Atmos. Meas. Tech., 5, 2069–2080, 2012www.atmos-meas-tech.net/5/2069/2012/doi:10.5194/amt-5-2069-2012© Author(s) 2012. CC Attribution 3.0 License.

AtmosphericMeasurement

Techniques

Evaluation of continuous water vaporδD and δ18O measurementsby off-axis integrated cavity output spectroscopy

N. Kurita 1,2, B. D. Newman1,*, L. J. Araguas-Araguas1, and P. Aggarwal1

1International Atomic Energy Agency (IAEA), P.O. Box 100, 1400 Vienna, Austria2Japan Agency for Marine-Earth Science and Technology, 2–15 Natsushima-cho, Yokosuka, 237-0061, Japan* now at: Los Alamos National Laboratory, Los Alamos, USA

Correspondence to:N. Kurita (nkurita@jamstec.go.jp)

Received: 21 March 2012 – Published in Atmos. Meas. Tech. Discuss.: 16 April 2012Revised: 26 July 2012 – Accepted: 7 August 2012 – Published: 27 August 2012

Abstract. Recent commercially available laser spectroscopysystems enabled us to continuously and reliably measure theδD and δ18O of atmospheric water vapor. The use of thisnew technology is becoming popular because of its advan-tages over the conventional approach based on cold trap col-lection. These advantages include much higher temporal res-olution/continuous monitoring and the ability to make directmeasurements of both isotopes in the field. Here, we evaluatethe accuracy and precision of the laser based water vapor iso-tope instrument through a comparison of measurements withthose found using the conventional cold trap method. A com-mercially available water vapor isotope analyzer (WVIA)with the vaporization system of a liquid water standard (Wa-ter Vapor Isotope Standard Source, WVISS) from Los GatosResearch (LGR) Inc. was used for this study. We found thatthe WVIA instrument can provide accurate results if (1) cor-rection is applied for time-dependent isotope drift, (2) nor-malization to the VSMOW/SLAP scale is implemented, and(3) the water vapor concentration dependence of the isotopicratio is also corrected. In addition, since the isotopic value ofwater vapor generated by the WVISS is also dependent onthe concentration of water vapor, this effect must be consid-ered to determine the true water vapor concentration effecton the resulting isotope measurement.

To test our calibration procedure, continuous water va-por isotope measurements using both a laser instrument anda cold trap system were carried out at the IAEA IsotopeHydrology Laboratory in Vienna from August to Decem-ber 2011. The calibrated isotopic values measured usingthe WVIA agree well with those obtained via the cold trapmethod. The standard deviation of the isotopic difference

between both methods is about 1.4 ‰ forδD and 0.28 ‰ forδ18O. This precision allowed us to obtain reliable valuesfor d-excess. The day-to-day variation ofd-excess measuredby WVIA also agrees well with that found using the coldtrap method. These results demonstrate that a coupled sys-tem, using commercially available WVIA and WVISS in-struments can provide continuous and accurate isotope data,with results achieved similar to those obtained using the con-ventional method, but with drastically improved temporalresolution.

1 Motivation

Global monitoring of oxygen and hydrogen isotope con-tents in precipitation has been carried out by the Interna-tional Atomic Energy Agency (IAEA) in cooperation withthe World Meteorological Organization (WMO) for morethan 40 yr through the IAEA/WMO Global Network of Iso-topes in Precipitation (GNIP). GNIP has led to a great dealof study on the temporal and spatial variations of environ-mental stable water isotopes (δ18O andδD) and is widelyused in various scientific disciplines, such as hydrology, cli-matology, ecology and biology (e.g.Araguas-Araguas et al.,2000; Hoffmann et al., 2000; Bowen et al., 2005). On theother hand, although water vapor is ubiquitous in the atmo-sphere, the availability of isotopic data on atmospheric wa-ter vapor is quite limited. One of the main reasons for thisscarcity is that water vapor sampling presents more difficul-ties than precipitation sampling. Traditionally, atmosphericwater vapor samples for isotope analysis were trapped in a

Published by Copernicus Publications on behalf of the European Geosciences Union.

2070 N. Kurita et al.: Evaluation of water vapor isotope analyzer

cold bath filled with a cooling agent, such as liquid N2 ordry ice/ethanol. Next, a collected water sample was extractedfrom the cold trap and analyzed using isotope ratio massspectrometry (IRMS) or laser spectroscopy methods from theliquid water samples. Because incomplete trapping of watervapor results in isotope fractionation, self-designed trap sys-tem (Schoch-Fischer et al., 1984; Yakir and Wang, 1996; Heand Smith, 1999; Uemura et al., 2008) have been developedto meet required trapping efficiency. In addition, extensivelaboratory work is necessary to extract a vapor sample from acold trap. More recently, alternative moisture trapping tech-niques have been reported, resulting in more sophisticatedmethods (Helliker et al., 2002; Han et al., 2006; Peters andYakir, 2010). However, these methods are still labor intensiveand require samples to be prepared in a laboratory.

A more recent alternative utilizes laser based instru-ments (e.g.Kerstel et al., 2006) and instruments based onoptical spectroscopy for water vapor isotope analysis (e.g.Griffith et al., 2006). These instruments can directly mea-sure atmospheric water vapor isotopes in real-time with hightemporal resolution and without the time-consuming physi-cal collection of condensed moisture. In addition, the latestinstruments are of a reasonably small size and power con-sumption is low, allowing for the application of this tech-nique to in-situ measurements (Lee et al., 2005; Gupta et al.,2009; Sturm and Knohl, 2010; Aemisegger et al., 2012). No-tably, Lee et al.(2006) have reported continuousδD watervapor records for New England, USA, for a period of oneyear, using a tunable diode laser (TDL) spectroscopy sys-tem. However, because of instrumental nonlinearity, drift-ing signal, dependence on water vapor concentration, tem-perature sensitivity, and pressure variations in the absorp-tion cell, regular and frequent calibration with a water va-por standard of known isotope content is required for long-term field observations. In addition, a required water vaporisotope standard of known isotopic content is not commer-cially available. Therefore, custom-made vaporizer systems,which can provide a frequent supply of water vapor of knownisotope content to an in-situ instrument, had to be devel-oped before such an instrument could be deployed in thefield (Lee et al., 2005; Iannone et al., 2009; Schmidt et al.,2010; Sturm and Knohl, 2010; Gkinis et al., 2010). This rep-resented a major limitation for the widespread use of thistechnique. However, a recently developed robust vaporizersystem, coupled with the laser instrument, has become com-mercially available.Tremoy et al.(2011) have tested the per-formance of the calibration system (Standard Delivery Mod-ule, SDM) produced by Picarro, Inc. coupled with their laserinstrument (Wavelength- Scanned Cavity Ring-Down Spec-troscopy, WS-CRDS).

