Isotopic Fractionation in Wintertime Orographic Clouds
1. Introduction
Precipitation amounts can increase when wintertime storms encounter mountains or plateaus. Orographic lifting in cold mountain clouds results in the production of additional condensate in the form of supercooled liquid water (SCLW) that becomes available for accretion (riming) by snow falling from higher levels, as was the case in the Park Range in northwestern Colorado (Rauber and Grant 1986a) and in the Mikuni Mountains of Japan (Kusunoki et al. 2005). The extent and liquid water content (LWC) of near-surface SCLW depend on the mountain aspect, wind speed, and direction relative to the mountain, and proximity to the moisture source. The intensity and location of the liquid cloud depend on stability, the occurrence of blocking, and small-scale topographical features (e.g., Colle et al. 2005; Houze and Medina 2005; Rotunno and Houze 2007). Liquid water contents as high as 1.5 and 2 g m−3 have been observed in wintertime orographic clouds in the Sierra Nevada and Cascade Mountains, respectively (Lamb et al. 1976; Hobbs 1975). In contrast, significantly lower SCLW levels (<0.4 g m−3) were reported in wintertime orographic clouds in the Park Range (Rauber and Grant 1986a; Borys et al. 2000, 2003).
The effect of aerosols on orographic cloud microphysics and snowfall has been examined by Borys et al. (2000, 2003), Muhlbauer et al. (2010), Lowenthal et al. (2011), and Saleeby et al. (2013). Model simulations by Saleeby et al. (2013) confirmed the empirical studies by Borys et al. (2000, 2003) that suggested that higher cloud condensation nucleus (CCN) concentration increased cloud droplet number concentration, decreased their size, thereby reducing riming efficiency and snowfall amount. Muhlbauer et al. (2010) applied three cloud-resolving models to orographic precipitation in an idealized 2D context to examine the aerosol effect on precipitation from mixed-phase orographic clouds. In most of the cases, the models showed a decrease in orographic precipitation with increasing aerosol number concentrations, but the sensitivity of precipitation to aerosol perturbations varied considerably among the models and scaled with the ice water path, suggesting that the presence of a well-developed mixed-phase layer may reduce the aerosol susceptibility of orographic precipitation due to compensating effects and microphysical buffers introduced by numerous ice-microphysical pathways.
Fractionation of stable isotopologues of water (H218O, H216O, HD16O [D, deuterium is 2H]) in the atmosphere occurs during phase changes between water vapor and liquid water or ice.
The heavier isotopologues (HD16O and H218O) are more plentiful in the condensed phase during the exchange between vapor and cloud drops or snow because their vapor pressures are lower than that of H216O. Further, kinetic fractionation occurs during nonequilibrium exchange (e.g., deposition onto ice in supersaturated conditions) due to the differing diffusivities of heavier and lighter water molecules (Jouzel and Merlivat 1984). The isotopic composition of stream and tree sap waters was used to explore the drying of air columns as they passed across mountain barriers (Smith et al. 2005; Smith and Evans 2007). Drier air masses require colder temperatures for condensation, and the resulting connection between temperature and isotopic composition has been found in observations of precipitation in the North Atlantic (Dansgaard 1964) and Antarctica (Picciotto et al.1960; Jouzel and Merlivat 1984; Masson-Delmotte et al. 2008). Warburton and DeFelice (1986) estimated the altitude of ice production in wintertime Sierra Nevada storms from the isotopic composition of snow collected at the surface, the empirical relationship between ice crystal habit and formation temperature, and atmospheric soundings. Demoz et al. (1991) extended the work of Warburton and DeFelice (1986) by incorporating information on riming derived from visual observation of ice crystals at the surface. They found that ice formation temperatures based on ice crystal habit were consistently lower than those inferred from the isotopic composition of snow collected at the surface. The difference was attributed to riming of crystals at altitudes lower and warmer than those at which the crystals nucleated. This framework for understanding the isotopic content of precipitation has served as a basis for inferring past climate characteristics from the isotopic composition of polar snow and ice (Lorius et al. 1979; Steen-Larsen et al. 2011), the inference of past mountain range heights relative to sea level (Rowley et al. 2001; Poage and Chamberlain 2001), and the contribution of orographic cloud water to precipitation in mountain forest ecosystems (Scholl et al. 2007).
Representations of isotopic exchange have been included in general circulation models (GCMs), starting with the work of Joussaume et al. (1984), and have been applied to study many issues including, for example, troposphere–stratosphere exchange (Schmidt et al. 2005) and the relationships between the parameterization of moist processes and isotopic composition (Lee et al. 2009a; Wright et al. 2009). Isotope-enabled GCMs and regional models have been widely applied in an effort to reproduce current and past climates (e.g., Noone and Simmonds 2002; Ciais and Jouzel 1994; Sturm et al. 2007; Risi et al. 2010; Werner et al. 2011; Yoshimura et al. 2011). They have also been used to gain understanding of specific paleoisotopic records, as in the recent work of Lee et al. (2009b) and Pausata et al. (2011). Progress has been slower for incorporating water isotopologues in cloud-resolving/large-eddy simulation models, which resolve at least the largest scales of motions within clouds. Such models remove many of the uncertainties associated with moist convection parameterizations in GCMs but are typically constrained to be run in limited geographical domains due to their higher spatial resolution. Building on earlier work by Gedzelman and Arnold (1994) and others, isotope-enabled cloud-resolving models have been developed by Smith et al. (2006) and Blossey et al. (2010) to study isotopic fractionation in the tropical tropopause layer. Recently, Pfahl et al. (2012) used Consortium for Small-Scale Modelling Isotope model (COSMOiso) to study the isotopic composition of a frontal system over the eastern United States and validated that model against observations included in the work of Gedzelman and Lawrence (1990). The simulations of Pfahl et al. (2012) were not cloud-resolving and required the integration of water isotopologues into the convective parameterization.
The Isotopic Fractionation in Snow (IFRACS) study at Storm Peak Laboratory (SPL) in northwestern Colorado was conducted during the winter of 2014 to explore the impacts of microphysical processes in mixed-phase orographic clouds on the water isotopic composition of falling snow. In-cloud observations are being complemented by simulations with an isotope-enabled version of the Weather Research and Forecasting (WRF) Model (Blossey et al. 2015; Moore et al. 2016). In the model, the evolution of water isotopic composition is followed from the water vapor source(s) through snowfall at SPL. Measurement of the isotopic composition of water vapor, cloud droplets, and snow in mixed-phase clouds coupled with simulations that follow these isotopic tracers through phase changes is intended to improve our understanding of snow growth processes, including riming and vapor deposition in microphysical parameterizations. The contribution of isotope-enabled microphysical modeling will allow for broader application of these methodologies in studies of general circulation and the water cycle. In this paper, we present and discuss the first simultaneous measurements of water isotopologues in water vapor, cloud droplets, and snow in mixed-phase wintertime orographic clouds.
2. Methods
A field study was conducted at the Desert Research Institute’s Storm Peak Laboratory (3210 m MSL; 40.456570°N, 106.739948°W) located on the summit of Mt. Werner in the Park Range near Steamboat Springs, Colorado, from 20 January to 27 February 2014. The study location is shown in Fig. 1 and in Wetzel et al. (2004). SPL is above cloud base in snowing clouds greater than 25% of the time during winter (Borys and Wetzel 1997). Synoptic-scale storms occur roughly weekly with snowfall under pre- and postfrontal conditions, with and without convection, and in large-scale stratiform cloud systems (Rauber et al. 1986b). Given sufficient moisture, orographic forcing typically produces a cap cloud at SPL, which may be embedded, and which produces persistent snowfall. During precipitation events, the flow is generally from the west or northwest. Cloud and precipitation under southwest flow are suppressed by the Flat Tops Range (maximum elevation: 3768 m MSL) (Fig. 1). The main source of moisture is the Pacific Ocean to the west. Snowing clouds were sampled during intensive operating periods (IOPs) with discrete durations averaging 37 min using methods described by Borys et al. (2000). Hydrometeor particle size distributions (PSDs) were measured at a frequency of 1 Hz with Droplet Measurement Technologies Inc. (DMT) cloud probes mounted on a rotating wind vane to orient them into the wind. The probes were calibrated at the factory prior to but not immediately after the field study. Cloud droplet PSDs (2–47 μm) were measured with an aspirated DMT SPP-100 (forward scattering spectrometer probe). Ice particle PSDs were measured with DMT cloud imaging probe (CIP; 25–1550 μm) and precipitation imaging probe (PIP; 100–6200 μm) optical array probes (OAP). An Applied Technologies Inc. SATI three-axis sonic anemometer mounted on the probe vane supplied the airspeed for the array probes. The 2D probe images were analyzed using the System for OAP Data Analysis (SODA2) analysis software obtained from the National Center for Atmospheric Research Mesoscale and Microscale Meteorology Laboratory (NCAR MMM), which is based on work by Heymsfield et al. (2013) and Delanoe et al. (2014).
