76

# Handbook of Humidity Measurement

Print publication date:  March  2018
Online publication date:  March  2018

Print ISBN: 9781138300217
eBook ISBN: 9780203731956

10.1201/b22369-3

#### Abstract

Optical hygrometry is a classic method of humidity measurement. Detection and determination of the water vapor concentration take place through the measurements of optical properties. The first known measurements of the water vapor absorption in the infrared (IR) were undertaken by Fowle in 1912. However, optical methods had the greatest development only in the last 30 years, when there appeared the low-cost optoelectronic components such as light-emitting diodes (LEDs), laser diodes (LDs), photodetectors (PDs), and a fast data-acquisition and the data-processing technique. Optical hygrometers form a sufficiently large class of devices, which can be divided into subgroups on the basis of (1) the waveband used for the detection of water vapor (visible, infrared, and ultraviolet), (2) the fraction of light, which is detected after interaction with the sample (i.e., the transmitted, reflected, or absorbed fraction), and (3) spectral width of light (e.g., monochromatic, polychromatic, and reference light with continuous spectrum). All these types of optical hygrometers, including their construction, principles of operation, and approaches to practical realization are discussed in the present chapter.

#### 3.1  Introduction

As it was shown in Chapter 2, the determination of humidity is of great importance. Therefore, humidity control becomes imperative in all fields of our activity, from production management to creating a comfortable environment for our living, as well for understanding the nature of the changes happening in the climate. In the second volume, Electronic and Electrical Humidity Sensors, it will be shown that there are a large number of various devices that allow controlling humidity. Along with conventional devices such as mechanical (hair) hygrometers, psychrometers, chilled mirror hygrometers, the Dunmore cells, coulometric hygrometers, and so on, humidity can be controlled using capacitive-, resistive-, surface acoustic wave (SAW)-, quartz crystal microbalance (QCM)-, cantilever-, field effect transistor (FET)-, Schottky barrier-, field emission-based humidity sensors, and many others. In the past decades, it was done a lot for the development of these devices and for the development of different measuring systems with an increased efficiency. However, the process of developing new humidity sensors as well as improving older types of devices used for humidity measurement is still ongoing. New technologies and the toughening of ecological standards require more sensitive instruments with faster response times, better selectivity, and improved stability. In this regard, optical humidity sensors, which are considered in this book, are of particular interest.

Optical hygrometry is a classic method of humidity measurement. However, unlike above-mentioned devices, the work of optical hygrometers is based on other principles. Detection and determination of the water vapor concentration take place through the measurements of optical properties. The first known measurements of the water vapor absorption in the infrared (IR) were undertaken by Fowle (1912) for the 1.13 and 1.37 μm bands using spectroscopic techniques. However, optical methods had the greatest development only in the last 30 years, when there appeared the low-cost optoelectronic components such as light-emitting diodes (LEDs), laser diodes (LDs), photodetectors (PDs), and a fast data acquisition, and the data processing technique.

Optical hygrometers form a sufficiently large class of devices, which can be divided into subgroups on the basis of (1) the waveband used for the detection of water vapor (visible, infrared [IR], and ultraviolet [UV]) (Table 3.1), (2) the fraction of light, which is detected after interaction with the sample (i.e., the transmitted, reflected, or absorbed fraction), and (3) spectral width of light (e.g., monochromatic, polychromatic, and reference light with continuous spectrum).

### Table 3.1   Spectral Regions Important for Optical Sensors

Range

, eV

λ, nm

ν, cm−1

Infrared (IR)

0.025–0.4

3000–50000

200–3300

Near-infrared (NIR)

0.4–1.6

780–3000

3100–13000

Visual (Vis)

1.6–3.3

380–780

13000–26000

Ultraviolet (UV)

2.3–6.2

200–380

26000–50000

λ is wavelength; ν = λ−1 is wave number; is energy of photons.

Source: Korotcenkov, G. (Ed.), Electrochemical and Optical Sensors, Momentum Press, New York, pp. 311–476.