In this study, we use another commercially available in-strument based on off-axis integrated cavity output spec-troscopy, manufactured by Los Gatos Research, Inc. (LGR)(Water Vapor Isotope Analyzer, WVIA) coupled with anaccessory device for the vaporization of a liquid water

standard (Water Vapor Isotope Standard Source, WVISS).The WVISS is programmable from the WVIA and thus thiscoupled system is capable of automatically conducting acalibration routine at specific intervals. A detailed assess-ment of the WVIA has already been undertaken in previousstudies (Wang et al., 2009; Sturm and Knohl, 2010; Ramboet al., 2011). Sturm and Knohl(2010) have reported a strongconcentration dependence and temperature sensitivity of theWVIA. Rambo et al.(2011) have used the same coupledsystem with the WVISS and proposed a routine calibrationprocedure to correct for both concentration dependence andtemperature sensitivity. Recently,Aemisegger et al.(2012)completed an evaluation of this coupled system through com-parison with the WS-CRDS, in conjunction with the SDM.These studies have highlighted the usefulness of this systemto the field monitoring, however long-term water vapor iso-tope monitoring testing across several seasons has not beenreported yet. In particular, the performance and reliability ofthe calibration system (WVISS) must be stringently exam-ined before it can be applied in field conditions. Here, we de-tail the results of extensive laboratory experiments to developa calibration procedure for routine field application. Then,the calibration procedure developed from the laboratory ex-periment was evaluated through a long-term monitoring pro-gram of water vapor isotopes, using both the laser instrumentand the conventional cryogenic moisture trapping method.

2 Experimental

2.1 Laser spectroscopy system

The laser spectroscopy system used to measure water vaporisotopes (δ18O andδD) was the model DLT-100 water va-por isotope analyzer (WVIA, LGR Inc., Mountain View, CA,USA) manufactured in June 2009 (model 908-0004). To opti-mize performance of the WVIA, the operating software wasupdated to the latest version in May 2011. This analyzer isbased on an off-axis integrated cavity output spectroscopysystem (OA-ICOS) using a semiconductor diode laser with awavelength of around 1.39 µm (e.g.Bear et al., 2002). Herewe describe the OA-ICOS measurement technique briefly. InOA-ICOS, the laser beam is directed off-axis with respectto the optical cavity. The laser wavelength is swept throughthe selected absorption line of each species (H2O, HDO, andH18

2 O) and the transmitted laser intensity through a cavityis recorded. In addition, to determine effective optical pathlength in the cavity at each wavelength, ring down-time mea-surement is made by rapidly switching off the laser diode,which is measuring decay of the light intensity over time.The typical ring-down time is around 10 µs at 10 000 ppmH2O concentration and it corresponds to the effective pathlength of 3 km in a 0.59 m-long cell. This path length is morethan 10 times longer than that of a multi-pass cell and resultsin a high signal to noise ratio for isotope measurement. The

Atmos. Meas. Tech., 5, 2069–2080, 2012 www.atmos-meas-tech.net/5/2069/2012/

N. Kurita et al.: Evaluation of water vapor isotope analyzer 2071

Fig. 1.Experimental setup for(a) laboratory test and(b) field evaluation.

mixing ratio χ of each water isotope species is determinedthrough integration of the absorption spectrum as follows:

χx =1

Sx Leff P

∫ln

(Io

Iν

)d ν

whereIν is the transmitted laser intensity at wavelengthν,Io is the initial intensity,P is gas pressure in the cell,S isabsorption line strength,Leff is effective optical path length.The subscriptx represents each water isotope species. Calcu-lated mixing ratios of water isotopologues are converted intoa δ value with respect to the international standard ViennaStandard Mean Ocean Water (V-SMOW) (Coplen, 1996).

δ =R

RVSMOW− 1

whereR is the atomic ratio, D/H or18O/16O, respectively.

2.2 Experimental setup for water vapor isotopemeasurement

A sample of air is introduced into an absorption cell (op-tical cavity) of the WVIA via an external pump (KNF,N920AP.29.18) downstream of the instrument at a constantflow rate of 0.5 l min−1. The volume of an absorption cellis around 830 ml, while inside pressure is controlled at37± 0.007 hPa via a pressure controller and temperature ismaintained at 49◦C to avoid the condensation of water va-por from ambient air in the cell. Based on these figures,the exchange time is estimated to be approximately 4 s. Thedata output frequency for water isotopologue measurementsis 1 Hz.

To introduce water vapor standards into the WVIA, weused an LGR Water Vapor Isotope Standard Source (WVISS)device, manufactured in April 2010 (see Fig.1a). It consistsof a two-stage air dryer unit and a heated evaporation jar

(1 l) for accepting water from a 0.5 l container via a nebu-lizer (part no. 800-1-005-01-00; Savillex Ltd., Eden Prairie,MN, USA). Compressed room air is passed through a des-iccant dryer (part no. MDH5-FLE-S37, Twin tower engi-neering, Broomfield, CU, USA) and a column of granularDrierite (size 8–20 mesh), providing dry air into the sys-tem (H2O concentration< 10 ppm). The compressed dry airstream from the drying unit is then split into two lines: onegoes into the evaporation jar and the other supplies air to thenebulizer. The nebulizer aspirates liquid water from a wa-ter reservoir at a rate of 50 µl min−1 and injects tiny waterdroplets into the evaporation jar. The jar is heated to about80◦C so that droplets are immediately evaporated; the result-ing vapor should have the same isotopic composition as thatof the source water. The humidity of the output air can be var-ied from 2500 to 25 000 ppm by adjusting the dry air mixingratio in the evaporation jar. The flow of dry air is controlledby a mass flow controller (part no. FMA5420, OMEGA En-gineering, Stamford, CT, USA, 0–5 l min−1).

The output tubing from the jar is connected to a T-shapesplitter via 3/8 inch (9.52 mm) Teflon tubing. One outlet con-stitutes an exhaust to room air, whereas the other outlet isdirected to the analyzer through a 3-way valve, which con-trols airflow by switching between the ambient air inlet andWVISS-generated standard gas. The WVISS-generated stan-dard gas is drawn by an external pump downstream of theanalyzer (see above). The tubing joining the WVISS and theanalyzer is made of 1/4 inch (6.35 mm) Teflon tubing (with alength of around 1 m). The dilution rate with dry air and valveoperation is controlled by the WVIA. Even though standardgas is released from the heated chamber (80◦C), there is noclear temperature shift in the optical cavity when the valveis switched from the ambient air inlet to the standard gasline. Raw data from the analyzer is stored on the internalhard disk, and is transferred via ethernet cable. Calibration is

www.atmos-meas-tech.net/5/2069/2012/ Atmos. Meas. Tech., 5, 2069–2080, 2012

2072 N. Kurita et al.: Evaluation of water vapor isotope analyzer

Table 1.δD andδ18O values of water standards measured using the IRMS and the WVIA.

