Comparison of semiclassical line-shape models to rovibrational ${\mathrm{H}}_2\mathrm{O}$ spectra measured by frequency-stabilized cavity ring-down spectroscopy

A single-mode cavity ring-down spectrometer, which incorporates a stabilized and tuneable comb of resonant frequencies and a continuous-wave external-cavity diode probe laser, was used to study rovibrational absorption line shapes within the and vibrational bands of water vapor. This spectrometer, which has a noise-equivalent absorption coefficient of and frequency resolution of , enables high-precision measurements of line-shape effects and pressure shifting of relatively weak absorption transitions. We investigated the room-temperature pressure dependence over the range from of two transitions perturbed by He, , and . Foreign-gas broadening and pressure-shift coefficients were determined for a relatively strong transition at , and for a weaker transition at the self- and -broadening and pressure-shift parameters were measured. In the low-pressure limit the room-temperature Doppler width was measured within 0.2% of its expected value. Doppler-free saturation effects were also observed with linewidths below . The data were compared to semiclassical line-shape models that considered the influence of Dicke narrowing as well as the speed dependence of pressure broadening and pressure shifting. Taking both of these effects into account gave the best agreement with our observations and allowed us to model the observed asymmetries of experimental profiles. Hard- and soft-collision as well as billiard-ball collision models were considered. These results allowed us to quantify systematic errors in line intensity and in pressure broadening associated with oversimplified models of line shape.

Figure 1

(Color online) Absorption spectrum of transition $A$ at the low-pressure limit and a least-squares Gaussian fit to the measurements: (a) weak transition close to transition $A$ measured at $p=7\times {10}^{-3}\mathrm{kPa}$ and (b) Lamb dip observed at the line center.

Figure 2

(Color online) Profiles of ${\mathrm{H}}_2\mathrm{O}$ transition $A$ perturbed by He at different pressures with fitted Galatry profiles (upper panel). Residuals for fits (lower panel) of VP, GP, and SDBBP to the $p=53.3\mathrm{kPa}$ profile.

Figure 3

(Color online) Dependence of fitted Lorentzian width (FWHM) ${\gamma }_L$ and shift $\Delta$ [graph (A)] and frequency of velocity-changing collisions ${\nu }_{\mathrm{opt}}$ [graph (B)] on He pressure $p$.

Figure 4

(Color online) Experimental and fitted profiles of ${\mathrm{H}}_2\mathrm{O}$ transition $A$ perturbed by ${\mathrm{N}}_2$ at different pressures (upper panel). Residuals corresponding to fits (lower panel) of VP, GP, and SDBBP to the $p=13.3\mathrm{kPa}$ profile.

Figure 5

(Color online) Dependence of fitted Lorentzian width (FWHM) ${\gamma }_L$ and shift $\Delta$ [graph (A)] and frequency of velocity-changing collisions ${\nu }_{\mathrm{opt}}$ [graph (B)] on ${\mathrm{N}}_2$ pressure $p$ for different theoretical profiles.

Figure 6

(Color online) Experimental and theoretical profiles of ${\mathrm{H}}_2\mathrm{O}$ transition $A$ perturbed by $\mathrm{S}{\mathrm{F}}_6$ at different pressures (upper panel). Residuals for fits of theoretical profiles (lower panel) to the $p=26.7\mathrm{kPa}$ data.

Figure 7

(Color online) Dependence of fitted Lorentzian width (FWHM) ${\gamma }_L$ and shift $\Delta$ [graph (A)] and frequency of velocity-changing collisions ${\nu }_{\mathrm{opt}}$ [graph (B)] on $\mathrm{S}{\mathrm{F}}_6$ pressure $p$ for different theoretical profiles.

Figure 8

(Color online) Relative deviation of the line area determined from fits of model profiles to experimental data for the ${\mathrm{H}}_2\mathrm{O}$ transition $A$ perturbed by $\mathrm{S}{\mathrm{F}}_6$ [graph (A)] and by ${\mathrm{N}}_2$ [graph (B)] at different pressures. The reference model is the SDNGP.

Figure 9

(Color online) Experimental profiles of transition $B$ and associated residuals for VP and GP fits to the data. Sample conditions correspond to $p=2\mathrm{kPa}$ of pure ${\mathrm{H}}_2\mathrm{O}$.

Figure 10

(Color online) Dependence of the self-broadening ${\gamma }_L$, shift $\Delta$, and ${\nu }_{\mathrm{opt}}$ characterizing Dicke narrowing on pressure for transition $B$ obtained from the fits of different profiles to the experimental data.

Figure 11

(Color online) Area $\mathcal{A}$ of transition $B$ is plotted against the ${\mathrm{H}}_2\mathrm{O}$ concentration $N$ for the VP, GP, and NGP fits. The dashed line corresponds to the HITRAN 2004 value of line intensity evaluated at $T=296\mathrm{K}$.

Figure 12

(Color online) Experimental and theoretical profiles of transition $B$ perturbed by ${\mathrm{N}}_2$ at different pressures and residuals for fits of theoretical profiles to the $p=13.3\mathrm{kPa}$ spectrum.

Figure 13

(Color online) Dependence of the fitted Lorentzian width (FWHM) ${\gamma }_L$ and shift $\Delta$ [graph (A)] and frequency of velocity-changing collisions ${\nu }_{\mathrm{opt}}$ [graph (B)] of transition $B$ on the ${\mathrm{N}}_2$ pressure $p$ for different theoretical profiles.