The most studied and the most popular are optical monochromatic hydrometers of absorption type, working in IR (1–14 μm) and UV (~100 nm) spectral ranges (Cerni 1994; Wiederhold 1997), where there is a large number of absorption bands, attributed to different gases (Figure 3.1). The areas of the electromagnetic (EM) spectrum that are absorbed by atmospheric gases such as oxygen, carbon dioxide, ozone, and other gases are known as absorption bands. Absorption bands corresponding to various gases in the IR и UV spectral regions are presented in Table 3.2. The main absorption bands, peculiar to the liquid water and the water vapor, are in the same band (Table 3.2 and Figure 3.1). It is important that the water molecule absorbs electromagnetic radiation both in a range of wavebands, and on discrete wavelengths.

Figure 3.1   Transmission spectra of atmosphere. It is seen that the atmospheric transparence is controlled mainly by absorption by atmospheric gases and vapors such CO2, O2 and H2O.

### Table 3.2   Main Visible and Near-IR Absorption Bands of Atmospheric Gases

Gas

Center

Band Interval (cm−1)

ν (cm−1)

λ (μm)

H2O

3703

5348

7246

9090

10638

12195

13888

visible

2.7

1.87

1.38

1.1

0.94

0.82

0.72

2500–4500

4800–6200

6400–7600

8200–9400

10100–11300

11700–12700

13400–14600

15000–22600

CO2

2526

3703

5000

6250

7143

4.3

2.7

2.0

1.6

1.4

2000–2400

3400–3850

4700–5200

6100–6450

6850–7000

O3

2110

3030

visible

4.74

3.3

2000–2300

3000–3100

10600–22600

O2

6329

7874

9433

13158

14493

15873

1.58

1.27

1.06

0.76

0.69

0.63

6300–6350

7700–8050

9350–9400

12850–13200

14300–14600

14750–15900

N2O

2222

2463

3484

4.5

4.06

2.87

2100–2300

2100–2800

3300–3500

CH4

3030

4420

6005

3.3

2.20

1.66

2500–3200

4000–4600

5850–6100

CO

2141

4273

4.67

2.34

2000–2300

4150–4350

Source: Korotcenkov, G. (Ed.), Electrochemical and Optical Sensors, Momentum Press, New York, pp. 311–476. With permission.

It was established that absorption of electromagnetic radiation by water depends on the state of the water; and absorption in different spectral ranges has different nature. The water molecule, in the gaseous state, has three types of transitions that can give rise to absorption of electromagnetic radiation:

• Rotational transitions, in which the molecule gains a quantum of rotational energy.
• Vibrational transitions in which a molecule gains a quantum of vibrational energy.
• Electronic transitions in which a molecule is promoted to an excited electronic state.

It is important to know that the electronic levels are associated with the energy of the electron subsystem. The transition from one electronic level to another can be considered as the transition of one of the electrons from one orbital to another. The vibration levels are related with the vibration motions of the molecules. The transitions between them have almost no effect on the electron subsystem. The rotational levels arise from rotations of molecules as a whole.

The energies of the transitions increase in order: rotational<vibrational<electronic (Tkachenko 2006). This is the order of decreasing mass of the considered subsystem: molecule<atom>electron. A smaller mass results in a higher frequency of oscillators (ν ~ 1/√m for harmonic oscillators). Typical ranges of the transition are summarized in Table 3.3 and presented in Figure 3.2.

Figure 3.2   Relations between different scales used in spectroscopy: Wavelength, λ, wavenumber, υ, and frequency, ν.

(Data extracted from Tkachenko, N.V., Optical Spectroscopy. Methods and Instrumentations, Elsevier, Amsterdam, the Netherlands, 2006.)

### Table 3.3   Energies and Spectral Ranges of Different Types of Transitions

Type

Energy, J

Frequency, Hz

Wavelength, μm

Electronic

(2–10)·10−19

(3–15)·1014

0.2–1.0

Vibrational

(2–20)·10−20

(3–30)·1013

1.0–10

Rotational

(2–20)·10−21

(3–30)·1012

10–100

Source: Tkachenko, N.V., Optical Spectroscopy. Methods and Instrumentations, Elsevier, Amsterdam, the Netherlands, 2006.

