Accuracy and precision of photoacoustic spectroscopy not guaranteed

Because of their accuracy and precision for measuring gas concentrations, gas chromatographs (GC) are standard analytical instruments used in investigations of nitrous oxide (N2O) and carbon dioxide (CO2) exchange between the soil and the atmosphere. Iqbal et al. (2012) indicate that photoacoustic spectroscopy (PAS) performs similar to GCs to this end. We welcome this addition to the literature, given the increasing number of studies using PAS (e.g. Predotova et al., 2009; Leytem et al., 2011) and the few comparative analyses available (Ambus & Robertson, 1998; Yamulki & Jarvis, 1999). However, poor performance of PAS in some assessments (Flechard et al., 2005; Akdeniz et al., 2009) and data from our own tests (reported below) raise questions about whether Iqbal et al.'s (2012) results are generally applicable to PAS instruments or unique to the experimental conditions and calibration of their instruments. We tested three PAS instruments (PAS 1, 2, and 3: INNOVA Lumasense Technologies models 1412, 1302, and 1312, respectively) for their accuracy and precision by measuring gas samples of known concentration in a laboratory environment. Experiments with PAS 1 and 2 took place under the same experimental conditions and tested differences in performance between two instruments. Gas samples of known concentration were passed directly into the instruments commensurate with manufacturer's specifications (Lumasence, 2009). Separate experiments were conducted on PAS 3 to examine how ambient air temperature and sample moisture content affect the instrument's performance. We generated gases of known concentration by mixing calibration gas with synthetic air. Two meters of common airline allowed for proper mixing. A dew point generator was used to vary sample moisture content. Each gas cylinder was equipped with a flow regulator and the gas flow was maintained constant throughout the experiments. An Aerodyne Quantum Cascade Laser was used to verify the concentration of the gas mixture and showed that the targeted and obtained concentrations differed less than 0.5%. The manufacturer calibrated PAS 1, 2, and 3, 1, 6, and 4 months prior to the experiments, respectively, within the recommended once-yearly factory calibration cycle. These experiments build on those by Iqbal et al. (2012) by testing instrument performance (i) at lower greenhouse gas concentrations; (ii) across a range of gas concentrations; and (iii) for instruments calibrated at different times.

---

• View source

---

DOI:

https://doi.org/10.1111/gcb.12332

---

Altmetric score:

Dimensions Citation Count: