<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN">
<html xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:mml="http://www.w3.org/1998/Math/MathML">
<link rel="stylesheet" type="text/css" href="../../../CSS_FULL/edps_full.css">
<body><div id="contenu_olm">

<!-- DOI: 10.1051/0004-6361/201015111 -->

<h2 class="sec">Online Material</h2>

<p></p><h2 class="sec"><a name="SECTION00080000000000000000"></a>
Appendix A: </h2>

<p>The rotational diagram in Fig.&nbsp;<a href="/articles/aa/full_html/2010/13/aa15111-10/aa15111-10.html#Comparison-App">A.1</a>
is complemented with ground-based data for the frequency range
300-400&nbsp;GHz observed at the CSO with comparable beam size by
Phillips et&nbsp;al. (<A NAME="aaref17"></A><a href="/articles/aa/full_html/2010/13/aa15111-10/aa15111-10.html#Phillips1992">1992</a>). The non-detection at 307&nbsp;MHz is surprising, but possibly an effect of optical depth. The data in Fig.&nbsp;<a href="/articles/aa/full_html/2010/13/aa15111-10/aa15111-10.html#Comparison-App">A.1</a>,
except 307&nbsp;GHz, are well fitted by a single rotational temperature
of 239&nbsp;K, suggesting that the observed levels are populated
according to an exponential distribution. The derived temperature and
column density infer an optical depth of<!-- MATH: $\tau < 0.1$ -->
 <IMG SRC="img23.png" ALT="$\tau &lt; 0.1$" ALIGN="MIDDLE" BORDER="0" HEIGHT="26" WIDTH="44">
for all lines except at 307&nbsp;GHz. The fitted line (dashed) corresponds to a column density of 
<!-- MATH: $8.5 (\pm 2)\times 10^{13}$ -->
<IMG SRC="img24.png" ALT="$8.5 (\pm 2)\times 10^{13}$" ALIGN="MIDDLE" BORDER="0" HEIGHT="30" WIDTH="85">&nbsp;cm<sup>-2</sup>, consistent with the value derived by Phillips et&nbsp;al. (<a href="/articles/aa/full_html/2010/13/aa15111-10/aa15111-10.html#Phillips1992">1992</a>). This leads to a beam averaged H<sub>3</sub>O<sup>+</sup> abundance of 
<!-- MATH: $4.2 (\pm 1)\times 10^{-10}$ -->
<IMG SRC="img25.png" ALT="$4.2 (\pm 1)\times 10^{-10}$" ALIGN="MIDDLE" BORDER="0" HEIGHT="30" WIDTH="92">
relative to&nbsp;H, to be compared with the theoretical value of 
<!-- MATH: $4 \times 10^{-10}$ -->
<IMG SRC="img26.png" ALT="$4 \times 10^{-10}$" ALIGN="MIDDLE" BORDER="0" HEIGHT="30" WIDTH="55">
reported by Bruderer et&nbsp;al. (<a href="/articles/aa/full_html/2010/13/aa15111-10/aa15111-10.html#Bruderer2010a">2010a</a>).

</p><p></p><div class="inset-old">
<table>
<tr><td><!-- init Label --><A NAME="Comparison-App"></A><!-- end Label--><A NAME="733"></A><A NAME="figure638" HREF="img27.png"><IMG SRC="Timg27.png" ALT="\begin{figure}
\par\resizebox{9cm}{!}{\includegraphics{15111FgA1.eps}}
\end{figure}" HEIGHT="88" WIDTH="101"></A><!-- HTML Figure number: 4 -->
</td>
<td class="img-txt"><span class="bold">Figure A.1:</span><p>
Rotational diagram of H<sub>3</sub>O<sup>+</sup>. Numbers indicate the frequencies in&nbsp;GHz of the observed lines. 
</td>
</tr><tr><td colspan="2"><a href="http://dexter.edpsciences.org/applet.php?pdf_id=4&DOI=10.1051/0004-6361/201015111" target="DEXTER">Open with DEXTER</a></td></tr>

</table></div>
<p></p><p></p>
</div></body></html>