As it is seen in Table 3.4, energy of transition can be presented in different units such as energy (ev), wavelength, and frequency. One should note that the wavelength, frequency, and energy are equivalent measures in spectroscopy. This takes place because the energy of a photon determines frequency (and wavelength) of the electromagnetic wave (by famous Planck formula E = or E = hc/λ in vacuum). Appropriate functions which can be used for conversion of these units are presented in Table 3.4. One practical inconvenience of this is that numerous units used to characterize one and the same parameter-transition energy. Usually the wavelengths and energy are being used in UV spectroscopy. IR spectroscopy usually uses the wavelength and the wavenumber. The wavenumber is used in IR spectroscopy, because the wavenumber is directly proportional to the frequency and energy (υ = ν/c = E/hc), and therefore it is convenient to use this unit when energy or frequency dependence is presented. The microwave and the terahertz (THz) spectroscope usually use the frequency.

### Table 3.4   Parameters and Units Usually Used in Spectroscopy

Quantity

Relationship

Usually Used Units

Wavelength

λ

µm, nm

Wavenumber

υ = 1/λ (cm)

cm−1

Frequency

ν = c/λ (m)

Hz

Energy

(c = 3 × 108 m·s−1)

eV

Velocity of light

= 1.24/λ (µm)

C = ν λ (m)

m s−1

Note: h is Planck’s constant (6.62606957 × 10−34 J·s; c is the speed of light in vacuum (299792458 m/s).

As it is seen in Figure 3.2, rotational transitions are responsible for absorption in the microwave and far-infrared (FIR), vibrational transitions in the mid-infrared (MIR) and near-infrared (NIR). In the liquid water, the rotational transitions are effectively quenched. In reality, vibrations of molecules in the gaseous state are accompanied by rotational transitions, giving rise to a vibration–rotation spectrum. Furthermore, vibrational overtones and combination bands occur in the NIR region. Therefore, vibrational bands have rotational fine structure. In crystalline ice the vibrational spectrum is also affected by the hydrogen bonding and there are lattice vibrations causing absorption in the FIR. Electronic transitions occur in the UV regions. Electronic transitions of gaseous molecules will show both vibrational and rotational fine structure.

UV and IR absorption optical hygrometers offer several advantages in various fields of application (Wernecke and Wernecke 2014). These devices are capable of observing a sample in its dynamic environment, no matter how distant, difficult to reach, or hostile this environment is. These devices are intrinsically safe, involving a low optical power, and are nonelectrical at the sensing point. These hygrometers are electrically passive and immune to electromagnetic disturbances, are geometrically flexible and corrosion-resistant, and are compatible with telemetry. In principle, these devices are capable of continuous use. These features impart to the optical hygrometers immense potential importance in biomedical, process, and environmental monitoring applications.

It should be noted that there are also hygrometers, which operate in the THz and microwave regions of the spectrum. These devices differ significantly from the optical hygrometers, because basically forced to use other radiants and radiation detectors. However, the base of these sensors operation is also the absorption of radiation by molecules of water vapor. Therefore, the techniques, which are often classified as optical hygrometry, more correctly to classify as electromagnetic radiation (EMR) absorption hygrometry (WMO 1992, 2011).

It is important that in contrast to the absorption bands, there are areas of the EM spectrum, where the atmosphere is transparent (little or no absorption of energy) to specific wavelengths (Figure 3.3). These wavelength bands are known as atmospheric windows because they allow the energy to easily pass through the atmosphere to the Earth’s surface. As it is seen in Figures 3.1 and 3.3, a clear electromagnetic spectral transmission window can be seen between 8 and 14 µm. A fragmented part of the window spectrum (one might say a louvered part of the window) can also be seen in the visible to mid-wavelength IR between 0.2 and 5.5 µm. You can see that there is plenty of atmospheric transmission of radiation at 0.5, 2.5, and 3.5 µm, but in contrast there is a great deal of atmospheric absorption at 2.0, 3.0, and about 7.0 µm. Atmospheric windows are also observed in spectral range from 4 cm to 20 m. It is understandable, that both passive and active remote-sensing technologies do best if they operate within the atmospheric windows. For example, the range from 4 cm to 20 m is used for operation of the radio telescopes, studying the Universe.

Figure 3.3   Rough plot of the Earth atmospheric transmittance (or opacity) to various wavelengths of electromagnetic radiation, including visible light.

(From http://www.nasa.gov. This file is in the public domain in the United States because it was solely created by NASA. NASA copyright policy states that “NASA material is not protected by copyright unless noted.”)

Among hygrometers operated using the absorption of electromagnetic radiation, it is also necessary to allocate the hygrometers, which are based on the detection of the fluorescence intensity of water molecules induced by the UV irradiation, and the hygrometers, using for the water vapors detection a remote sounding principles such as light detection and ranging (LIDAR).

#### 3.2  Principles of Operation of Electromagnetic Resonance Hygrometers Based on Absorption

The principle of the method, based on the absorption of electromagnetic radiation, is to determine the attenuation of radiation in a waveband that is specific to water–vapor absorption, along the path between a source of the radiation and a receiving device. Thus, the transmitted radiation carries the sought for information on the analyte. In particular, its intensity (I) is related to the analyte concentration (C), and this relation is expressed by the Beer–Lambert law, which is a combination of the Bouguer–Lambert law and the Beer law. The former gives an expression for the light intensity attenuation dI due to absorption and scattering along an optical path of dl. The latter describes the dependency of the intensity attenuation dI from the concentration c of sample. The differential Beer–Lambert law Equation 3.1 states that a variation in intensity dI along a differential path length dl is related to an absorption coefficient α(λ) and to the concentration C of the sample:

3.1 ()

Integration of both sides of Equation 3.1 yields

3.2 ()
3.3 ()

where:

• I 0 is the intensity of the incident radiation
• α(λ) is molar absorptivity
• l is the length of the optical path in the absorbing phase

The molar absorptivity is characteristic of a species at a given wavelength. Thus, the analyte concentration is a linear function of the quantity log(I 0/I), which is defined as absorbance (A).

In our case, the Beer–Lambert principle states that when light energy at certain wavelengths travel through gas, a certain amount of the energy is absorbed by the water within the path. The amount of light energy lost is related to the concentration of water.

One should note that this equation is valid only under the assumptions of

• Parallel and monochromatic incident light
• Homogeneous absorption in the medium
• No other interaction processes within the medium
• Negligible scattering and reflection at entrance and exit windows

Figure 3.4 shows the two transmittance measurements that are necessary to use absorption to determine the concentration of water vapors in the gas. The diagram in Figure 3.4a is for gas only, that is, dry gas, and the diagram in Figure 3.4b is for an absorbing sample in the same gas, that is, for air or gas contenting the water vapors. In this example, IS is the source light power that is incident on a sample; I is the measured light power after passing through the wet gas or air, and sample holder; and I 0 is the measured light power after passing through only the dry gas and sample holder. The measured transmittance in this case is attributed to only the analyte, that is, water vapors.

Figure 3.4   (a,b) Two transmittance measurements necessary for determining the concentration of water vapors in gas.

With regard to the limitations to Beer’s law, then one can select the following: at first, Beer’s law has fundamental limitations. This law is valid only for low concentrations of analyte. At higher concentrations, the individual particles of analyte no longer behave independently of each other. The resulting interaction between particles of analyte may change the analyte’s absorptivity. In addition, the analyte’s concentration via the sample’s refractive index (RI) can influence the values of

. Only for sufficiently low concentrations of analyte, the RI is essentially constant and the calibration curve is linear. At second, Beer’s law has chemical limitations. A chemical deviation from Beer’s law may occur if the analyte is involved in an equilibrium reaction. At third, Beer’s law has instrumental limitations. There are two principal instrumental limitations to Beer’s law. The first limitation is that Beer’s law assumes that the radiation reaching the sample is of a single wavelength—that is, that the radiation is purely monochromatic. However, even the best wavelength selector passes radiation with a small, but finite effective bandwidth. Polychromatic radiation always gives a deviation from Beer’s law, but the effect can be reduced to make absorbance measurements at the top of a broad absorption peak. Stray radiation is the second contribution to instrumental deviations from Beer’s law. Stray radiation arises from imperfections in the wavelength selector that allow light to enter the instrument and reach the detector without passing through the sample.

