaud M., Dame T. M., Thaddeus P. 1989, Astrophys. J., 347, 231

Griffin M. J., Abergel A., Abreu A. et al. 2010, Astron. Astrophys., 518, L3

Hacar A., Clark S., Heitsch F. et al. 2022, Initial Conditions for Star Formation: A Physical Description of the Filamentary ISM, arXiv:2203.09562

Hennebelle P. 2013, Astron. Astrophys., 556, A153

Inutsuka S.-I., Miyama S. M. 1997, Astrophys. J., 480, 681

Kirk, JM; Ward-Thompson, D; Palmeirim, P et al. Mon. Not. R. Astron. Soc.; 2013; 432, 1424.2013MNRAS.432.1424K [DOI: https://dx.doi.org/10.1093/mnras/stt561]

Könyves, V; André, P; Men’shchikov, A et al. Astron. Astrophys.; 2015; 584, A91. [DOI: https://dx.doi.org/10.1051/0004-6361/201525861]

Könyves, V; André, P; Men’shchikov, A et al. Astron. Astrophys.; 2015; 584, A91. [DOI: https://dx.doi.org/10.1051/0004-6361/201525861]

Larson R. B. 2005, Mon. Not. R. Astron. Soc., 359, 211

Li X.-M., Zhang G.-Y., Men’shchikov A., et al. 2023

Marsh, KA; Griffin, MJ; Palmeirim, P et al. Mon. Not. R. Astron. Soc.; 2014; 439, 3683.2014MNRAS.439.3683M [DOI: https://dx.doi.org/10.1093/mnras/stu219]

Men’shchikov, A. Astron. Astrophys.; 2013; 560, A63.2013A&A..560A.63M [DOI: https://dx.doi.org/10.1051/0004-6361/201321885]

Men’shchikov, A. Astron. Astrophys.; 2021; 649, A89.2021A&A..649A.89M [DOI: https://dx.doi.org/10.1051/0004-6361/202039913]

Men’shchikov, A; André, P; Didelon, P et al. Astron. Astrophys.; 2012; 542, A81. [DOI: https://dx.doi.org/10.1051/0004-6361/201218797]

Molinari, S; Swinyard, B; Bally, J et al. Astron. Astrophys.; 2010; 518, L100.2010A&A..518L.100M [DOI: https://dx.doi.org/10.1051/0004-6361/201014659]

Padoan, P; Juvela, M; Goodman, AA; Nordlund, Å. Astrophys. J.; 2001; 553, 227.2001ApJ..553.227P [DOI: https://dx.doi.org/10.1086/320636]

Panopoulou, GV; Clark, SE; Hacar, A et al. Astron. Astrophys.; 2022; 657, L13.2022A&A..657L.13P [DOI: https://dx.doi.org/10.1051/0004-6361/202142281]

Pattle, K; Ward-Thompson, D; Kirk, JM et al. Mon. Not. R. Astron. Soc.; 2016; 464, 4255.2017MNRAS.464.4255P [DOI: https://dx.doi.org/10.1093/mnras/stw2648]

Planck Collaboration, Abergel A., Ade P. A. R. et al. 2014, Astron. Astrophys., 571, A11

Poglitsch, A; Waelkens, C; Geis, N et al. Astron. Astrophys.; 2010; 518, L2.2010A&A..518L..2P [DOI: https://dx.doi.org/10.1051/0004-6361/201014535]

Pokhrel R., Myers P., Dunham M. et al. 2017, Astrophys. J., https://doi.org/10.3847/1538-4357/aaa240

Roy A., André P., Palmeirim P. et al. 2014, Astron. Astrophys., 562, A138

Sharma, E; Maheswar, G; Dib, S. Astron. Astrophys.; 2022; 658, A55.2022A&A..658A.55S [DOI: https://dx.doi.org/10.1051/0004-6361/202140495]

Shimajiri Y., André P., Palmeirim P. et al. 2019, Astron. Astrophys., 623, A16

Ward-Thompson D., Kirk J. M., André P. et al. 2010, Astron. Astrophys., 518, L92

Zhang, G-Y; André, P; Men’shchikov, A; Wang, K. Astron. Astrophys.; 2020; 642, A76. [DOI: https://dx.doi.org/10.1051/0004-6361/202037721]

Zhang C., Zhang G.-Y., Li J.-Z., Li X.-M. 2022, Res. Astron. Astrophys., 22, 055012