LFH

Humidity Reference Instruments

1. Lyman alpha Resonance Fluorescence Hygrometer

For the water vapor measurements a Lyman-alpha fluorescence hygrometer (LFH) [Kley et al., 1978], is installed in the simulation chamber as reference instrument. The principle of the instrument is based on the absorption measurement of H₂O at the Lyman-alpha line of hydrogen (l =121.6 nm) in combination with the detection of the OH (l =310 nm) fluorescence which is produced by photolysis of H₂O at the Lyman-alpha radiation. This combination of detection techniques makes a two beam absorption instrument unnecessary and allows for measurements of extremely low H₂O concentrations.

Figure 1:  Cross section of Lyman-alpha fluorescence hygrometer perpendicular to the flow channel of the instrument. The radiation of the lamp enters the air stream channel through a pinhole and is partly reflected by a MgF₂ plate to cell B. After traversing the flow duct the light beam leaves it through a pinhole and reaches cell B. The scattered radiation inside the flow duct is detected by the photomultiplier looking perpendicular to the light beam in the duct.

Figure 1 shows a cross section of the instrument, perpendicular to the flow duct. In a rigid ground plate an air flow duct is mounted. The radiation from the Lyman-alpha lamp enters the flow duct through a hole and runs through a second hole to the NO cell A. This establishes the main absorption path x_A of the instrument. For a secondary light path (x_B) a small part of the radiation is reflected by a MgF₂ plate to a second NO cell B. Light path x_B is shorter than x_A.

Absorption Mode

The Lyman-alpha intensities I_A and I_B at cell A respectively B in this set-up are then given by

where I_o the lamp intensity, [O₂] and [H₂O] the concentrations of oxygen and water, and s _O2 and s_H2O the corresponding absorption cross sections. C is a constant, which accounts for the different beam geometry, reflectivity of the MgF₂ mirror and sensitivity of NO cell B relative to NO cell A.

If factor C is known, the water vapor concentration can be determined by

Under conditions of very low water vapor concentration ([H₂O] » 0), the factor C is obtained from

A first estimate of C can therefore be taken from the measurements of the lowest water vapor concentrations measured during an experiment. For a precise determination of C also the fluorescence measurement has to be taken into account.

Fluorescence Mode

The Lyman-alpha radiation dissociates H₂O molecules into OH and H

H₂O + hn    Þ  OH(A^{2S +}) + H(²S).

One part of the electronically excited OH gives a fluorescence at 310 nm wavelength, while the rest is electronically quenched by N₂ and O₂

OH(A^{2S +})  Þ   OH(X^2P ) + hn    (l = 310 nm)

The 310 nm fluorescence is observed with a photomultiplier mounted in the flow duct at right angle to the direct light beam traversing the flow duct on path A.

The measured fluorescence signal S (after subtraction of the background counting rate) is proportional to the H₂O volume mixing ratio m_H2O for moderate H₂O concentrations. It is not exactly linearly dependent on m_H2O, due to Ly(a)-light absorption between lamp and fluorescence volume. Therefore the relation for S can be written as

where x_F is the length of the light path from the lamp to the middle of the fluorescence volume in the flow duct.

The exponential expression describes the "pre-absorption" of Ly(a)-light before it photodissociates H₂O in the fluorescence volume and gives rise to the OH fluorescence at l =310 nm. Since the layout of the instrument provides that x_B = x_F, the pre-absorption is proportional to the signal of cell B and relation 2-9 can be written as

By dividing the measured fluorescence counting rate through the measured I_B signal one has effectively linearized the response of the instrument with respect to "pre-absorption" and eliminated any changes due to fluctuations in lamp intensity [Kley et al., 1979]. Thus the corrected counting rate S_corr of the fluorescence can be expressed as

which is now exactly linear in m_H2O. Constant B is the sensitivity of the corrected fluorescence count rate. An estimated m_H2O can be computed by equation 3 with an estimate of constant C. S_corr is linearly correlated with the estimated volume mixing ratio m_H2O showing normally also an offset. Using this offset and the slope of the linear regression, the correct C can be computed and B in equation [7] can be determined. This is done for each temperature level, at which the calibration procedure is executed. The slope B in equation [7] of the final linear regression gives a calibration of the fluorescence count rate for the H₂O mixing ratio.

The accuracy of the water vapor mixing ratio measurements in the fluorescence mode of the instrument is ± 4%, including random and systematic erors with inclusion of the accuracy of the absorption cross sections for water vapor and oxygen at l =121.6 nm wavelength [Kley, 1984].

2. Dew Point Hygrometer

For humidities larger than 0.5g/kg, i.e. lower tropospheric conditions, an optical condensation dew point hygrometer (General Eastern 1311DR) is used, allowing microprocessor-based control, measurement and display. It incorporates a chilled mirror dew point sensor for measurements in a dew point range of -75°C to +95°C. It is calibrated against dew point standards that are traceable to the U.S. National Institute of Standards and Technology.

The principle of the detection method is the optical observation of a metallic mirror, which can be controlled within a wide range of temperatures by Peltier elements and a resistance heater. The dew layer building up, when the mirror temperature is falling below the dew point temperature, is optically detected and its temperature than held at that temperature and measured with a precise Platinum resistance thermometer, delivering the dew/frost point temperature. The optical detection of the dew layer is done by monitoring the light of a high intensity LED, reflected from the mirror surface, with a photo detector. When the dew layer forms, diffuse reflection leads to an attenuated signal of the photo detector. The specifications of the instrument are shown in Table 1.

The air to be monitored is sucked with a stainless steel tube (6 mm Æ) from the sample flow of the of the Lyman-alpha hygrometer (LFH) into the dew point hygrometer (DPH) by a small bellows pump with 2 l/min (STP). The tube is electrically heated with a coaxial resistance heater on its full length to avoid memory effects of the wall inside the tube.

Table 1:  Specifications of dew point mirror hygrometer General Eastern D1311R

Figure 2:  Set points of air temperature, wall temperature, and relative humidity with resulting measured dew-point temperature (dew point hygrometer) and relative humidity (Lyman-alpha hygrometer) as function of simulation time.

Figure 2 gives an impression of the instrument performance in comparison with the Lyman-alpha hygrometer measurement in the simulation chamber together with pressure and temperature as function of the simulation time. For temperatures below -20°C there is a positive offset of the dew point hygrometer in the order of 1% RH compared with the Lyman-alpha. For relative humidities above 4% the reading of the dew point hygrometer is close to that of the Lyman-alpha for all temperatures. At steep changes of the humidity the response time of the dew point instrument is long, increasing with decreasing temperature, and needs some minutes of recovery. At relative humidity levels below 4% and fast changes of humidity the dew point hygrometers accuracy is not sufficient for calibrations, while at higher humidity levels the relative deviation between dew point hygrometer and Lyman-alpha hygrometer is systematically below 10% of the reading, provided fast humidity changes are excluded.

References

Kley, D. and Stone, E.J., Measurement of water vapor in the stratosphere by photodissociation with Ly (a) (1216 A) light, Rev. Sci. Instrum., 49, 691-697, 1978.

Kley, D., Stone, E.J., Henderson, W.R., Drummond, J.W., Harrop, W.J., Schmeltekopf, A.L., Thompson, T.L., and Winkler, R.H., In Situ Measurements of the Mixing Ratio of Water Vapor in the Stratosphere, J. Atm. Sci. 36,2513-2524, 1979.

Kley, D., Ly (a) absorption cross-section of H₂O and O₂, J. Atm. Chem., 2, 203-210, 1984.