#### 3.3  Instrumentation and Configuration of Optical Hygrometers—General View

In principle optical devices, which can be used as optical hygrometers, can be classified depending on the way how the specific wavelength of interest is extracted from the light source. On one hand, a dispersive spectrometer includes an optical device such as a prism or a diffraction grating to spread the light spectrum and isolate a specific wavelength band. It uses their absorption characteristics to measure ingredients and quantity of a sample. On the other hand, nondispersive spectrometer isolates the specific wavelength band of the light that corresponds to the absorption spectrum of the target gas using narrowband transmission filters. Fourier transform spectroscopy represents another gas detection technique, especially in the IR spectra Fourier transform infrared (FTIR) spectroscopy, which is widely used for multipurpose applications. It is based on Michelson interferometer and a later Fourier transform to obtain the spectrum. Of course, nondispersive technique as the easiest to use, does not require the use of expensive and bulky spectrometers and enables the development of compact devices. It is the most suitable for the development of optical hygrometers.

The instrumentation associated with optical hygrometer is similar to that associated with conventional spectrophotometric techniques (Smith 2007). Typical optical hygrometers consist of an IR or UV light source (L), one or more optical filters (F), entrance and exit windows, a chopper for light pulsing (C), and an IR or UV detectors (D) with associated electronics (signal converter and amplifier, and digital and analog processing units). The components are assembled together into a sampling chamber (S) that acts as a sealed optical bench, while allowing humidified gas to flow through (Figure 3.5). The utilized wavelength has to be fitted to the specific measurement situation, taking into account technical capabilities (the presence of appropriate source and detector of radiation), and the presence of other components that may also interact with the radiation in selected range, and expected percent of water vapor in the sample.

Figure 3.5   Scheme of principal setups for optical humidity measurement in gas: 1-light source; 2-focusing elements; 3-modulator; 4-chopping; 5-optical filters; 6-sample (air, gas); 7-measuring cell; 8-measurement path; 9-detector.

Photomultiplier tubes, photodiodes, photodiode arrays, phototransistors, and the charge-coupled devices (CCDs) are used as optical detectors. Photomultiplier tubes are the most sensitive photodetection systems and are preferred for a low-level light detection. Phototransistors and photodiodes (which are also known as quantum detectors) offer compactness and miniaturization of analytical systems, and are useful for measurements in the UV, visible, and NIR spectral regions. Photodiode arrays and CCDs are multichannel detectors that can perform simultaneous detection of the dispersed optical radiation.

The optical filters are among typical wavelength selectors employed in optical chemical hygrometer. The optical filters are selected to optimize detection only in that part of the spectrum around the water absorption lines.

At present, a multitude of configurations have been adopted in optical hygrometers. However, they are mostly attributed to two principal methods used for determination of the degree of attenuation of the radiation, such as (WMO 1992, 2011):

1. Transmission of narrow-band radiation at a fixed intensity to a calibrated receiver. This approach is illustrated in Figure 3.5.
2. Transmission of radiation at two wavelengths, one of which is strongly absorbed by water vapor and the other being either not absorbed or only very weakly absorbed. If a single source is used to generate the radiation at both wavelengths, the ratio of their emitted intensities may be accurately known, so that the attenuation at the absorbed wavelength can be determined by measuring the ratio of their intensities at the receiver.

The simplest embodiment of the measurement method is shown in Figure 3.6. The two beams, each with a different wavelength, are created by a radiation source with a broad spectrum (e.g., lamp of LED) and an electronic or mechanical chopper that switches between two optical filters F1 and F2, which are transparent for different wavelengths. This results in a wavelength modulation between λ1 and λ2 (or more). This approach can also be realized using two lasers with different spectrum of radiation.

Figure 3.6   Setup of a two-beam measurement device: 1-measurement path, 2-reference path. The two-channel transmission cell combines reference and signal pathways. One cell is filled with reference gas, which does not interact with infrared radiation. The gas to be measured can be inserted into other cell through the gas inlet. Two detectors D1 and D2 act as receivers to record the transmitted intensity through both pathways.

