# Quantitative inference of the $H_2$ column densities from 3 mm molecular emission: A case study towards Orion B

**Authors:** Pierre Gratier, Jérôme Pety, Emeric Bron, Antoine Roueff, Jan H. Orkisz, Maryvonne Gerin, Victor de Souza Magalhaes, Mathilde Gaudel, Maxime Vono, Sébastien Bardeau, Jocelyn Chanussot, Pierre Chainais, Javier R. Goicoechea, Viviana V. Guzmán, Annie Hughes, Jouni Kainulainen, David Languignon, Jacques Le Bourlot, Franck Le Petit, François Levrier  
**arXiv:** [2008.13417](https://arxiv.org/abs/2008.13417)  
**Date:** 2020-08-31  
**Summary page:** https://www.summarizepaper.com/en/paper/2008.13417v1/

## Key points

- Study focuses on accurately measuring molecular hydrogen (H2) column densities in cold molecular clouds
- Team explores using multi-molecule line emission from radio observations to infer NH2
- Supervised machine learning methods, specifically Random Forests, used on wide-field hyperspectral IRAM-30m observations of Orion B
- Results show predictor for NH2 can predict within a factor of 1.2 from Herschel-based estimates for conditions similar to those in Orion B
- Key molecular lines such as $^{13}$CO(1-0), $^{12}$CO(1-0), C$^{18}$O(1-0), and HCO$^+$(1-0) play significant roles in predicting NH2
- Incorporating N$_2$H$^+$(1-0) and CH$_3$OH(20-10) lines improves predictions for dense core conditions
- Research presents a promising approach to directly infer important physical parameters from millimeter domain molecular line emissions

## Summary

In their study "Quantitative Inference of the $H_2$ Column Densities from 3 mm Molecular Emission: A Case Study towards Orion B," Pierre Gratier and a team of researchers tackle the challenge of accurately measuring molecular hydrogen (H2) column densities in cold molecular clouds. Currently, measurements rely on dust emission observation in the far-IR or star counting due to H2 being unobservable in these conditions. To address this limitation, the team explores the use of multi-molecule line emission from radio observations to infer NH2. Using supervised machine learning methods, specifically Random Forests, on wide-field hyperspectral IRAM-30m observations of the Orion B molecular cloud, they train a predictor for NH2. Their results show that by using a limited set of molecular lines as input and Herschel-based dust-derived NH2 as ground truth output, they are able to predict NH2 within a typical factor of 1.2 from the Herschel-based estimates for conditions similar to those in Orion B. The analysis also reveals that certain key molecular lines such as $^{13}$CO(1-0), $^{12}$CO(1-0), C$^{18}$O(1-0), and HCO$^+$(1-0) play significant roles in predicting NH2. Furthermore, when differentiating between diffuse, translucent, filamentary, and dense core conditions, incorporating N$_2$H$^+$(1-0) and CH$_3$OH(20-10) lines improves predictions for dense core conditions. Overall, this research presents a promising approach to directly infer important physical parameters from millimeter domain molecular line emissions. Moving forward, further testing is needed to simultaneously infer multiple parameters, such as NH2 and far-UV illumination field, to enhance the applicability and accuracy of this method in diverse astronomical contexts.