Sample δD δ18O N

δIRMS δWVIA 1δ δIRMS δWVIA 1δ

Std 11 −0.1 0.08Std 6 −61.3 −62.4± 1.3 1.07 −8.65 −8.69± 0.23 0.04 15Std 12 −86.4 −86.4± 0.7 0.04 −12.03 −11.93± 0.14 0.10 15Std X1 −138.3 −138.5± 1.0 0.22 −18.36 −18.35± 0.16 0.02 20Std 9 −189.2 −190.0± 0.7 0.80 −24.76 −24.69± 0.16 0.07 15Std X2 −225.7 −226.3± 0.6 0.62 −29.45 −29.44± 0.15 0.01 20Std 13 −257.7 −257.0± 1.4 0.68 −33.39 −33.22± 0.24 0.17 15Std X3 −332.7 −331.1± 0.6 1.52 −42.44 −42.35± 0.17 0.09 20Std 10 −397.9 −50.87

undertaken as a post-processing step using the external com-puter system.

2.3 Performance test of the WVISS

To test the assumption that no fractionation occurs in theWVISS, air generated in the unit was physically collectedusing a cold trap (see Fig.1a). A water reservoir was filledwith an in-house laboratory water standard (δD =−78 ‰ andδ18O =−11.3 ‰), then the WVISS was run continuouslyduring vapor trapping (2–6 h) to obtain a sufficient amountof water for isotope analysis. WVISS air exhaust was drawnby an external pump at a flow rate of 1.5 l min−1, then wa-ter vapor was collected with a glass trap submerged intoa dry-ice/isopropanol mixture at−78◦C. This experimentwas repeated more than 10 times at 6 levels of water vaporconcentration (3000 ppm, 4000 ppm, 5500 ppm, 6000 ppm,8000 ppm, 10 000 ppm). In this study, we used a custommanufactured glass trap for vapor collection. This trap hasbeen used for water vapor sample collection in various fieldprojects (Kurita and Yamada, 2007; Kurita, 2011; Kuritaet al., 2011). Along with this, we undertook a trapping testusing the WVISS/WVIA system before starting the labora-tory experiment. A glass trap was inserted between WVISSand WVIA and output humidity from a cold trap was mon-itored by the WVIA. This pretest confirmed complete trap-ping of water vapor because the reported humidity from theWVIA is less than 10 ppm during such a trapping period.

The collected amount of water was less than 1.5 g. Sam-pling time varied from two to six hours, depending on theconcentration. After sample collection, water in the trap wassubsequently thawed and poured into a 9-ml glass vial. Iso-topic analysis of the collected water was undertaken at theIAEA Isotope Hydrology Laboratory using IRMS (Deltaplus, Thermo Scientific, Bremen, Germany) with a locallydeveloped equilibration device and through cavity ring-downspectroscopy isotopic water analysis (model L1102-i; PicarroInc., Sunnyvale, CA, USA) with a CTC Analytics autosam-pler (model HTC-PAL; Leap Technologies, Carrboro, NC,

USA). Measurement precision, including internal and exter-nal variation, is better than± 0.15 ‰ forδ18O and± 2 ‰ forδD.

During vapor trapping, the WVISS generated vapor wasintroduced into the WVIA for 10 min in every hour. Regard-ing comparison with the values of water vapor samples col-lected by cold trap method, the time-averaged isotopic valuesof WVIA measurements were calculated from the last 5 minof each 10 min continuous measurement run every hour.

2.4 Accuracy and precision of the WVIA measurement

Isotope measurements of nine in-house water standards withdifferent δ18O and δD contents (Table 1) were performedto help determine a routine calibration procedure for thecoupled WVIA and WVISS system. The water standardsranged in isotope values from−397.9 to−0.1 ‰ for δDand from−50.87 to−0.08 ‰ for δ18O. Each water stan-dard was stored in a 9-ml glass vial. The nebulizer feed tubewas inserted into a vial of standard, and the mixing ratioof the WVISS-generated vapor was set to 10 000 ppm. Va-por was continuously introduced into the WVIA and the iso-tope data for the last 10 min of the 20 min measuring peri-ods was averaged. After a 20 min measurement period, thewater bottle with the isotope standard was replaced. To re-duce memory effects from residual water, water in the tub-ing was flushed away with dry air before the tube was in-serted into a next standard. As described inSturm and Knohl(2010), the WVIA response time is short, thus we did not ob-serve any memory effect in the isotopic data during the final10 min period, even though the most negative water standard(STD10) was measured just after the least negative waterstandard (STD11). Measurements of the nine water standardswere performed during the same day (the measurement se-quence was random) and the experiment was repeated morethan 15 times.

Atmos. Meas. Tech., 5, 2069–2080, 2012 www.atmos-meas-tech.net/5/2069/2012/

N. Kurita et al.: Evaluation of water vapor isotope analyzer 2073

2.5 Ambient air analysis

To evaluate the overall calibration procedure, we measuredwater vapor isotopes in ambient air using the WVIA-WVISScoupled system from August to December 2011. A samplingtube for outdoor air (5 m length of BEV-A-Line V tubing,1/4 inch O.D.) was connected to a 3-way valve attached tothe WVISS (see Fig.1b). The outdoor air tube was used tosample the ambient water vapor present in Vienna, Austriaoutside our laboratory (tubing was routed through a window).A 1.0 µm filter with PTFE membrane (part no. 10463523,Whatman, Dassel, German) was placed where the tube entersthe WVISS. Air was drawn via the WVIA external pump ata flow rate of 0.5 l min−1. Every 50 min, the 3-way valve au-tomatically switched from ambient/outdoor inlet to WVISSstandard air, whereupon standard air with a H2O concentra-tion of 10 000 ppm was introduced to the WVIA for 10 min.After finishing the reference gas measurement, the valveswitches back and ambient air sampling is resumed.

Atmospheric water vapor sampling using the cold trapmethod was also carried out to evaluate the validity oflaser based water isotope measurements. Air sampling tubing(10 m length of PVC tubing, 14 mm O.D.) with an inlet at thesame location as the WVISS ambient inlet was attached to aglass trap, which is the same as that used in the WVISS per-formance test (Fig.1b). Air was drawn at a rate of 1.5 l min−1

through the trap immersed into a dry-ice/isopropanol bath for6 to 15 h. The amount of water vapor trapped ranged from1.6 to 13.2 g. We collected 154 samples from August to De-cember 2011. The isotopic analysis of these samples was car-ried out using the procedure described above.

3 Results and discussion

3.1 Performance test of the WVISS

The mean isotopic values and the statistical distribution ofthe WVISS-generated air collected by cold trap, at severalwater vapor concentrations are shown in Fig.2. There wasa large spread in isotopic values, greater than analytical er-ror, arising from water isotope measurement at each vaporconcentration (see Sect. 2.3). To examine whether a largespread in isotopic values in WVISS-generated air is relatedto WVISS vapor production, the values of water vapor sam-ples using the cold trap (δTRAP) technique are compared withWVIA-reported raw isotopic values (δWVIA ) in Fig. 3. ForδD, the spread of values from the cold trap samples is almosttwice as big as that of the WVIA reported values. The vari-ation in δDWVIA values is less than±2 ‰ from the meanvalue. TheδDTRAP values include analytical error, whichtake place during liquid water isotope measurement (around2 ‰). In addition, decanting small amounts of trap water(less than 1.5 g) into vials leads to additional scatter. Thevariation in δDTRAP is similar or less than the analytical

−12.0

−11.5

−11.0

−10.5

3000 4000 5500 6000 8000 10000

H2O [ppm]

−80

−79

−78

−77

−76

−75

−74

−73

−72

Fig. 2. Analytical variations in theδD and δ18O of cold trappedvapor sampled from a WVISS exhaust line with different H2O con-centrations. Repeated analysis of standard water (δD =−78 ‰ andδ18O =−11.3 ‰) was run at 6 levels of vapor concentration from3000 to 10 000 ppm. The broken line represents known isotopic val-ues determined via IRMS measurement. The box and whisker plotsof δ18O data represent the median (dark bar), 25–75 % quantilerange (inter-quantile range [IQR] shown as the box), and maximumand minimum values (whiskers) for repeated experiments.

uncertainty of cold trap samples. This uncertainty may resultin a large spread and thus data points plot under a 1 : 1 line.However, there is a positive correlation betweenδDWVIA andδDTRAP (R = 0.369) and a similar feature that theδDWVIAvalues at 3000 ppm showed slightly enriched values, similarto theδDTRAP (see purple dots in Fig.3a). Although the am-plitude of a spread may be overestimated due to analyticaluncertainty of cold trapped samples, the variation ofδDTRAPvalues may reflect some isotopic shift in WVISS-generatedair.

For δ18O, the relationship betweenδ18OWVIA andδ18OTRAP is scattered across the entire water vapor con-centration range (3000 to 10 000 ppm) because of the time-dependent isotope variation (see next section). However, alinear correlation betweenδ18OWVIA andδ18OTRAP was ob-served in the measurements at 10 000 ppm when the time-dependent isotopic variation was relatively small. The pointsat 10 000 ppm measured between 27–29 November 2011 (redcircles) are clearly distributed along a 1:1 line and the spreadis significantly larger than the analytical error (0.15 ‰)(Fig. 3b). This suggests that a large spread of 10 000 ppmmay reflect isotopic changes in WVISS-generated air. Thisis supported by the fact thatδ18OTRAP values negatively cor-relate with water vapor concentration in WVISS-generatedair (R =−0.55). The mean water vapor concentration during

www.atmos-meas-tech.net/5/2069/2012/ Atmos. Meas. Tech., 5, 2069–2080, 2012

2074 N. Kurita et al.: Evaluation of water vapor isotope analyzer

−11.4

−11.2

−11.0

−10.8

−10.6

−10.4

−10.2

−11.8 −11.6 −11.4 −11.2 −11.0 −10.8 −10.6

−80

−78

−76

−74

−72

−80 −78 −76 −74 −72

COLD TRAP COLD TRAP

LAS

ER

(R

AW

)

1:1

10000 ppm

Nov27-29

3000 ppm

3000 - 10000 ppm

Fig. 3. A comparison of cold trap vapor analysis with direct vapor measurement using the WVIA:(a) δD in a concentration range of3000–10 000 ppm; the purple points represent values at 3000 ppm; and(b) δ18O at 10 000 ppm only; the red plots represent data from 27–29 November 2011. Broken lines represent the 1 : 1 relationship. These WVIA measurements are the uncorrected and uncalibratedδD andδ18O values.

vapor trapping varied, ranging from 9500 ppm to 10 500 ppmwithin each experiment. The maximum (minimum) value ofδ18OTRAP corresponds to the lowest (highest) vapor con-centration. From these results, we can infer that the largespread inδ18O values of WVISS-generated air at each con-centration may be primarily related to WVISS vapor pro-duction. The average standard deviation in the range of3000 to 10 000 ppm is about 0.25 ‰ forδ18O. In addition toa large spread in isotopic values, Fig.2 shows that the meanisotopic value of WVISS-generated air varied in responseto changing water vapor concentration. Measuredδ18O athigher concentrations (6000 to 10 000 ppm) is higher thanthe known value, and at lower water vapor concentrations(3000 to 5500 ppm),δ18O values tend to be more negativethan for higher concentrations. The mean value shifts from−11.1 ‰ at higher concentrations to−11.4 ‰ at lower con-centrations. The difference between them is statistically sig-nificant (p < 0.05). Interestingly, this tendency runs oppositeto the gradual decrease inδ18OTRAP with water vapor con-centrations of 10 000 ppm. This suggests that factors result-ing in a largeδ18O spread are not primary drivers of concen-tration dependence; several effects may contribute to chang-ing the isotopic values of WVISS-generated air. However,identification of the source of this bias and explaining thelarge spread in isotopic values of WVISS-generated air arebeyond the scope of this paper. On the other hand,δD datado not suggest the same behavior. A positive bias is observedat all concentrations. There is also no systematic difference inthe isotopic contents of water vapor for concentrations rang-ing from 3000 to 10 000 ppm.

3.2 Accuracy and precision of the WVIA measurement

3.2.1 Long-term stability

The long-term stability of the WVIA was evaluated viarepeated measurements of a water standard over a periodof four months. The uncalibrated hourly measurement

time series of a water standard (δD =−78 ‰ andδ18O =−11.3 ‰) at a constant concentration (10 000 ppm) isshown in Fig.4. Each plot represents the last 5 min averagevalue of a 10 min measurement. Measured isotopic valuesrange from−77 ‰ to −82 ‰ for δD and from−8.0 ‰ to−12.5 ‰ forδ18O. The spread inδ18O is three times largerthan the difference between the maximum and minimumof WVISS-generated air at the same concentration (Fig.2),indicating the variation stems mainly from the WVIA.Because an analyzer was installed in an air-conditionedlaboratory, the observed large isotopic variation is too largeto be explained by temperature variations (Sturm and Knohl,2010). It is not clear yet what is causing this largeδ18Ovariation, however it is clear that the system does havesignificant non-linear drift. Thus, frequent measurement ofa water standard is crucial to correct for time-dependentisotope variation.

Next, we examined the similarity between the time-dependent variability in water standards with different iso-topic values. The temporal variations inδD andδ18O of thesix water standards (see Table 1) are shown in Fig.5. Thedata are represented as a normalized anomaly from the long-term mean value. The temporal trends of each standard wa-ter matched each other well for both isotopes, and they donot depend on the isotopic content of water. Thus, repeatedmeasurements of a single reference water sample should beenough to remove time-dependent drift. To move the run-to-run variability, the WVIA-reported isotopic value of thewater standard was converted to the VSMOW scale using awater standard (STD-11, see Table 1).

δWVIA =δsample/raw− 1

δstd11/raw− 1(δstd11/VSMOW− 1) − 1

where δsample/raw and δSTD11/raw are the rawδD or δ18Ovalues of the sample and STD-11, respectively, andδSTD11/VSMOW is the VSMOW scale value determined byIRMS measurements.

Atmos. Meas. Tech., 5, 2069–2080, 2012 www.atmos-meas-tech.net/5/2069/2012/

N. Kurita et al.: Evaluation of water vapor isotope analyzer 2075

Fig. 4. Time series of uncalibratedδD andδ18O values of a single water standard (δD =−78 ‰ andδ18O =−11.3 ‰) measured using theWVIA from late August through mid-December 2011. The data represent the uncorrected and uncalibratedδD andδ18O values recorded bythe WVIA and show the amount of instrument drift over the multi-month period.

Fig. 5.Comparison of time-dependentδD andδ18O variations of sixwater standards with different isotopic compositions. Plotted valuesare expressed as a normalized anomaly, that is, a departure from thelong-term mean for the analysis period, divided from the standarddeviation during this period.

3.2.2 Linearity

To test the linearity of the WVIA-WVISS coupled sys-tem, the WVIA-measured isotopic values of several waterstandards were compared with those obtained via IRMSmeasurements. The differences between values from the

WVISS/WVIA and known values of each standard are shownin Fig. 6. For δD, except STDX3 (δD =−332.7 ‰), the cal-culated VSMOW scale value of each water standard agreedwell with the known value. However, the WVISS/WVIA-δ18O values significantly underestimated the true values, andthe amplitude of this underestimation decreased linearly withincreasingδ values. A two-point calibration method was ap-plied to normalize WVIA-reported data (δWVIA ) to the inter-national VSMOW/SLAP scale. The calibration line was de-termined from two different water standards, one of whichhas more negative isotope contents (STD10) and anotherwith more enriched isotopic values (STD11). The averagedlines obtained during the laboratory test are as follows:

δDIRMS = (1.001± 0.003) δDWVIA − 0.85 ± 1.03 (1)

δOIRMS = (0.946± 0.005) δOWVIA − 0.96 ± 0.89. (2)

The linear relationship between IRMS and WVIA measure-ments has already been reported inSturm and Knohl(2010),however the slope for both isotopes are substantially differentin this study. This finding suggests that linearity calibrationswill likely be instrument specific. During the period of thislaboratory test, the calibration line was determined every dayand then normalized to the VSMOW/SLAP scale, becauseeach obtained calibration line significantly differed from theaverage values shown in Eqs. (1) and (2). The results of thewater standards measurements are summarized in Table 1(Fig. 6 for δ18O). The isotopic difference between knownand measured values is less than 1.5 ‰ forδD and 0.2 ‰ forδ18O. Thus, the corrected WVIA values agree well withknown values within analytical error. In summary, obtainingaccurate isotope data from WVISS/WVIA systems will re-quire, at a minimum, correcting for time-dependent isotopic

www.atmos-meas-tech.net/5/2069/2012/ Atmos. Meas. Tech., 5, 2069–2080, 2012

2076 N. Kurita et al.: Evaluation of water vapor isotope analyzer

Fig. 6. The difference in WVISS basedδD andδ18O values from known IRMS based values of each water standard (blue circle). In bothfigures, the x-axis represents known IRMS based SMOW/SLAP scale values. Forδ18O, collected SMOW scale values are also plotted at thetop of the figure (white circle).

drift using a frequently repeated reference standard, and nor-malization of the drift corrected data to the VSMOW/SLAPscale using a two standard approach. The recommended cal-ibration interval to determine the VSMOW/SLAP scale isdaily.

3.2.3 Concentration dependence

Previous studies have reported water vapor concentrationdependence ofδ values, and highlighted this factor as theprimary source of analytical uncertainty (Sturm and Knohl,2010; Rambo et al., 2011). Through the WVISS perfor-mance test a clear concentration dependence forδ18O aris-ing from the calibration system was uncovered, howeverthis effect has not been fully considered in previous studies.The water vapor concentration effect, determined from rawWVIA isotopic values through the measurement of WVISS-generated air, together with the isotopic changes in WVISS-generated air collected by cold trap are shown in Fig.7.The responses to changes in water vapor concentration areshown as differences in isotope values from those mea-sured at 10 000 ppm. Regardingδ18O there is a increasingtrend in δ18O values with decreasing water vapor concen-tration for the WVISS/WVIA system that is a nearly lin-ear match to that of WVISS-generated air up to a water va-por concentration of 6500 ppm. However, in the lower rangeof the concentration (lower than 6500 ppm), theδ18O val-ues of WVISS-generated air decreased inversely with wa-ter vapor concentration. These findings suggest that con-centration dependence determined by WVIA-WVISS mea-surements is underestimated for low water vapor concentra-tion. The corrected concentration effect arising from WVIAmeasurements was determined through subtraction of iso-topic changes in WVISS-generated air. These results showthat there is no effect in the concentration range of 6500 to10 000 ppm and it gradually increases in the concentration

Fig. 7. δ18O anomaly with changing water vapor concentra-tions. Anomalies are calculated by subtracting the mean value at10 000 ppm from measured values at a given water vapor concen-tration. Error bars show the standard deviation of repeated experi-ments. The WVISS was run with different water vapor concentra-tions and the resultant vapor was measured with the WVIA (redline). To escape from the influence of WVIA temporal drift, thisexperiment was completed within a 30 min time frame after start-ing calibration. Moisture from WVISS exhaust air was collectedusing the cold trap system and analyzed off line (blue). To col-lect a sufficient amount of water for isotope analysis, vapor trap-ping continued more than two hours. A large spread in the errorbar at each vapor concentration may result from isotopic changes inWVISS-generated air and analytical error. Humidity bias stemmingsolely from the WVIA (green line) was calculated by subtracting themean values of water vapor generated by the WVISS from WVIA-measured values at a given water vapor concentration.

range of 6500 to 4500 ppm. Theδ18O values at 3000 ppmwere similar to those found at 4500 ppm. This non-linearityeffect forδ18O appears to be similar to that found bySturmand Knohl(2010), although the amplitude of isotopic varia-tions is smaller.

Atmos. Meas. Tech., 5, 2069–2080, 2012 www.atmos-meas-tech.net/5/2069/2012/

N. Kurita et al.: Evaluation of water vapor isotope analyzer 2077

For δD, WVIA-reported raw isotopic values slightly in-crease with decreasing water vapor concentration from10 000 ppm to 3000 ppm (not shown); this tendency is con-sistent with the result of cold trap samples (Fig.2). How-ever, the spread of values from cold trap samples is largerthan that of the WVIA reported values in the whole concen-tration range (Fig.3) and thus we cannot conclude that thiswater vapor concentration dependency is related to WVISSvapor production. Because the amplitude of this shift is small(1 ‰ from 10 000 ppm to 3000 ppm), we do not considerthe concentration dependence in this study. The amplitudeis quite small compared to that reported bySturm and Knohl(2010). The concentration effect results likely from the spec-tral fitting procedures used, including the removal of inter-ferences in the WVIA. The spectral fitting procedure of ourinstrument was updated in May 2011 and the upgraded in-strument may be less sensitive to the water vapor concentra-tion effect.

3.3 Ambient air analysis

3.3.1 Calibration procedure

The calibration procedure developed from laboratory exper-iments was applied to the continuous measurement of watervapor isotopes. First, the WVIA-reported raw isotopic val-ues were converted using the reference standard values tocorrect time-dependent isotopic drift. In the laboratory test,the reference standard was measured just once a day, how-ever more frequent measurement of the reference is recom-mended for continuous measurement under routine condi-tions (see Fig.4). For the measurements of outdoor air, thereference gas (which is set to a water vapor concentrationof 10 000 ppm) was measured every hour for 10 min. Theisotopic contents of the source water used as a referencewere determined by IRMS measurement (δD =−79 ‰ andδ18O =−11.3 ‰). The hourly reference data was interpo-lated to 10 min steps and the data at measuring time wasused for reference scale conversion. Forδ18O, because hu-midity bias cannot be ignored, the observed non-linear func-tion was applied to time-interpolated reference data and thecorresponding reference valueδref at the concentration mea-sured in ambient air was calculated as

δDref =(D/H)raw

(D/H)ref− 1

δOref =

(18O

/16O

)raw

f (H2O)(18O

/16O

)ref

− 1

where D/Hraw (18O/16Oraw) and D/Href (18O/16Oref) are theWVIA-reported isotopic ratios of ambient air and water va-por reference generated by the WVISS. Thef (H2O) rep-resents a calibration function for the concentration effectshown in Fig.7 (green line). As described in the previous

section, we applied this function when the water vapor con-centration of ambient air dropped lower than 6500 ppm, oth-erwisef (H2O) = 1.

Next, the reference scale value was converted to theVSMOW scale value and then normalized to the VS-MOW/SLAP scale value using Eqs. (1) and (2) in Sect. 3.2.2.The measurement of multiple water standards was not per-formed during the measurement of ambient air, thus the long-term mean slope and intercept forδD and δ18O value ob-tained from the laboratory test was used. The possible er-ror in isotopic measurements when applying a constant slopeand intercept was evaluated through a comparison with theresults shown in Table 1. Without daily calibration, the stan-dard deviation of water standards increased a little, rangingfrom 0.79 to 1.32 ‰ forδD and from 0.19 to 0.39 ‰ forδ18O. However, this error is relatively small compared to thatarising from the uncertainty of the humidity bias correction.

3.3.2 Field data evaluation

To further test our calibration procedure, corrected WVIAdata was compared with the results of cryogenic vapor col-lection/IRMS measurement. The continuous measurement ofatmospheric vapor based on a laser instrument and ambi-ent air sampling using the cold trap method was carried outfrom August to December 2011. During this period, morethan 150 water vapor samples were collected using the coldtrap method with water vapor concentrations ranging from3000 to 15 000 ppm.

Even though several sources of uncertainty contribute tothe degraded accuracy of water vapor isotope measurementsusing the WVIA, the primary source of uncertainty may beassociated with non-linear dependency on the humidity biascorrection. Because this function includes the uncertaintiesarising from both the WVIA and the WVISS (Fig.7), therobustness of this function should be tested before applyingit to field data. In Fig.8a, the non-linear function was com-pared with the humidity bias obtained from ambient air mea-surement. For this comparison, we used WVIA data with-out correcting the water vapor concentration effect; thesewere integrated over time during vapor trapping. Althoughthe data points are widely scattered in the low concentrationrange (σ > 0.35 ‰ at 4500 ppm, whereasσ < 0.30 ‰ be-tween 6500 and 15 000 ppm), a clear negative effect of wa-ter vapor concentration onδ18O contents is seen in the fielddata. In addition, this humidity bias is in very good agree-ment with the calibration curve obtained from the laboratoryexperiment. Thus, applying the calibration function leads toa decrease in systematic offset depending on concentration(Fig. 8b). This finding emphasizes that the non-linear func-tion reasonably represents concentration dependence arisingfrom the WVIA. For δD, the concentration effect was notclearly seen in the field data, although the points are morescattered towards the region of decreasing water vapor con-centration (Fig.9).

www.atmos-meas-tech.net/5/2069/2012/ Atmos. Meas. Tech., 5, 2069–2080, 2012

2078 N. Kurita et al.: Evaluation of water vapor isotope analyzer

Fig. 8. The δ18O difference in atmospheric water vapor betweenWVIA data and cold trap vapor samples as a function of changingwater vapor concentrations.(a) WVIA data without a water vaporconcentration correction and(b) corrected data. WVIA data wasintegrated during vapor trapping. The broken red line with crossesshows average values at each vapor concentration from 3500 ppmto 14 500 ppm. For comparison, the water concentration bias arisingfrom WVIA and shown as a green line in Fig.7 was superimposedin the uncorrected version.

Over the observation period, theδ18O (δD) values ob-served from samples collected by the cold trap method var-ied widely from−27.1 ‰ (−199 ‰) to−13.1 ‰ (−94 ‰),and these large temporal changes were also successfully re-flected by the WVIA values which in fact reveal even greatertemporal dynamics than were possible to measure using thetrapping approach (not shown). In addition, the time-seriesof the d-excess parameter (d = δD − 8× δ18O), defined byDansgaard(1964) show reasonable consistency between thecold trap and WVIA analysis over a large range (5∼ 25 ‰).The similarity between trap based values and WVIA basedvalues is strong evidence that high quality high temporal res-olution data can be obtained over a wide range of isotopevalues and humidity conditions with the WVISS/WVIA sys-tem. To further evaluate the accuracy and precision ofδDandδ18O measured with the WVIA, we calculated the time-averaged isotopic values during vapor trapping and then sub-tracted from the cold trap values for the same sampling peri-ods (Fig.10). The plots are scattered randomly in the rangeof −0.65 to 0.77 ‰ forδ18O and from−3.8 to 2.9 ‰ forδD,

−4

−3

−2

−1

0

1

2

3

4

δDla

ser−δD

trap

3000 6000 9000 12000 15000

H2O ppmv

Fig. 9.The same as Fig.8b, but forδD values.

and there is no offset. The mean value of the deviations of the154 samples is 0.06± 0.28 ‰ forδ18O and−0.3± 1.4 ‰ forδD, respectively. Thus, although there is substantial scatterthe mean values for both isotopes are close to zero suggest-ing that our developed calibration procedure can provide ac-curate isotope data, similar to the conventional method. Therelatively large random error may result from the severalsources of uncertainty. As discussed earlier, there are uncer-tainties related to how WVIA/WVISS data were correctedand calibrated. More frequent linearity calibration may helpto reducing the random error. Regardingδ18O the uncertaintyof the isotopic value of WVISS-generated air is 0.25 ‰ andmay contribute to random error. Further development of thecalibration system may play a key role in reducing this ran-dom error. Finally, we evaluated thed-excess values by com-paring thed-excess measured by the WVIA with that fromthe cold trap method (Fig.11). Most data plot along the 1 : 1line which indicates that our calibration procedure can pro-vide d-excess data that are consistent with the conventionaltrap approach.

4 Summary and conclusions

Water vapor stable isotope measurements based on a com-mercially available laser instrument (WVIA) in conjunctionwith a calibration system (WVISS) was evaluated throughlaboratory experiments using several water standards. Wefound that WVIA instrument can provide accurate results ifcorrections are applied for: (1) temporal drift, (2) normal-ization to the VSMOW/SLAP scale, and (3) dependency ofisotope results on water vapor concentration. Periodic VS-MOW/SLAP normalization using two standards (lineariza-tion calibration) is especially important forδ18O (and d-excess), because the slope of the calibration line between theWVIA-reported values and the IRMS value differs substan-tially from 1 in addition to drift correction for temporal vari-ation using a frequently measured single reference standardas described byRambo et al.(2011). Another remarkable

Atmos. Meas. Tech., 5, 2069–2080, 2012 www.atmos-meas-tech.net/5/2069/2012/

N. Kurita et al.: Evaluation of water vapor isotope analyzer 2079

−4

−2

0

2

4

−0.8 −0.6 −0.4 −0.2 0.0 0.2 0.4 0.6 0.8

Fig. 10. Comparison ofδD andδ18O of water vapor collected viathe cold trap method and directly measured using the WVIA. Forimproved comparison, the time-averaged isotopic values of WVIAmeasurements during vapor trapping periods are plotted.

finding of our study is that concentration dependence stemsfrom the WVISS. Theδ18O values of vapor generated by theWVISS vary with time and depending on the output humidityconcentration when vapor concentration becomes lower than6500 ppm. In this study, we determined the “true” concen-tration effect stemming from the WVIA and then produced anew calibration procedure for WVIA measurement.

To test our calibration procedure and the instrumentation,continuous measurement of water vapor isotopes in ambi-ent outdoor air was carried out from summer to winter usingthe WVISS/WVIA system together with water vapor sam-pling using the conventional cold trap method. Despite watervapor concentrations that varied from 3000 to 15 000 ppm,the results from both methods agree well throughout the en-tire observation period. The precision estimated from thecomparison of both measurements is 0.28 ‰ forδ18O and1.4 ‰ for δD, respectively. These uncertainties allow calcu-lation of representatived-excess values and to assess naturalvariability. Thed-excess variations from the WVIA measure-ments reasonably match those obtained from cold trapped va-por. The fact that accurated-excess values can be obtained isgood confirnamtion of the overall performance of the instru-ment and calibration approach.

In summary, accurate water vapor isotope data, with re-sults comparable to those achieved using conventional meth-ods, can be obtained from commercially available systemswith high time resolution. This suggests that isotopes in wa-ter vapor can be measured without special operational skills.In addition, laser based instruments do not require the manualeffort and hassles. Thus, the laser based technique has sub-stantial potential to increase the availability of water vapor

Fig. 11.Comparison ofd-excess in water vapor using the cold trapand WVIA methods. The WVIA-measuredd-excess values werederived from corrected and calibratedδD andδ18O values and thentime-averaged values were calculated.

isotope data which can significantly advance scientific un-derstanding related to the water cycle.

Acknowledgements.We would like to thank Liliana Poltensteinand Michael van Duren for liquid water isotope analysis in ourlaboratory, Manfred Jaklitsh for his help in the laboratory testand ambient air sampling, and Maureen Macneill for editing theoriginal text.

Edited by: D. Griffith

References

Aemisegger, F., Sturm, P., Graf, P., Sodemann, H., Pfahl, S., Knohl,A., and Wernli, H.: Measuring variations ofδ18O andδ2H in at-mospheric water vapour using two commercial laser-based spec-trometers: an instrument characterisation study, Atmos. Meas.Tech., 5, 1491–1511,doi:10.5194/amt-5-1491-2012, 2012.

Araguas-Araguas, L., Froehlich, F., and Rozanski, K.: Deuteriumand oxygen-18 isotope composition of precipitation and atmo-spheric moisture, Hydrol. Process., 14, 1341–1355, 2000.

Bear, D., Paul, J., Gupta, M., and O’Keefe, A.: Sensitive absorp-tion measurements in the near-infrared region using off-axisintegrated-cavity-output-spectroscopy, Appl. Phys. B-Lasers O.,75, 261–265, 2002.

Bowen, G. J., Wassenaar, L. I., and Hobson, K. A.: Global applica-tion of stable hydrogen and oxygen isotopes to wildlife forensics,Oecologia, 143, 337–348, 2005.