Topographical profile of the study area and the location of Storm Peak Laboratory (red star) and the town of Steamboat Springs (black circle) in the Park Range in northwestern Colorado.
Citation: Journal of Atmospheric and Oceanic Technology 33, 12; 10.1175/JTECH-D-15-0233.1
Super-cooled cloud water was collected with monofilament sieves that collect droplets with diameters > 2 μm but not snow crystals (Hindman et al. 1992). If necessary, sequential sieve samples were taken during a snow collection period to avoid overloading the cloud sieves (Hindman et al. 1992). Immediately after sampling, rime ice was scraped off the sieves, placed in a storage bag, and transferred to a freezer. Snow was collected in polyethylene bags mounted in two 15-cm-diameter tubes oriented into the wind by a vane (Borys et al. 1988). This collector acts as a virtual impactor that excludes cloud droplets. Ice water content (IWC) was derived from the wind speed and the diameter of the tubes. Snow samples were transferred to a storage bag and stored in a freezer. After the study, aliquots of cloud and snow water samples were sent to the Institute of Arctic and Alpine Research (INSTAAR) in Boulder, Colorado, for analysis of stable isotopes of water (Lowenthal et al. 2011).
A weather station measured 5-min average temperature, relative humidity (RH), and wind speed and direction. The NCAR GPS Advanced Upper-Air Sounding System (GAUS) balloon-sonde system was deployed in the valley approximately 6.1 km upwind of SPL (2067 m MSL; 40.456250°N, 106.813863°W). Balloons were launched every 2 h during IOPs. Particle light absorption (Bap) was measured with a NOAA/Global Monitoring Division (GMD) continuous light absorption photometer (CLAP) that measures light absorption by particles at three wavelengths (467, 528, 652 nm). The CLAP is similar to the particle soot absorption photometer (PSAP; Bond et al. 1999), but it is designed to run unattended 8 times longer.
a. Water vapor isotopic composition
Two secondary standards were used to calibrate ambient vapor measurements: 1) distilled deionized water (DDI); and 2) filtered Antarctic ice water (POLAR) from the European Project for Ice Coring in Antarctica (EPICA) B40 core taken at a depth of 80 m (~1200 AD). Each secondary standard was divided into ten 125-mL bottles and a small aliquot from each was reserved and sent to INSTAAR for calibration against primary standards [Vienna Standard Mean Ocean Water (VSMOW)], Vienna Standard Light Antarctic Precipitation (VSLAP), and Greenland Ice Sheet Project (GISP)]. The isotopic composition was δD = −68.68 ± 0.20‰ and −342.80 ± 0.46‰ and δ18O = −7.68 ± 0.059‰ and −43.38 ± 0.06‰ for DDI and POLAR, respectively, where the uncertainties are the standard deviations of the 10 replicates [per mil (‰)]. Nearly all of the uncalibrated vapor measurements were within the range of the two standards: 99.2% for δD and 99.97% for δ18O. During routine operation, the SDM injected and vaporized the two secondary standards into the analyzer every 4–6 h for 75% of the time, every 12 h for 14% of the time, and within 24 h for the remainder of the time. The standards were vaporized and mixed with room air dried with Drierite. A critical orifice in the SDM maintained a constant airflow rate. The vapor concentration during standardization was set by regulating the liquid flow rate into the vaporizer. During the experiment, standardizations were done at a vapor concentration of ~20 000 ppmv at a liquid flow rate of 0.04 μL s−1. Each of the standards was measured for 20 min, with 10 min of ambient air sampling between the two. The instrument allows the standard to stabilize for 10 min before reporting data for 20 min. The first 10 min after the second standard were eliminated to allow for stabilization of the ambient vapor measurements. In this manner, memory effects were minimized. The Picarro L-2130-i reports water vapor concentration (H2O), although detailed calibrations beyond those performed by the manufacturer were not done. The average indicated water vapor concentration at which the standards were measured was 21 780 ± 500 ppmv.
The uncertainties of the vapor isotopic measurements have two components: accuracy and precision (repeatability of the measurement of a constant attribute). Accuracy, or bias, is the difference between the true (VSMOW, VSLAP, and GISP) and unknown values, which in this case is the secondary standards (POLAR and DDI). INSTAAR reported upper limits on accuracy of 0.09‰ for δ18O and 0.78‰ for δD. There is no error to minimize for a two-point calibration line. However, the calibration precision is a function of variability of the secondary standards (σ2Std), as determined by INSTAAR, and the variability of the measured values of the standards over the calibration period (σ2Meas). The uncertainties of the slope and intercept of the calibration line can be obtained by propagating σ2Std and σ2Meas through the calculation using the effective variance approach [Eq. (2) in Lowenthal et al. 1987]. The contributions of the calibration process to ambient vapor measurement precision were 0.33‰ and 1.73‰ for δ18O and δD, respectively.
The response and measurement precision of the Picarro L-2130-i vary as a function of water vapor concentration (H2O). Aemisegger et al. (2102) characterized these effects for Picarro analyzers with detailed laboratory experiments. They found that the response of the Picarro L-2130-i analyzer was nearly flat for δ18O and δD for H2O between 2500 and 30 000 ppmv. To address these issues, the POLAR standard was introduced to the analyzer to produce indicated vapor concentrations from ~600 to ~15 000 ppmv by diluting the vaporized standard with ultradry air and by varying the liquid standard flow rate (Bastrikov et al. 2014). This was done at the end of the field study on 5 and 6 March 2014 at SPL after the instrument had been repaired remotely by Picarro. Standard measurements at low vapor concentration could not be reliably done by only reducing the liquid flow rate below 0.01 μL s−1. Low vapor concentrations were obtained by removing the critical orifice in the SDM and manually regulating the flow of ultradry airflow into the vaporizer (Bastrikov et al. 2014). In this manner, we were able to obtain eight and seven data points for δ18O and δD, respectively. The isotopic response with respect to the indicated water vapor concentration is shown in Fig. 2. The response was flat above ~4000 ppmv, but it deviated nonlinearly below that concentration. The solid lines in Fig. 2 represent second- and third-order polynomial fits for δ18O and δD, respectively, for the indicated H2O concentration between 644 and 4203 ppmv. One δD measurement at 1747 ppmv (Fig. 2) was not included in the regression. It is not known why δD was heavier than expected during this run. It is possible that there were problems with the vaporization, resulting in fractionation or that the manual regulation of the ultradry dilution air was not stable. Isotopic composition was corrected for the response bias at H2O less than 4203 ppmv using these functions. Note that there were only 9 hourly average H2O concentrations less than 1000 ppm and that these occurred during the clear period of 20–26 January 2014. During IOPs, there were 13 and 1 hourly average concentrations less than 3000 and 2000 ppmv, respectively.
Isotopic response of the Picarro L-2130-i (a) δ18O and (b) δD to water vapor concentration. Error bars are the standard deviations of the isotopic composition at each H2O concentration. Data points in red were not included in the polynomial regressions.