As follows from the previous discussions, optical measurements of the transmittance of the gaseous medium serve as the physical basis for instrumentally determining the volume concentration of a specific gaseous component of the atmosphere from its absorption bands. However, experiments have shown that optimal measurements took place only for transmittance changes in the range from 15% to 85%; outside this range, the errors caused by errors in the photometry and the zero setting of the device became substantial. Unfortunately, in practice, we usually encounter other situations, and during direct absorption measurements in the atmosphere with low concentration of water vapor we have to resolve very small changes in a large signal; essentially, this is the equivalent of finding a needle in a haystack. Thus, measuring low concentrations of the sample material is limited by the noise present in the measurement of the background. It was found that a good way to solve the above-mentioned problem is to measure not the transmittance, but its derivative (differential). That is why optical humidity analyzers mainly present a differential optical absorption spectroscopy (DOAS) (Sigrist 1994; Kosterev and Tittel 2002). In a DOAS, to reduce errors due to fluctuations of the source, temperature, and so on, a second channel with cell is utilized for reference. The measurement signal and reference signal can then be combined to give a measurement value. A simplified optical scheme for such a humidity hygrometer is shown in Figure 3.7. The optical resolution is determined by the total length and precision of mirror displacement.

Figure 3.7   Optical scheme for moisture measurement based on differential spectroscopy: 1, light source; 2 and 3, semitransparent mirrors; 4 and 7, attenuators; 5 and 8, mirrors; 6, sample cell; 6b, reference cell; 9, analyzer.

A structural two-channel layout can be implemented in which the signal is formed as the difference between the working and reference fluxes, using a null method of comparison, with optical, electrical, or gaseous compensation. This method assumes that the signals from the two channels are equal when there is no absorbing gas in the measurement cell. So, in comparison with direct spectroscopy, DOAS produces a signal which is directly proportional to the concentration of the species. The dynamic range of gas concentrations that can be measured in the best devices with zero compensation is limited by errors in setting and maintaining in time the equality of the channels, as well as by the photometric accuracy, which is usually maintained to within 1%–3% (Mirumyants and Maksimyuk 2002). The incorporation of various automatic compensation systems makes it possible to use simple methods to compensate for the errors associated with ambient temperature variation, the dustiness of the optical elements of the system, and so on.

However, it should be recognized that the most effective approach to the development of optical hygrometers is an approach based on the use of tunable diode lasers (TDLs). The hygrometers developed based on this principle are called a tunable diode laser absorption spectrometers (TDLAS), or TDLAS hygrometers. Advantages of such devices are listed in Table 3.5. Examples of the TDLAS hygrometers realization will be considered in Section 3.4.2. As is clear from the title, the basis of these devices constitutes the TDLs appeared in the 1960s. The main advantage of TDLs is the ability to change the wavelength of the radiation. In principle, it is possible to change the frequency of the emitted light by changing the temperature. But in TDLs, the changing of its wavelength is reached by ramping the injection current. Emission bandwidths for the TDLs are on the order of 10−4–10−5 cm−1. During the measurements, the laser temperature is kept constant by using of a thermoelectric (Peletier) heat pump array. The laser is also modulated at high frequency. Consequently, by passing light at the water absorption frequency through a sample chamber containing air or natural gas of a certain moisture content, it is possible to precisely establish the water content by measuring the amount of loss in the absorption spectrum obtained by adjustment of the laser radiation. One should note that TDLAS has the ability to rapidly tune the lasers, so techniques such as wavelength modulation spectroscopy (WMS), which yields dramatic sensitivity enhancements over a direct absorption approach, are easily implemented (Linnerud et al. 1998). In WMS, the wavelength of the optical source is modulated, so that a small region of the absorption spectrum of an analyte, containing a distinct absorption peak, is repetitively scanned. As the wavelength of the source scans through the spectral feature of interest the relative attenuation of the source power is varied. Thus, by modulating the wavelength of the source, a time-varying signal (i.e., the optical power) is produced at the detector. A phase-sensitive detection scheme is used to detect the presence of, and quantify, the time-varying signal produced by the presence of the analyte species. Since the line widths of the lasers are much narrower than the individual absorption lines observed for the gas-phase molecules, the TDLAS technique obtains an extremely high spectral resolution, which results in the ability to isolate a single rotation–vibrational transition line of the analyte species (Figure 3.8), thereby reducing (close to eliminating) the background interference encountered by conventional IR–NIR spectrometers. Such ability of TDLAS provides both extremely high specificity for the analyte, even in a complex sample matrix, and detection limit that is several orders below a conventional absorption measurement.

Figure 3.8   Basic setup of tunable laser absorption spectroscopy (TLAS).

(Idea from Bange, J. et al., Airborne Measurements for Environmental Research, Wiley-VCH, Weinheim, Germany, 2013.)