Coplen, T. B.: New reporting of stable hydrogen, carbon, and oxy-gen isotopic abundances, Geochim. Cosmochim. Acta, 60, 3359–3360, 1996.

Dansgaard, W.: Stable isotopes in precipitation, Tellus, 16, 436–468, 1964.

www.atmos-meas-tech.net/5/2069/2012/ Atmos. Meas. Tech., 5, 2069–2080, 2012

2080 N. Kurita et al.: Evaluation of water vapor isotope analyzer

Gkinis, V., Popp, T. J., Johnsen, S. J., and Blunier, T.: A contin-uous stream flash evaporator for the calibration of an IR cavityring-down spectrometer for the isotopic analysis of water, Isotop.Environ. Health S., 46, 463–475, 2010.

Griffith, D. W. T., Jamie, I., Esler, M., Wilson, S. R., Parkes, S. D.,Waring, C., and Bryant, G. W.: Real-time field measurements ofstable isotopes in water and CO2 by Fourier transform infraredspectroscopy, Isotop. Environ. Healt. S., 42, 9–20, 2006.

Gupta, P., Noone, D., Galewsky, J., Sweeney, C., and Vaughn, B. H.:Demonstration of high-precision continuous measurements ofwater vapor isotopologues in laboratory and remote field deploy-ments using wavelength-scanned cavity ring-down spectroscopy(WS-CRDS) technology, Rapid Commun. Mass Sp., 23, 2534–2542, 2009.

Han, L.-F., Groning, M., Aggarwal, P., and Helliker, B. R.: Reli-able determination of oxygen and hydrogen isotope ratios in at-mospheric water vapor absorbed on 3A molecular sieve, RapidCommun. Mass Sp., 20, 3612–3618, 2006.

He, H. and Smith, R. B.: Stable isotope composition of water vapourin the atmospheric boundary layer above the forests of New Eng-land, J. Geophys. Res., 104, 11657–11673, 1999.

Helliker, B. R., Roden, J. S., Cook, C., and Ehleringer, J. R.: Arapid and precise method for sampling and determining the oxy-gen isotope ratio of atmospheric water vapor, Rapid Commun.Mass Sp., 16, 929–932, 2002.

Hoffmann, G., Jouzel, J., and Masson, V.: Stable water isotopesin atmospheric general circulation models, Hydol. Process., 14,1385–1406, 2000.

Iannone, R. Q., Romanini, D., Kassi, S., Maijer, H., and Kerstel, E.R. T.: A microdrop generator for the calibration of a water vaporisotope ratio spectrometer, J. Atmos. Ocean. Tech., 26, 1275–1288, 2009.

Kerstel, E. R. T., Iannone, R. Q., Chenevier, M., Kassi, S., Jost,H. J., and Romanini, D.: A water isotope (2H, 17O, and18O)spectrometer based on optical-feedback cavity enhanced absorp-tion for in-situ airbone applications, Appl. Phys. B-Lasers O., 85,397–406, 2006.

Kurita, N.: Origin of Arctic water vapor during the ice-growing season, Geophys. Res. Lett., 38, L02709,doi:10.1029/2010GL046064, 2011.

Kurita, N. and Yamada, H.: The role of local moisture recyclingevaluated using stable isotope data from over the middle of theTibetan Plateau during the monsoon season, J. Hydrometeorol.,9, 760–775, 2007.

Kurita, N., Noone, D., Risi, C., Schmidt, G. A., Yamada, H., andYoneyama, K.: Intraseasonal isotopic variation associated withthe Madden-Julian Oscillation, J. Geophys. Res., 116, D24101,doi:10.1029/2010JD015209, 2011.

Lee, X., Sargent, S., Smith, R., and Tanner, B.: In situ measurementof the water vapor18O/16O isotope ratio for atmospheric andecological applications, J. Atmos. Ocean. Tech., 22, 555–565,2005.

Lee, X., Smith, R., and Williams, J.: Water vapour18O/16O isotoperatio in surface air in New England, USA, Tellus B, 58, 293–304,2006.

Peters, L. I. and Yakir, D.: A rapid method for the sampling of at-mospheric water vapour for isotopic analysis, Rapid Commun.Mass Sp., 24, 103–108, 2010.

Rambo, J., Lai, C.-T., Farlin, J., Schroeder, M., and Bible, K.: On-site calibration for high precision measurements of water vaporisotope ratios using off-axis cavity-enhanced absorption spec-troscopy, J. Atmos. Ocean. Tech., 28, 1448–1457, 2011.

Schmidt, M., Maseyk, K., Lett, C., Biron, P., Richard, P., Bariac,T., and Seibt, U.: Concentration effects on laser-basedδ18O andδ2H measurements and implications for the calibration of vapourmeasurements with liquid standards, Rapid Commun. Mass Sp.,24, 3553–3561, 2010.

Schoch-Fischer, H., Rozanski, K., Jacob, H., Sonntag, C., Jouzel,J.,Ostlund, G., and Geyh, M.: Hydrometeorological factors con-trolling the time variation of D,18O and3H in atmospheric watervapour and precipitation in the northern westwind belt, in: Iso-tope Hydrology 1983, IAEA-publication, Vienna, Austria, 3–30,1984.

Sturm, P. and Knohl, A.: Water vaporδ2H and δ18O measure-ments using off-axis integrated cavity output spectroscopy, At-mos. Meas. Tech., 3, 67–77,doi:10.5194/amt-3-67-2010, 2010.

Tremoy, G., Vimeux, F., Cattani, O., Mayaki, S., Souley, I., andFavreau, G.: Measurements of water vapor isotope ratios withwavelength-scanned cavity ring-down spectroscopy technology:new insights and important caveats for deuterium excess mea-surements in tropical areas in comparison with isotope-ratio massspectrometry, Rapid Commun. Mass Sp., 25, 3469–3480, 2011.

Uemura, R., Matsui, Y., Yoshimura, K., Motoyama, H., andYoshida, N.: Evidence of deuterium excess om water vapor asa indicator of ocean surface condition, J. Geophys. Res., 113,9114,doi:10.1029/2008JD010209, 2008.

Wang, L., Caylor, K. K., and Dragoni, D.: On the calibration of con-tinuous, high-precisionδ18O andδ2H measurements using anoff-axis integrated cavity output spectrometer, Rapid Commun.Mass Sp., 23, 530–536, 2009.

Yakir, D. and Wang, X. F.: Fluxes of CO2 and water between terres-trial vegetation and the atmosphere estimated from isotope mea-surements, Nature, 380, 515–517, 1996.

Atmos. Meas. Tech., 5, 2069–2080, 2012 www.atmos-meas-tech.net/5/2069/2012/