Citation: Journal of Atmospheric and Oceanic Technology 33, 12; 10.1175/JTECH-D-15-0233.1
Figure 2 shows that measurement precision decreased with decreasing H2O concentration below ~4000 ppmv. Table 1 gives the precision (the standard deviation of measurements of the standard over an approximately 20-min period at constant H2O) as a function of H2O concentration for the points in the H2O calibration. The precision of the POLAR standard measurement was 0.54‰ and 2.16‰ for δ18O and δD, respectively, at 4203 ppmv and 2.78‰ and 15.23‰ for δ18O and δD, respectively, at 644 ppmv. These values could be parameterized or interpolated to estimate measurement precision for the ambient data, but there are very few data points. Measurement precision may also be considered as the variance of ambient measurements over a time period where δ18O and δD are assumed to be constant, that is to say, over repeated measurements. To simulate repeated measurements, ambient vapor δ18O and δD were averaged over 5-min periods and a standard deviation was calculated. On average, there were 50 data points per 5-min period. The squares of these standard deviations were averaged over ranges of H2O concentration. The square root of this average is the root-mean-square error (RMSE). The results are presented in Table 2, which gives the RMSE precision for δ18O and δD, and the number of 5-min averaging periods in each H2O concentration range. Only 1.2% and 0.14% of the measurements had indicated H2O concentrations less than 1000 and greater than 9000 ppmv, respectively. Figure 3 shows second-order polynomial fits of 5-min precision of ambient δ18O and δD measurements as a function of H2O concentration (as the midpoints of the ranges in Table 2). For both of these regressions R2 was greater than 0.999. To compare the two approaches for estimating precision, the data points from Table 1 are overlaid with the values from Table 2 in Fig. 3. The two approaches agree well; however, precision based on short-term averaging of the ambient data is clearly more robust. Measurement precision for each ambient data point was estimated from these functions. The overall precision was obtained by propagating the precisions associated with the standardization and the ambient measurements. These precisions were assigned to the 5-min average data points. Longer averages were calculated from the 5-min data, where the precision was reduced as the square root of the ratio of the number of minutes in the averaging period divided by 5.
Precision (the standard deviation of measurements of the standard over a 20-min period) as a function of H2O concentration (ppmv) for the calibration of instrument response.
Average ambient vapor measurement precisiona as a function of the indicated H2O concentration (ppmv).
Precision (standard deviation) of 5-min ambient measurements of (a) δ18O and (b) δD as a function of water vapor (H2O) concentration (open circles). Precision based on the calibration of isotopic composition vs H2O concentration taken from Table 1 (closed circles).
Citation: Journal of Atmospheric and Oceanic Technology 33, 12; 10.1175/JTECH-D-15-0233.1
b. Back trajectories
Hybrid Single-Particle Lagrangian Integrated Trajectory model (HYSPLIT) back trajectories based on 12-km-resolution North American Mesoscale Forecast System (NAM) data were used to locate the source of water vapor at SPL during IOPs (Draxler and Rolph 2015). The 72-h isentropic back trajectories were run (online) from an elevation of 3500 m ASL every 2 h beginning 2 h before and ending 2 h after each IOP.
3. Results and discussion
a. Isotopic composition
Continuous vapor measurements with regular standardization began on 20 January 2014, although conditions were clear until 27 January 2014. There were 44 cloud sampling periods over 14 days from 27 January to 27 February 2014, representing 27.1 h when SPL was in cloud. Snowfall occurred during 35 of these periods for 21.7 h. Figure 4 presents daily average δ18O and δD compositions of water vapor, cloud water, and snow during the study. While there was considerable temporal variation, both δ18O and δD in the three phases tracked one another. The Pearson correlation coefficients (r) between daily vapor and cloud, vapor and snow, and cloud and snow δ18O were 0.97, 0.94, and 0.93, respectively. The corresponding r values for δD were 0.96, 0.92, and 0.93. Vapor was the most isotopically depleted (more negative δ, i.e., lighter) in the heavy isotopologues (HD16O and H218O), followed by snow, and then cloud. Daily average differences between vapor and cloud δ18O and between vapor and snow δ18O were 12.1 ± 0.8‰ and 10.9 ± 1.3‰, respectively. The corresponding differences for δD were 98.2 ± 7.0‰ and 89.7 ± 12.0‰, respectively. The δ18O and δD were highly correlated during vapor (r = 1.00), cloud (r = 1.00), and snow (r = 0.99) sample periods. Figure 5 presents the relationships between δ18O and δD in cloud, snow, and vapor samples during IOPs. The regressions shown in Fig. 5 are close to the global meteoric water line (GMWL) (δD = 8 δ18O + 10) found by Craig (1961) in natural waters and precipitation, although the intercept for vapor is lower.
Variation of daily average water vapor, cloud, and snow oxygen isotopic composition (δ18O and δD).
Citation: Journal of Atmospheric and Oceanic Technology 33, 12; 10.1175/JTECH-D-15-0233.1
Relationship between δ18O and δD in cloud, snow, and vapor samples. Solid line is the global meteoric water line of Craig (1961): δD = 8 δ18O + 10.
Citation: Journal of Atmospheric and Oceanic Technology 33, 12; 10.1175/JTECH-D-15-0233.1
The temporal variation of vapor δ18O and δD reflects the history of precipitation (i.e., the fraction of vapor removed) and any processes that introduce vapor to the air mass en route to SPL. Daily average vapor δ18O during IOPs varied from −36.0‰ to −23.8‰. Based on equilibrium fractionation coefficients presented in Clark and Fritz (1997) from Majoube (1971), a VSMOW-derived source vapor δ18O at 100% RH would vary by only 1.73‰ for sea surface temperatures between 30°C and 10°C. As vapor moves inland, it becomes more and more isotopically depleted after successive precipitation events (Dansgaard 1964). Since the saturation vapor pressure is lower at lower temperatures, the final vapor δ18O and δD at SPL are determined by the temperature gradient between the vapor source and SPL (Dansgaard 1964; Harmon 1979). This gradient is enhanced by orographic lifting (Siegenthaler and Oeschger 1980). Dansgaard (1964) demonstrated that δ18O in precipitation decreased linearly with decreasing mean annual air temperature with a slope of 0.7‰ °C−1 at 38 coastal stations in the North Atlantic and Greenland, and that this decrease was theoretically consistent with the isotopic lightening of an air mass as it cools during transport from the oceanic vapor source. Isotopic fractionation during successive events of vapor condensation under equilibrium conditions followed by removal of the condensate is known as Rayleigh distillation. Similar relationships were found by Picciotto et al. (1960), Warburton and DeFelice (1986), and Lowenthal et al. (2011).
The expected isotopic lightening following removal of water from air masses was demonstrated in coastal Iceland (Steen-Larsen et al. 2015) and on the eastern slope of the Colorado Rockies (Noone et al. 2013). This was also the case during IFRACS. Figure 6 presents a time series and scatterplot of daily average water vapor concentration (H2O) and δ18O. The relationship was direct with a moderate correlation (r = 0.62), suggesting different airmass histories from day to day. However, Fig. 6 illustrates the effect of Rayleigh distillation upwind of SPL, whether or not there was cloud at SPL. While the isotopic composition of cloud water and snow at SPL varied from day to day, the differences between phases was relatively constant. This is consistent with Rayleigh distillation of the vapor with equilibrium fractionation en route to SPL.
The relationship between daily water vapor concentration (H2O) and vapor δ18O: (a) time series, also indicating days on which cloud and snow sampling (IOPs) occurred; and (b) scatterplot for days without (clear, black circle) and with cloud and snow (IOPs, red circle) sampling (r = Pearson correlation).