• Very fast response in both directions: dry to wet or wet to dry
• Noncontact measurements
• No sensing surface to degrade due to exposure during exploitation in various atmospheres
• High long-term stability
• No zero or span gases needed
• Based on fundamental measurement Immune to many contaminants present in the air or natural gas

• Relatively expensive
• Must be calibrated using a test gas with the same basic major components of the measured gas
• Measurement is made a close to atmospheric pressure
• Strong pressure and temperature control
• Skilled support is required because troubleshooting and repair can be very difficult

Phase-sensitive detection, normally employed in WMS, permits the measurement of different multiples of the drive frequency used to modulate the output of the tunable diode laser. Most implementations of TDLAS technology have made use of the second harmonic signal (known as 2F) for two main reasons. First, if the 1F spectrum is a very close approximation to the derivative of the absorbance peak, whereas the second harmonic signal approximates the second derivative of the absorption spectrum. Such operation removes sloping backgrounds and offsets that are the result of the less than ideal output characteristics of the TDLs. This means that the 2F signal produces a zero baseline signal. And second, the second harmonic spectra display a peak, which coincides with the peak in the absorbance spectrum. Moreover, the 2F peak height is directly proportional to the partial pressure of water in the absorption cell. By simultaneously measuring the cell total pressure the concentration of water vapor can be determined. The transmitted light intensity can be related to the concentration of the absorbing water vapor by the Lambert–Bouguer law. The simultaneous measurement of the gas temperature and total pressure and other humidity parameters as absolute humidity, a dew point can be determined with a high degree of precision using psychrometric equations.

One should note that TDLAS hygrometers can be designed using different optical schemes such as single path, multipath beam, single modulation, double modulation, open path, and closed path (Buchholz et al. 2013). Some of them will be described in Section 3.4. The optical path length necessary for TDLAS depends on the required S/N ratio, the dynamic range, the line strength of the transition, the temporal response, but also on detector properties, choice of electronics, optical noise by fibers and other optical elements, and many other factors, making the TDLAS system design a nontrivial optimization problem (Buchholz et al. 2013). Path length folding, such as in White (White 1942; Lübken et al. 1999), Herriott (Herriot et al. 1964; May 1998), and astigmatic Herriott (McManus et al. 1995) arrangements, is a frequently applied principle, but difficult to optimize in applications.

Additional information related to TDLAS, one can find in numerous reviews and original articles (Allen 1998; Werle 2004; Buchholz et al. 2013; White and Cataluna 2015).

It is noted that conventional IR and NIR spectrometers do not have such high wavelength resolution (Figure 3.9), and therefore such hygrometers one can use only the spectral ranges, where there is not much overlap in the spectra of the absorbing species.

Figure 3.9   Illustration showing TDLAS narrow emission bandwidth versus conventional IR techniques.

TDLAS hygrometers are capable of fast response. The optical response is about 2 s. However, it takes time to purge the absorption cell and sampling system. Therefore, typical system response times are in the range of 3–10 min for a 90% step change. The typical accuracy of a TDLAS hygrometer is 2% of reading in terms or the mole fraction or ppmv. As a rule, modern TDLAS hygrometers use a self-calibrating data evaluation strategy based on the first principles approach and known parameters such as the absorption line strength, pressure, gas temperature, and absorption path length. The best TDLAS hygrometers elaborated recently have the detection limit ~50 ppb (0.05 ppm). This is achieved using a multipass absorption cell to achieve a longer optical path length and operating the cell at a vacuum pressure.

