A decade of CH₄, CO and N₂O in situ measurements at Lauder, New Zealand: assessing the long-term performance of a Fourier transform infrared trace gas and isotope analyser

• Main Authors: D. Smale, V. Sherlock, D. W. T. Griffith, R. Moss, G. Brailsford, S. Nichol, M. Kotkamp
• Format: Article
• Language: English
• Published: Copernicus Publications 2019-01-01
• Series: Atmospheric Measurement Techniques
• Online Access: https://www.atmos-meas-tech.net/12/637/2019/amt-12-637-2019.pdf

<p>We present a 10-year (January 2007–December 2016) time series of continuous in situ measurements of methane (<span class="inline-formula">CH₄</span>), carbon monoxide (CO) and nitrous oxide (<span class="inline-formula">N₂O</span>) made by an in situ Fourier transform infrared trace gas and isotope analyser (FTIR) operated at Lauder, New Zealand (45.04&thinsp;S, 169.68&thinsp;E, 370&thinsp;m&thinsp;a.&thinsp;m.&thinsp;s.&thinsp;l.). Being the longest continuous deployed operational FTIR system of this type, we are in an ideal position to perform a practical evaluation of the multi-year performance of the analyser. The operational methodology, measurement precision, reproducibility, accuracy and instrument reliability are reported.</p> <p>We find the FTIR has a measurement repeatability of the order of 0.37&thinsp;ppb (1<span class="inline-formula"><i>σ</i></span> standard deviation) for <span class="inline-formula">CH₄</span>, 0.31&thinsp;ppb for CO and 0.12&thinsp;ppb for <span class="inline-formula">N₂O</span>. Regular target cylinder measurements provide a reproducibility estimate of 1.19&thinsp;ppb for <span class="inline-formula">CH₄</span>, 0.74&thinsp;ppb for CO and 0.27&thinsp;ppb for <span class="inline-formula">N₂O</span>. FTIR measurements are compared to co-located ambient air flask samples acquired at Lauder since May 2009, which allows a long-term assessment of the FTIR data set across annual and seasonal composition changes. Comparing FTIR and co-located flask measurements show that the bias (FTIR minus flask) for <span class="inline-formula">CH₄</span> of <span class="inline-formula">−1.02</span>&thinsp;<span class="inline-formula">±</span>&thinsp;2.61&thinsp;ppb and CO of <span class="inline-formula">−0.43</span>&thinsp;<span class="inline-formula">±</span>&thinsp;1.60&thinsp;ppb are within the Global Atmospheric Watch (GAW)-recommended compatibility goals of 2&thinsp;ppb. The <span class="inline-formula">N₂O</span> FTIR flask bias of <span class="inline-formula">−0.01</span>&thinsp;<span class="inline-formula">±</span>&thinsp;0.77&thinsp;ppb is within the GAW-recommended compatibility goals of 0.1&thinsp;ppb and should be viewed as a serendipitous result due to the large standard deviation along with known systematic differences in the measurement sets. Uncertainty budgets for each gas are also constructed based on instrument precision, reproducibility and accuracy. In the case of <span class="inline-formula">CH₄</span>, systematic uncertainty dominates, whilst for CO and <span class="inline-formula">N₂O</span> it is comparable to the random uncertainty component.</p> <p>The long-term instrument stability, precision estimates and flask comparison results indicate the FTIR <span class="inline-formula">CH₄</span> and CO time series meet the GAW compatibility recommendations across multiple years of operation (and instrument changes) and are sufficient to capture annual trends and seasonal cycles observed at Lauder. The differences between FTIR and flask <span class="inline-formula">N₂O</span> measurements need to be reconciled. Trend analysis of the 10-year time series captures seasonal cycles and the secular upward trend of <span class="inline-formula">CH₄</span> and <span class="inline-formula">N₂O</span>. The <span class="inline-formula">CH₄</span> and CO time series have the required precision and accuracy at a high enough temporal resolution to be used in inversion models in a data-sparse region of the world.</p>