Citation: Journal of Atmospheric and Oceanic Technology 33, 12; 10.1175/JTECH-D-15-0233.1
b. Altitude of snow growth
The mass-weighted altitude (MWA) of snow growth by vapor deposition and riming may be inferred from the relationship between isotopic composition and temperature (Picciotto et al. 1960; Warburton and DeFelice 1986; Lowenthal et al. 2011). Based on δ18O in precipitation and surface temperature at 38 coastal North Atlantic and Greenland sites, Dansgaard (1964, p. 443) concluded: “If the mean surface temperature [T] is supposed to vary parallel to the mean condensation temperature from one place to another,” then dδ18O/dT ≈ 0.7‰ °C−1. When snow is more depleted than cloud droplets at SPL (Fig. 4), it follows that it formed aloft at a temperature colder than that at SPL. This temperature can be estimated as the difference (Δ) between δ18O in cloud and snow divided by 0.7. To convert temperature aloft to altitude above SPL, Δ is multiplied by the average of the inverse ambient lapse rate derived from sounding data within 2 km above SPL during IOPs. This value was relatively stable over the experiment (158 ± 20 m °C−1). Figure 7 presents daily MWA, which ranged from just below SPL (−53 m) to 922 m above SPL. On 14 and 16 February, negative heights could represent growth of ice crystals from isotopically enriched vapor or drops in the upslope flow below SPL. Also plotted in Fig. 7 are the daily differences (Δ) between cloud and snow δ18O. The seeder–feeder model (Reinking et al. 2000) suggests that ice nucleation occurs at a higher altitude than the liquid cloud, which, based on the soundings, extended from 1 to 2.5 km above SPL on 10 of 14 sampling days. While ice nucleation may occur above this level, most of the ice mass and its isotopic signature accrue much closer to the surface.
Daily MWA above SPL of snow growth (bars). The solid circles are the differences between δ18O in cloud and snow.
Citation: Journal of Atmospheric and Oceanic Technology 33, 12; 10.1175/JTECH-D-15-0233.1
c. Isotopic fractionation
Under nonequilibrium conditions—that is, evaporation into an unsaturated atmosphere or condensation of vapor to liquid or deposition to ice from a supersaturated atmosphere—a kinetic effect arises from the lower diffusivity of HD16O and H218O relative to H216O (Merlivat 1978; Jouzel and Merlivat 1984). During condensation (deposition), H216O becomes enriched relative to HD16O and H218O in the condensed phase. During evaporation, H216O becomes enriched relative to H218O and HD16O in the vapor phase. Given the much stronger equilibrium fractionation of HD16O compared to H218O, and to a lesser extent the slightly higher diffusivity of HD16O than that of H218O, these kinetic effects, relative to their equilibrium fractionation, are more important for H218O than for HD16O and are manifested in the deuterium excess (d), defined as δD − 8 δ18O (Craig 1961). Thus, evaporation leads to higher d in the vapor phase (lower d in the condensed phase) and condensation (deposition) leads to higher d in the condensed phase (lower d in the vapor phase). Note that while the δD − 8 δ18O definition of d is a good indicator for nonequilibrium conditions at temperatures higher than −20°C (typical of conditions during IFRACS), curvatures in the Rayleigh distillation curve can generate nonzero d even for equilibrium processes.
The isotopic composition of cloud water samples was estimated from the average vapor δ18O and δD during those sample periods and equilibrium fractionation factors from Majoube (1971), and from the combination of equilibrium and kinetic fractionation factors, as described above. Measured and estimated values were compared on a daily average basis. Figure 8 shows good agreement between measured and estimated cloud water δ18O and δD considering equilibrium fractionation only (Fig. 8, Calc.1). Lowenthal et al. (2002) estimated a typical supersaturation with respect to water of 0.5% in wintertime orographic clouds at SPL. The kinetic effect was evaluated assuming Swater = 1.005. Because Swater was so close to unity, the kinetic effect was negligible (Fig. 8, Calc.2), consistent with the work of Stewart (1975). A similar analysis was done for snow growth by vapor deposition only using equilibrium fractionation factors from Ellehoj et al. (2013) at the temperature and vapor composition at SPL. The maximum S with respect to ice was estimated assuming RH with respect to water equal to 100.5% using equations from Vaisala (2013). The results are presented in Fig. 9. Considering equilibrium fractionation only, estimated δ18O and δD were 4.6‰ and 29.5‰ less depleted than measured δ18O and δD, respectively. When kinetic fractionation was included, these differences decreased to 0.97‰ and 16.1‰, respectively.
Measured and estimated (a) δ18O and (b) δD for cloud water using the vapor composition at SPL and equilibrium fractionation only (Calc.1) and equilibrium and kinetic fractionation (Calc.2).
Citation: Journal of Atmospheric and Oceanic Technology 33, 12; 10.1175/JTECH-D-15-0233.1
Measured and estimated (a) δ18O and (b) δD for snow using the vapor composition at SPL and equilibrium fractionation only (Calc.1) and equilibrium and kinetic fractionation (Calc.2).
Citation: Journal of Atmospheric and Oceanic Technology 33, 12; 10.1175/JTECH-D-15-0233.1
This treatment for snow is unrealistic in part because it does not account for snow growth by riming, which can be significant at SPL in mixed-phase clouds (Borys et al. 2003; Lowenthal et al. 2011; Saleeby et al. 2013). Snow will incorporate the isotopic signature of the cloud droplets at SPL to the extent that the snow is rimed by those drops. The daily average depletions of δ18O and δD in snow with respect to cloud were 1.44 ± 1.33‰ and 11.1 ± 10.6‰, respectively. The difference for δ18O was less than 1‰ on 7 of 14 days, suggesting that the rimed mass fraction was high and riming occurred close to SPL. Larger differences could be caused by riming of snow by more depleted droplets or by deposition of vapor that is lighter than at SPL. As discussed above, vapor is expected to be lighter aloft where temperatures are lower than at SPL. Preliminary modeling of a case study with a large difference between cloud and snow δ18O on 13 February 2014 indicated vertical stratification (depletion) of δ18O in water vapor above SPL (Blossey et al. 2015). On 13 February 2014, the difference between snow (−17.3‰) and cloud (−14.0‰) δ18O was 3.3‰, the vapor δ18O was −25.5‰, and the temperature at SPL was −4.7°C. The δ18O estimated assuming equilibrium and equilibrium plus kinetic fractionation were −10.5‰ and −12.7‰, respectively. If vapor deposition occurred at a higher altitude but from the same vapor as at SPL, estimated δ18O would be heavier than at SPL because of the increase in the equilibrium fractionation coefficient with decreasing temperature. Assuming an inverse lapse rate of 158 m °C−1 (above), the temperature at 1 km above SPL (above the highest mass-weighted altitude of snow growth in Fig. 7) would be −11°C. This corresponds to a vapor δ18O of −31.5‰ based on the relationship between δ18O and the temperature described above. Assuming that all vapor deposition occurred at this elevation, estimated snow δ18O would be −15.5‰ and −19.7‰ considering equilibrium fractionation and equilibrium plus kinetic fractionation, respectively. While these estimates bracket the measured snow δ18O, they do not fully account for the effect of riming on the snow’s isotopic composition. While the isotopic composition of cloud water is theoretically consistent with equilibrium fractionation of the vapor at SPL, this is not the case for snow. Model estimates of the vertical structure of the vapor isotopic composition will provide better understanding of the mechanisms and level in the atmosphere where snow forms.
d. Deuterium excess
Deuterium excess d in polar ice has been used to infer source location, meteorological conditions, and long-term climate variations (Jouzel et al. 1982; Dansgaard et al. 1989; Vimeux et al. 1999; Jouzel et al. 2007), water vapor circulation and atmospheric mixing (Blossey et al. 2010), and cloud processes (Bolot et al. 2013; Ciais and Jouzel 1994; Samuels-Crow et al. 2014). Benetti et al. (2014) concluded that d was inversely related to RH above the sea surface and that it increased significantly as a function of increasing surface roughness, which is a function of wind speed. Pfahl and Sodemann (2014) concluded that d varied with RH but not sea surface temperature (SST). Steen-Larsen et al. (2014) found a relationship between d and RH on Bermuda, but the effect of wind speed on d was ambiguous. Steen-Larsen et al. (2015) concluded that neither wind speed nor SST had an effect on d in Iceland. The global meteoric water line of Craig (1961) is consistent with a d of 10‰ in water vapor over an oceanic source at RH~85% (Clark and Fritz 1997).
The temporal variations of daily average vapor, cloud, and snow d during IOPs are shown in Fig. 10a. Average vapor, cloud, and snow d were 11.1 ± 1.7‰, 12.6 ± 2.8‰, and 13.0 ± 3.4‰, respectively. On a daily basis, vapor d at SPL was roughly consistent with oceanic source vapor evaporated at 85% RH (Clark and Fritz 1997). This was also the case for cloud and snow d, although there were higher values prior to 9 February 2014. Figure 10a indicates the direction of the source area with respect to SPL derived from ensemble HYSPLIT back trajectories during IOPs on each day. All 72-h trajectories ended in the Pacific Ocean except for several on 29 February 2014, which ended in Nevada and California. For the 12 days with vapor measurements, the source directions were west, northwest, and west–southwest on 8, 3, and 1 days, respectively. Deuterium excess during IOPs was estimated for cloud (Fig. 10b) and snow (Fig. 10c) from the estimated δ18O and δD, shown in Figs. 8 and 9, respectively. Cloud d after 6 February 2014 was estimated to within 1.3‰ and 0.3‰, on average, based on equilibrium and equilibrium plus kinetic fractionation, respectively. Cloud d was underestimated on 27 January, and 4 and 6 February 2014. As was the case for δ18O and δD (Fig. 9), snow d could not be reproduced assuming growth by vapor deposition under equilibrium or equilibrium plus kinetic fractionation.
(a) Variation of daily d in vapor, cloud, and snow during IOP sample periods; comparison of measured and calculated d in (b) cloud and (c) snow. Calc.1 and Calc.2 are as in Figs. 9 and 10. Wind directions from ensemble HYSPLIT trajectories on each day are shown in Fig. 10a (W stands for west, NW is for northwest, WSW is for west-southwest).
Citation: Journal of Atmospheric and Oceanic Technology 33, 12; 10.1175/JTECH-D-15-0233.1
Welp et al. (2012) reported diurnal cycles of d in surface water vapor during summer at six urban, agricultural, and rural sites in the United States, Canada, and China. Deuterium excess increased from noon to midafternoon at all sites, with amplitudes ranging from 3.5‰ to 17‰. This was attributed to entrainment of upper-level air as the mixed layer deepened and/or evapotranspiration from plants.
The diurnal variation of water vapor d during IFRACS was examined using grand hourly averages for all hours from 20 January to 19 February 2014. The average hourly vapor d was 8.9 ± 3.8‰. The diurnal cycles of d, temperature, water vapor concentration (H2O), and the particle light absorption coefficient at 528 nm (Bap), an indicator of surface pollution, are shown in Fig. 11a. Peaks in Bap occurred from 2300 to 0100 mountain standard time (MST). These peaks were probably caused by snow grooming operations that occurred late at night and into the early morning. After eliminating the highest 1% of the data (i.e., statistical outliers), these peaks disappeared. The uncertainties of d were propagated from those of δ18O and δD. The correlations r between temperature and H2O and between H2O and d were 0.74 and 0.71, respectively. Maxima and minima in d, temperature, H2O, and Bap occurred during the afternoon (1300–1700 MST) and early morning (0400–0900 MST), respectively. Average afternoon and morning d and H2O were 10.3 ± 0.4‰ and 4600 ± 140 ppmv and 7.7 ± 0.4‰ and 4200 ± 140 ppmv, respectively. Thus, the diurnal ranges of d and H2O were only 2.5‰ and 400 ppmv. Note that based on the relationship shown in Fig. 6b, the diurnal variation in water vapor concentration (400 ppmv) corresponds to a change in vapor δ18O of only 0.49‰. Water vapor concentration varies by 40% on a daily basis but only by 11% over the diurnal cycle. Figure 11b shows the diurnal variation of δ18O and δD. There was no clear diurnal trend for δ18O and δD (consistent with the small diurnal variation of water vapor concentration), although their divergence between 1300 and approximately 2000 MST is responsible for the afternoon peak in vapor d. The small variation of water vapor concentration and its isotopic composition at night supports the suggestion in section 3a (Fig. 6b) that their daily variation is related to Rayleigh distillation during transport to SPL.
Diurnal variation of (a) deuterium excess d, temperature, water vapor (H2O), and Bap for the bottom 99% of the data; and (b) diurnal variation of δ18O and δD.
Citation: Journal of Atmospheric and Oceanic Technology 33, 12; 10.1175/JTECH-D-15-0233.1
SPL is above the surface inversion overnight and into the morning hours. Lower temperature and Bap during that period represent background or free-tropospheric air (Borys et al. 1986; Lowenthal et al. 2002; Obrist et al. 2008). During the daytime, surface-derived aerosols (as indicated by Bap) are mixed up to the level of SPL. As discussed above, kinetic fractionation during condensation increased estimated d in the condensate—for example, cloud (Fig. 10b) and snow (Fig. 10c)—although the effect is much greater for snow. The effect is opposite for d in the vapor. Figure 12 shows a weak but direct relationship between daily average H2O concentration and vapor d at SPL. This is expected if the fraction (f) of water vapor remaining after transport to SPL is proportional to the vapor concentration at SPL and if kinetic fractionation occurred during condensation and precipitation upwind of SPL. This mechanism, which is seen on a day-to-day basis, is consistent with the diurnal variation in H2O and vapor d. The diurnal variation in d suggests vertical stratification of water vapor, which becomes drier and more isotopically depleted with a lower d with increasing altitude. Another process that could elevate d during the daytime is evaporation from upwind surface snow. Kinetic fractionation could occur if there was melting followed by evaporation at the snow surface. This is certainly possible as even under cloudy conditions at SPL during IOPs, the average surface temperature and RH at the GAUS launch site were 1.3°C and 73%, respectively. Relatively small day-to-day variation of d in vapor, cloud water, and snow suggests a relatively constant vapor source. However, the diurnal variation of d demonstrates the influence of colder, drier free-tropospheric air with lower d during the nighttime and early morning hours in contrast with higher d in moister surface air during the afternoon hours.
Relationship between daily average water vapor concentration (H2O) and vapor d.
Citation: Journal of Atmospheric and Oceanic Technology 33, 12; 10.1175/JTECH-D-15-0233.1
4. Conclusions
Water isotopologues (HD16O and H218O) were measured in water vapor, supercooled cloud droplets, and falling snow in mixed-phase wintertime orgographic clouds at Storm Peak Laboratory (SPL, 3210 m MSL) from 20 January to 27 February 2014 during the Isotopic Fractionation in Snow (IFRACS) study. A Picarro L-2130-i analyzer continuously measured δ18O and δD in atmospheric water vapor until 19 February. Cloud and snow were sampled when SPL was enveloped in snowing cloud in discrete intervals totaling 21.7 h over 14 days.
a. Isotopic composition
Water vapor was the most isotopically depleted, followed by snow, and then cloud. The δ18O and δD were nearly perfectly correlated (r > 0.99) in all three phases and closely followed the global meteoric water line of Craig (1961). There were strong correlations between daily δ18O in vapor and cloud (r = 0.97), vapor and snow (r = 0.94), and cloud and snow (r = 0.93). The differences between vapor and cloud and between vapor and snow δ18O were similar from day to day, consistent with Rayleigh distillation and equilibrium fractionation during transport to SPL. This also suggests relatively consistent cloud condensation and snow growth processes during the experiment.
b. Altitude of snow growth
The altitude at which the snow accrued most of its mass was inferred from relationships between cloud and snow δ18O, a literature-based dδ18O/dT of 0.7‰ °C−1, and a consistent ambient lapse rate over the course of the field study. The estimated mass-weighted altitude of snow growth was within 1 km of SPL on all days and within 500 m of SPL on 9 out of 14 study days. These results are similar to those found in a previous study at SPL (Lowenthal et al. 2011). While ice crystals may nucleate under subsaturated conditions higher in the atmosphere, they gain most of their ice mass by vapor deposition and riming at much lower levels in the orographic cloud.
c. Isotopic fractionation
The isotopic composition of cloud water at SPL was closely reproduced from the vapor composition at SPL and literature-based equilibrium fractionation factors. Estimated δ18O decreased by only 0.2‰, on average, if the potential effects of kinetic fractionation were considered. Estimating the isotopic composition of snow was more difficult. Assuming depositional growth only, the vapor composition observed at SPL, and equilibrium fractionation, the estimated snow composition was significantly heavier than that observed. Adding kinetic fractionation increased the estimated depletion; however, this treatment is unrealistic in part because snow growth occurs by riming as well as by vapor deposition.
The difference between cloud and snow δ18O was less than 1‰ on 7 of 14 days, suggesting that snow on these days gained much of its mass and isotopic signature from riming close to SPL. Thus, near-surface riming can have a significant effect on the isotopic composition of deposited snow. For larger differences between cloud and snow δ18O, growth likely occurred by deposition of vapor that was isotopically lighter than that observed at SPL or by riming of droplets derived from that vapor. This implies that the vapor and its isotopic composition were vertically stratified. Vapor depletion with increasing altitude is consistent with previous observations and our understanding of isotopic fractionation as a function of temperature.
d. Deuterium excess
The daily average deuterium excess (d) in vapor, cloud, and snow was 11.1 ± 1.7‰, 12.6 ± 2.8‰, and 13.0 ± 3.4‰, respectively. The small variation of vapor d from day to day is consistent with Rayleigh distillation and equilibrium fractionation of water vapor from a similar source during transport of air masses to SPL. Cloud d was estimated to within 2.4 and 1.4‰ assuming equilibrium fractionation and equilibrium plus kinetic fractionation, respectively. In these cases, the vapor d signal was well preserved in the cloud water. The highest cloud d (19.9‰) was estimated to within 3.4‰, assuming both equilibrium and kinetic fractionation. The correlation between measured and estimated cloud d was 0.76. A diurnal trend in vapor d, peaking in the early to midafternoon, coincided with trends in temperature, water vapor concentration, and aerosol light absorption (Bap). This demonstrates a transition between surface air during the day and free-tropospheric air at night at SPL, and suggests vertical stratification of water vapor with lower concentration, more isotopic depletion, and lower deuterium excess at higher altitude. These observations will be used to help validate simulations of storms with an isotope-enabled mesoscale model.
Acknowledgments
This work was supported by National Science Foundation Division of Atmospheric Sciences Grants AGS-1260462, AGS-1260368, and AGS-1260380. Logistical assistance from the Steamboat Ski and Resort Corporation is greatly appreciated. Special thanks to the support personnel from NCAR EOL. The authors gratefully acknowledge the NOAA Air Resources Laboratory (ARL) for the provision of the HYSPLIT transport and dispersion model and/or READY website (http://www.ready.noaa.gov) used in this publication. Thanks also to Betsy Andrews at NOAA/CIRES in Boulder, Colorado, for providing SPL Bap data. The Desert Research Institute is an equal opportunity service provider and employer, and is a permittee of the Medicine Bow and Routt National Forests.
REFERENCES
Aemisegger, F., Sturm P. , Graf P. , Sodemann H. , Pfahl S. , Knohl A. , and Wernli H. , 2012: Measuring variations of δ18O and δ2H in atmospheric water vapour using two commercial laser-based spectrometers: An instrument characterisation study. Atmos. Meas. Tech., 5, 1491–1511, doi:10.5194/amt-5-1491-2012.
Bastrikov, V., Steen-Larsen H.-C. , Masson-Delmotte V. , Gribanov K. , Cattani O. , Jouzel J. , and Zakharov V. , 2014: Continuous measurements of atmospheric water vapour isotopes in western Siberia (Kourovka). Atmos. Meas. Tech., 7, 1763–1776, doi:10.5194/amt-7-1763-2014.
Benetti, M., Reverdin G. , Pierre C. , Merlivat L. , Risi C. , Steen-Larsen H. C. , and Vimeux F. , 2014: Deuterium excess in marine water vapor: Dependency on relative humidity and surface wind speed during evaporation. J. Geophys. Res. Atmos., 119, 584–593, doi:10.1002/2013JD020535.
Blossey, P. N., Kuang Z. , and Romps D. M. , 2010: Isotopic composition of water in the tropical tropopause layer in cloud-resolving simulations of an ideal tropical circulation. J. Geophys. Res., 115, D24309, doi:10.1029/2010JD014554.
Blossey, P. N., and Coauthors, 2015: Isotopic Fractionation in Snow (IFRACS) at Storm Peak Laboratory. 2015 Fall Meeting, San Francisco, CA, Amer. Geophys. Union, Abstract PP11B-2224.
Bolot, M., Legras B. , and Moyer E. J. , 2013: Modelling and interpreting the isotopic composition of water vapour in convective updrafts. Atmos. Chem. Phys., 13, 7903–7935, doi:10.5194/acp-13-7903-2013.
Bond, T. C., Anderson T. L. , and Campbell D. , 1999: Calibration and intercomparison of filter-based measurements of visible light absorption by aerosols. Aerosol Sci. Technol., 30, 582–600, doi:10.1080/027868299304435.
Borys, R. D., and Wetzel M. A. , 1997: Storm Peak Laboratory: A research, teaching, and service facility for the atmospheric sciences. Bull. Amer. Meteor. Soc., 78, 2115–2123, doi:10.1175/1520-0477(1997)078<2115:SPLART>2.0.CO;2.
Borys, R. D., Lowenthal D. H. , and Rahn K. A. , 1986: Contributions of smelters and other sources to pollution sulfate at a mountaintop site in northwestern Colorado. Acid Deposition in Colorado—A Potential or Current Problem; Local versus Long-Distance Transport into the State, R. Pielke, Ed., Colorado State University, 167–174.
Borys, R. D., Hindman E. E. , and DeMott P. J. , 1988: The chemical fractionation of atmospheric aerosols as a result of snow crystal formation and growth. J. Atmos. Chem., 7, 213–239, doi:10.1007/BF00130931.
Borys, R. D., Lowenthal D. H. , and Mitchell D. L. , 2000: The relationships among cloud microphysics, chemistry, and precipitation rate in cold mountain clouds. Atmos. Environ., 34, 2593–2602, doi:10.1016/S1352-2310(99)00492-6.
Borys, R. D., Lowenthal D. H. , Cohn S. A. , and Brown W. O. J. , 2003: Mountaintop and radar measurements of aerosol effects on snow growth and snowfall rate. Geophys. Res. Lett., 30, 1538, doi:10.1029/2002GL016855.
Ciais, P., and Jouzel J. , 1994: Deuterium and oxygen 18 in precipitation: Isotopic model, including mixed cloud processes. J. Geophys. Res., 99, 16 793–16 803, doi:10.1029/94JD00412.
Clark, I., and Fritz P. , 1997: Environmental Isotopes in Hydrology. CRC Press/Lewis Publishers, 328 pp.
Colle, B. A., Wolfe J. B. , Steenburgh W. J. , Kingsmill D. E. , Cox J. A. , and Shafer J. C. , 2005: High-resolution simulations and microphyscial validation of an orographic precipitation event over the Wasatch Mountains during IPEX IOP3. Mon. Wea. Rev., 133, 2947–2971, doi:10.1175/MWR3017.1.
Craig, H., 1961: Isotopic variations in meteoric waters. Science, 133, 1702–1703, doi:10.1126/science.133.3465.1702.
Dansgaard, W., 1964: Stable isotopes in precipitation. Tellus, 16A, 436–468, doi:10.1111/j.2153-3490.1964.tb00181.x.
Dansgaard, W., White J. W. C. , and Johnsen S. J. , 1989: The abrupt termination of the Younger Dryas climate event. Nature, 339, 532–534, doi:10.1038/339532a0.
Delanoe, J. M., Heymsfield A. J. , Protat A. , Bansemer A. R. , and Hogan R. J. , 2014: Normalized particle size distribution for remote sensing application. J. Geophys. Res. Atmos., 119, 4204–4227, doi:10.1002/2013JD020700.
Demoz, B. B., Warburton J. A. , and Stone R. H. , 1991: The influence of riming on the oxygen isotopic composition of ice-phase precipitation. Atmos. Res., 26, 463–488, doi:10.1016/0169-8095(91)90039-Y.
Draxler, R. R., and Rolph G. D. , 2015: HYSPLIT model access. NOAA Air Resources Laboratory. [Available online at http://www.arl.noaa.gov/HYSPLIT.php.]
Ellehoj, M. D., Steen-Larsen H. C. , Johnsen S. J. , and Madsen M. B. , 2013: Ice-vapor equilibrium fractionation factor of hydrogen and oxygen isotopes: Experimental investigations and implications for stable water isotope studies. Rapid Commun. Mass Spectrom., 27, 2149–2158, doi:10.1002/rcm.6668.
