The Nearby Evolved Stars Survey III: First data release of JCMT CO-line observations

TL;DR

This paper presents the first JCMT CO data release for the Nearby Evolved Stars Survey (NESS), a volume-limited program targeting ~850 AGB and RSG stars within 3 kpc to quantify gas and dust mass loss. It introduces a dedicated JCMT data-reduction pipeline and an empirical framework to derive CO line parameters, $^{12}$CO/$^{13}$CO ratios, and gas mass-loss rates (MLR) using the Ramstedt et al. (2008) formalism, along with dust-production rates (DPR) from SED fitting. The 485 sources analyzed show high CO detection rates (≈81% for CO(2-1) and ≈75% for CO(3-2)), a predominance of soft-parabola line profiles, and a broad range of MLRs and gas-to-dust ratios with notable differences between O-rich and C-rich chemistries. The study finds a broken-power-law relation between MLR and DPR with a saturation around ≈$5.2\times10^{-6}$ M$_\odot$ yr$^{-1}$ and a separate saturation in MLR versus $v_{inf}$ near ≈17 km s$^{-1}$, highlighting that there is no single gas-to-dust ratio across the population and that literature samples are biased toward brighter, higher-MLR sources. Overall, the NESS dataset demonstrates the value of a volume-limited approach for a complete view of the local AGB population and provides a robust baseline for future radiative-transfer modelling and population studies.

Abstract

Low- to intermediate-mass ($\sim$0.8$-$8 M$_\odot$) evolved stars contribute significantly to the chemical enrichment of the interstellar medium in the local Universe, making accurate mass-return estimates in their final stages crucial. The Nearby Evolved Stars Survey (NESS) is a large multi-telescope project targeting a volume-limited sample of $\sim$850 stars within 3 kpc in order to derive the dust and gas return rates in the Solar Neighbourhood, and to constrain the physics underlying these processes. We present an initial analysis of the CO-line observations, including detection statistics, carbon isotopic ratios, initial mass-loss rates, and gas-to-dust ratios. We describe a new data reduction pipeline to analyse the available NESS CO data from the JCMT, measuring line parameters and calculating empirical gas mass-loss rates. We present the first release of the available data on 485 sources, one of the largest homogeneous samples of CO data to date. Comparison with a large literature sample finds that high mass-loss rate and especially carbon-rich sources are over-represented in literature, while NESS is probing significantly more sources at low mass-loss rates, detecting 59 sources in CO for the first time and providing useful upper limits. CO line detection rates are 81% for the CO (2--1) line and 75% for CO (3--2). The majority (82%) of detected lines conform to the expected soft parabola shape, while eleven sources show a double wind. Calculated mass-loss rates show power-law relations with both the dust-production rates and expansion velocities up to $\sim 5 \times 10^{-6}$~\msunyr. Median gas-to-dust ratios of 250 and 680 are found for oxygen-rich and carbon-rich sources, respectively. Our analysis of CO observations in this first data release highlights the importance of our volume-limited approach in characterizing the local AGB population as a whole.

Figures (13)

• Figure 1: Distance vs. dust production rate (DPR) for the full NESS sample, in grey. Overlaid are the sources which have been observed and are included in the current analysis, coloured by tier. Cyan crosses show the locations of the carbon-rich sources.
• Figure 2: Three typical examples of the different CO line shapes, with the best-fit soft parabola shown with a red dotted line. From left to right: LP And (IRAS 23320+4316) shows a soft parabola profile, V360 And (IRAS 01556+4511) shows a double-wind profile, and ST Her (IRAS 15492+4837) shows an asymmetric profile with line wings.
• Figure 3: Distributions of MLRs derived from CO lines with a soft parabola shape, a double wind, or other non-soft-parabola shape.
• Figure 4: Plot of the $^{12}$CO/$^{13}$CO ratio as a function of mass-loss rate. Note that the optically thin C-rich IRAS 19008+0726 with a ratio of 69 and MLR of $4.6 \times 10^{-6}$ M$_\odot$ yr$^{-1}$ has been excluded for clarity.
• Figure 5: Plot of log(MLR) including error bars vs log(DPR) with a best-fit broken power law in dashed black. The grey lines show random draws from the posterior distribution, as an indication of the uncertainty in the fit. The marginal distributions for the MLR and DPR are shown as histograms. A histogram (dashed) is also shown for the cases where the MLRs are upper limits.