Gedzelman, S. D., and Lawrence J. R. , 1990: The isotopic composition of precipitation from two extratropical cyclones. Mon. Wea. Rev., 118, 495–509, doi:10.1175/1520-0493(1990)118<0495:TICOPF>2.0.CO;2.
Gedzelman, S. D., and Arnold R. , 1994: Modeling the isotopic composition of precipitation. J. Geophys. Res., 99, 10 455–10 471, doi:10.1029/93JD03518.
Hallar, A. G., Mccubbin I. B. , Novosselov I. , Gorder R. , and Ogren J. , 2016: A high elevation aerosol manifold modeling study and inter-comparison. 2012 Fall Meeting, San Francisco, CA, Amer. Geophys. Union, Abstract A51A-0016.
Harmon, R. S., 1979: An isotopic study of groundwater seepage in the central Kentucky Karst. Water Resour. Res., 15, 476–480, doi:10.1029/WR015i002p00476.
Heymsfield, A. J., Schmitt C. , and Bansemer A. , 2013: Ice cloud particle size distributions and pressure-dependent terminal velocities from in situ observations at temperatures from 0° to −86°C. J. Atmos. Sci., 70, 4123–4154, doi:10.1175/JAS-D-12-0124.1.
Hindman, E. E., Carter E. J. , Borys R. D. , and Mitchell D. L. , 1992: Collecting super-cooled droplets as a function of droplet size. J. Atmos. Oceanic Technol., 9, 337–353, doi:10.1175/1520-0426(1992)009<0337:CSCDAA>2.0.CO;2.
Hobbs, P. V., 1975: The nature of winter clouds and precipitation in the Cascade Mountains and their modification by artificial seeding. Part I: Natural conditions. J. Appl. Meteor., 14, 783–804, doi:10.1175/1520-0450(1975)014<0783:TNOWCA>2.0.CO;2.
Houze, R. A., and Medina S. , 2005: Turbulence as a mechanism for orographic precipitation enhancement. J. Atmos. Sci., 62, 3599–3623, doi:10.1175/JAS3555.1.
Joussaume, S., Sadourny R. , and Jouzel J. , 1984: A general circulation model of water isotope cycles in the atmosphere. Nature, 311, 24–29, doi:10.1038/311024a0.
Jouzel, J., and Merlivat L. , 1984: Deuterium and oxygen 18 in precipitation: Modeling of the isotopic effects during snow formation. J. Geophys. Res., 89, 11 749–11 757, doi:10.1029/JD089iD07p11749.
Jouzel, J., Merlivat L. , and Lorius C. , 1982: Deuterium excess in an East Antarctic ice core suggests higher relative humidity at the oceanic surface during the last glacial maximum. Nature, 299, 688–691, doi:10.1038/299688a0.
Jouzel, J., and Coauthors, 2007: The GRIP deuterium-excess record. Quat. Sci. Rev., 26, 1–17, doi:10.1016/j.quascirev.2006.07.015.
Kumler, A., and Ogren J. , 2016: A comparison of inlet setups at Storm Peak Laboratory. NOAA, 1 p. [Available online at http://www.esrl.noaa.gov/gmd/publications/annual_meetings/2016/abstracts/24-160406-B.pdf.]
Kusunoki, K., Murakami M. , Orikasa N. , Hoshimoto M. , Tanaka Y. , Mizuno H. , Hamazu K. , and Watanabe H. , 2005: Observations of quasi-stationary and shallow orographic snow clouds: Spatial distributions of supercooled liquid water and snow particles. Mon. Wea. Rev., 133, 743–751, doi:10.1175/MWR2874.1.
Lamb, D., Nielsen K. W. , Klieforth H. E. , and Hallett J. , 1976: Measurement of liquid water content in cloud systems over the Sierra Nevada. J. Appl. Meteor., 15, 763–775, doi:10.1175/1520-0450(1976)015<0763:MOLWCI>2.0.CO;2.
Lee, J.-E., Pierrehumbert R. , Swann A. , and Linter B. R. , 2009a: Sensitivity of stable water isotopic values to convective parameterization schemes. Geophys. Res. Lett., 36, L23801, doi:10.1029/2009GL040880.
Lee, J.-E., Johnson K. , and Fung I. , 2009b: Precipitation over South America during the Last Glacial Maximum: An analysis of the “amount effect” with a water isotope-enabled general circulation model. Geophys. Res. Lett., 36, L19701, doi:10.1029/2009GL039265.
Lorius, C., Merlivat L. , Jouzel J. , and Pourchet M. , 1979: A 30,000 yr isotopic climate record from Antarctic ice. Nature, 280, 644–648, doi:10.1038/280644a0.
Lowenthal, D. H., Hanumara R. C. , Rahn K. A. , and Currie L. A. , 1987: Effects of systematic error, estimates and uncertainties in chemical mass balance apportionments: Quail Roost II revisited. Atmos. Environ., 21, 501–510, doi:10.1016/0004-6981(87)90033-3.
Lowenthal, D. H., Borys R. D. , and Wetzel M. A. , 2002: Aerosol distributions and cloud interactions at a mountaintop laboratory. J. Geophys. Res., 107, doi:10.1029/2001JD002046.
Lowenthal, D. H., Borys R. D. , Cotton W. , Saleeby S. , Cohn S. A. , and Brown W. O. J. , 2011: The altitude of snow growth by riming and vapor deposition in mixed-phase orographic clouds. Atmos. Environ., 45, 519–522, doi:10.1016/j.atmosenv.2010.09.061.
Majoube, M., 1971: Fractionnement en oxygène-18 et en deutérium entre l’eau et sa vapeur. J. Chim. Phys., 68, 1423–1436.
Masson-Delmotte, V., and Coauthors, 2008: A review of Antarctic surface snow isotopic composition: Observations, atmospheric circulation, and isotopic modeling. J. Climate, 21, 3359–3387, doi:10.1175/2007JCLI2139.1.
Merlivat, L., 1978: Molecular diffusivities of H2 160, HD160, and H2 180 in gases. J. Chem. Phys., 69, 2864–2871, doi:10.1063/1.436884.
Moore, M., Blossey P. , Muhlbauer A. , and Kuang Z. , 2016: Microphysical controls on the isotopic composition of wintertime orographic precipitation. J. Geophys. Res. Atmos., 121, 7235–7253, doi:10.1002/2015JD023763.
Muhlbauer, A., Hashino T. , Xue L. , Teller A. , Lohmann U. , Rasmussen R. M. , Geresdi I. , and Pan Z. , 2010: Intercomparison of aerosol-cloud-precipitation interactions in stratiform orographic mixed-phase clouds. Atmos. Chem. Phys., 10, 8173–8196, doi:10.5194/acp-10-8173-2010.
Noone, D., and Simmonds I. , 2002: Associations between δ18O of water and climate parameters in a simulation of atmospheric circulation for 1979–95. J. Climate, 15, 3150–3169, doi:10.1175/1520-0442(2002)015<3150:ABOOWA>2.0.CO;2.
Noone, D., and Coauthors, 2013: Determining water sources in the boundary layer from tall tower profiles of water vapor and surface water isotope ratios after a snowstorm in Colorado. Atmos. Chem. Phys., 13, 1607–1623, doi:10.5194/acp-13-1607-2013.
Obrist, D., Hallar A. G. , McCubbin I. , Stephens B. B. , and Rahn T. , 2008: Atmospheric mercury concentrations at Storm Peak Laboratory in the Rocky Mountains: Evidence for long-range transport from Asia, boundary layer contributions, and plant mercury uptake. Atmos. Environ., 42, 7579–7589, doi:10.1016/j.atmosenv.2008.06.051.
Pausata, F. S. R., Battisti D. S. , Nisancioglu K. H. , and Bitz C. M. , 2011: Chinese stalagmite controlled by changes in the Indian monsoon during a simulated Heinrich event. Nat. Geosci., 4, 474–480, doi:10.1038/ngeo1169.
Pfahl, S., and Sodemann H. , 2014: What controls deuterium excess in global precipitation? Climate Past, 10, 771–781, doi:10.5194/cp-10-771-2014.
Pfahl, S., Wernli H. , and Yoshimura K. , 2012: The isotopic composition of precipitation from a winter storm—A case study with the limited-area model COSMOiso. Atmos. Chem. Phys., 12, 1629–1648, doi:10.5194/acp-12-1629-2012.
Picciotto, E., De Maere X. , and Friedman I. , 1960: Isotopic composition and temperature of formation of Antarctic snows. Nature, 187, 857–859, doi:10.1038/187857a0.
Poage, M. A., and Chamberlain C. P. , 2001: Empirical relationships between elevation and the stable isotope composition of precipitation and surface waters: Considerations for studies of paleoelevation change. Amer. J. Sci., 301, 1–15, doi:10.2475/ajs.301.1.1.
Rauber, R. M., and Grant L. O. , 1986a: The characteristics and distribution of cloud water over the mountains of northern Colorado during wintertime storms. Part II: Microphysical characteristics. J. Climate Appl. Meteor., 25, 489–504, doi:10.1175/1520-0450(1986)025<0489:TCADOC>2.0.CO;2.
Rauber, R. M., Grant L. O. , Feng D. , and Snider J. B. , 1986b: The characteristics and distribution of cloud water over the mountains of northern Colorado during wintertime storms. Part I: temporal variations. J. Climate Appl. Meteor., 25, 468–488, doi:10.1175/1520-0450(1986)025<0468:TCADOC>2.0.CO;2.
Reinking, R. F., Snider J. B. , and Coen J. L. , 2000: Influences of storm-embedded orographic gravity waves on cloud liquid water and precipitation. J. Appl. Meteor., 39, 733–759, doi:10.1175/1520-0450(2000)039<0733:IOSEOG>2.0.CO;2.
Risi, C., Bony S. , Vimeux F. , Frankenberg C. , Noone D. , and Worden J. , 2010: Understanding the Sahelian water budget through the isotopic composition of water vapor and precipitation. J. Geophys. Res., 115, D24110, doi:10.1029/2010JD014690.
Rotunno, R., and Houze R. A. , 2007: Lessons on orographic precipitation from the Mesoscale Alpine Programme. Quart. J. Roy. Meteor. Soc., 133, 811–830, doi:10.1002/qj.67.
Rowley, D. B., Pierrehumbert R. T. , and Currie B. S. , 2001: A new approach to stable isotope-based paleoaltimetry: Implication for paleoaltimetry and paleohypsometry of the high Himalaya since the Late Miocene. Earth Planet. Sci. Lett., 188, 253–268, doi:10.1016/S0012-821X(01)00324-7.
Saleeby, S. M., Cotton W. R. , Lowenthal D. , Messina J. , and Benedict K. B. , 2013: Aerosol impacts on the microphysical growth processes of orographic snowfall. J. Appl. Meteor. Climatol., 52, 834–852, doi:10.1175/JAMC-D-12-0193.1.
Samuels-Crow, K. E., Galewsky J. , Sharp Z. D. , and Dennis K. J. , 2014: Deuterium excess in subtropical free troposphere water vapor: Continuous measurements from the Chajnantor Plateau, northern Chile. Geophys. Res. Lett., 41, 8652–8659, doi:10.1002/2014GL062302.
Schmidt, G. A., Hoffmann G. , Shindell D. T. , and Hu Y. , 2005: Modeling atmospheric stable water isotopes and the potential for constraining cloud processes and stratosphere-troposphere water exchange. J. Geophys. Res., 110, D21314, doi:10.1029/2005JD005790.
Scholl, M. A., Giambelluca T. W. , Gingerich S. B. , Nullet M. A. , and Loope L. L. , 2007: Cloud water in windward and leeward mountain forests: The stable isotope signature of orographic cloud water. Water Resour. Res., 43, W12411, doi:10.1029/2007WR006011.
Siegenthaler, U., and Oeschger H. , 1980: Correlation of 18O in precipitation with temperature and altitude. Nature, 285, 314–317, doi:10.1038/285314a0.
Smith, J. A., Ackerman A. S. , Jensen E. J. , and Toon O. B. , 2006: Role of deep convection in establishing the isotopic composition of water vapor in the tropical transition layer. Geophys. Res. Lett., 33, L06812, doi:10.1029/2005GL024078.
Smith, R. B., and Evans J. P. , 2007: Orographic precipitation and water vapor fractionation over the southern Andes. J. Hydrometeor., 8, 3–19, doi:10.1175/JHM555.1.
Smith, R. B., Barstad I. , and Bonneau L. , 2005: Oregon’s precipitation and Oregon’s climate transition. J. Atmos. Sci., 62, 177–191, doi:10.1175/JAS-3376.1.
Steen-Larsen, H. C., and Coauthors, 2011: Understanding the climate signal in the water stable isotope records from the NEEM shallow firn/ice cores in northwest Greenland. J. Geophys. Res., 116, D06108, doi:10.1029/2010JD014311.
Steen-Larsen, H. C., and Coauthors, 2014: Climatic controls on water vapor deuterium excess in the marine boundary layer of the North Atlantic based on 500 days of in situ, continuous measurements. Atmos. Chem. Phys., 14, 7741–7756, doi:10.5194/acp-14-7741-2014.
Steen-Larsen, H. C., and Coauthors, 2015: Moisture sources and synoptic to seasonal variability of North Atlantic water vapor isotopic composition. J. Geophys. Res. Atmos., 120, 5757–5774, doi:10.1002/2015JD023234.
Stewart, M. K., 1975: Stable isotope fractionation due to evaporation and isotopic exchange of falling waterdrops: Applications to atmospheric processes and evaporation of lakes. J. Geophys. Res., 80, 1133–1146, doi:10.1029/JC080i009p01133.
Sturm, C., Hoffman G. , and Langmann B. , 2007: Simulation of stable water isotopes in precipitation over South America: Comparing regional to global circulation models. J. Climate, 20, 3730–3750, doi:10.1175/JCLI4194.1.
Vaisala, 2013: Humidity conversion formulas. Vaisala Oyj Doc. B210973EN-F, 17 pp. [Available online at http://www.vaisala.com/Vaisala%20Documents/Application%20notes/Humidity_Conversion_Formulas_B210973EN-F.pdf.]
Vimeux, F., Masson V. , Jouzel J. , Stievenard M. , and Petit J. R. , 1999: Glacial–interglacial changes in ocean surface conditions in the Southern Hemisphere. Nature, 398, 410–413, doi:10.1038/18860.
Warburton, J. A., and DeFelice T. P. , 1986: Oxygen isotopic composition of central Sierra Nevada precipitation, I. Identification of ice-phase water capture regions in winter storms. Atmos. Res., 20, 11–22, doi:10.1016/0169-8095(86)90004-9.
Welp, L. R., and Coauthors, 2012: A meta-analysis of water vapor deuterium-excess in the midlatitude atmospheric surface layer. Global Biogeochem. Cycles, 26, GB3021, doi:10.1029/2011GB004246.
Werner, M., Langebroek P. M. , Carlsen T. , Herold M. , and Lohmann G. , 2011: Stable water isotopes in the ECHAM5 general circulation model: Toward a high-resolution isotope modeling on a global scale. J. Geophys. Res., 116, D15109, doi:10.1029/2011JD015681.
Wetzel, M., and Coauthors, 2004: Mesoscale snowfall prediction and verification in mountainous terrain. Wea. Forecasting, 19, 806–828, doi:10.1175/1520-0434(2004)019<0806:MSPAVI>2.0.CO;2.
Wright, J. S., Sobel A. H. , and Schmidt G. A. , 2009: Influence of condensate evaporation on water vapor and its stable isotopes in a GCM. Geophys. Res. Lett., 36, L12804, doi:10.1029/2009GL038091.
Yoshimura, K., Frankenberg C. , Lee J. , Kanamitsu M. , Worden J. , and Röckmann T. , 2011: Comparison of an isotopic atmospheric general circulation model with new quasi-global satellite measurements of water vapor isotopologues. J. Geophys. Res., 116, D19118, doi:10.29/2011JD016035.