One should note that application of lasers with such cell allows developing very compact, robust, and lightweight instruments. Advantages, indicated previously and listed in Figure 3.9, as well as performances of TDLAS make it very suitable for a number of various applications. For example, using TDL spectrometry, one can conduct the measurement of water vapor H2 16O and H2 18O isotopes (Wang et al. 2014). For these purposes, Wang et al. (2014), proposed to use a field-deployable QCL-based spectrometer operated in the 7.12 μm region. The target lines shown in Figure 3.10 have been selected based on four criteria: (1) both lines should provide approximately equal peak absorption to allow probing both isotopes with significantly different natural abundance using the same optical path, (2) the lower state energies of both transitions should be within ±6.8 cm−1 to relax the requirement for the sample gas temperature stability to only ±1 K (assuming the targeted precision of the delta value is Δδ = 0.1‰), (3) minimal interference from other species present in the air, and (4) center frequencies of the selected lines should be within ∼1 cm−1 from each other to enable simultaneous access to both lines within a single current scan of a typical single-frequency distributed feedback (DFB) QCL. The system optical layout and functional block diagram of the water vapor isotope analyzer designed by Wang et al. (2014) are detailed in Figure 3.11. Experiment has shown that the achieved 1 s detection limits for H2 16O, H2 18O, and δ18O measurements using this instrument were 2.2 ppm, 7.0 ppb, and 0.25%, respectively. The Allan deviation analysis indicated that after 160 s averaging the detection limits achievable for H2 16O, H2 18O, and δ18O measurements can be 0.6 ppm, 1.7 ppb, and 0.05%, respectively.

Figure 3.10   The selected lines for H2 16O and H2 18O are: transition 514 ←541 at 1390.52 cm−1 and transition 432 ←541 at 1389.91 cm−1, respectively, in the ν2 rotational-vibrational band of H2O.

(Reprinted with permission from Wang, W.E. et al., Rev. Sci. Instrum., 85, 093103, 2014. Copyright 2014 by the American Institute of Physics.)

Figure 3.11   System schematic and optical layout. M, mirror; BS, beam splitter; QCL, quantum cascade laser; DCM, dichroic mirror; MCT detector, mercury cadmium telluride detector; OAPM, off-axis parabolic mirror; NI-DAQ, National Instruments data acquisition board.

(Reprinted with permission from Wang, W.E. et al., Rev. Sci. Instrum., 85, 093103, 2014. Copyright 2014 by the American Institute of Physics.)

One should note the measurement of water vapor isotopes is important for atmosphere monitoring. Measurements of water vapor isotopes (H2 16O and H2 18O) enable the fingerprinting of water fluxes in the urban environment, including the partitioning of latent heat flux between evaporation (bare soil and impervious surfaces) and the plant transpiration. Thus, stable isotope measurements can provide transformative observational capabilities for environmental monitoring and climatological assessments at regional and global scales. Furthermore, water vapor isotopes can be used to assess the role of spatial heterogeneities of the land surface composition on turbulent fluxes of water vapor and sensible heat. These capabilities are particularly important for examination of the water cycle in urban environments. Isotopic measurements are critical for hydrological cycle studies, the gas exchange and the transport processes between vegetation and the atmosphere, and understanding spatial and temporal variations (Gat 1996; Yakir and Sternberg 2000; Wen et al. 2008, 2012)

With regard to the disadvantages of these devices, they are listed in Table 3.4. In addition, the measurement is not flow dependant, but care must be taken to control the sample temperature and pressure. Changes in the sample temperature and pressure affect the line shape of the rotational–vibrational transitions and will cause changes in the instrument readings. Drawback to this technology is also the temperature sensitivity of the TDL. The junction temperature of the laser diode is critical to the measurement; any small changes in temperature of the diode will shift the center wavelength of the emission and can result in erroneous measurements or alarm conditions in the analyzer. While most manufacturers use conventional thermoelectric coolers and thermocouples to maintain a stable temperature of the laser diode, the use of on-board moisture reference cells has been used to provide the TDLAS system with line-lock capability and provide feedback to the temperature control loop of the TDL to increase reliability and confidence in the measurement. Part of the laser beam is passed through the reference cell assembly where the spectra of the analyte sample in the reference cell are monitored and any shift in the observed peak is used as a feedback signal for the temperature control of the tunable laser diode. A final disadvantage is the range and accuracy capability of the TDLAS device. Although it is acceptable to use these instruments in the traditional pipeline natural gas applications, their use for low-level detection of water vapor is limited due to the detection capability in a methane-based background. Specialized techniques such as differential spectroscopy add maintenance, cost, and potential errors to the measurement systems.

#### 3.4.1  General View

IR light is defined as electromagnetic radiation with a wavelength longer that in the range of visible light, that is λ > 800 nm. The IR range can be further divided into

 Near-infrared (NIR) λ = 780–3000 nm (0.78–3.0 μm) Middle-infrared (MIR) λ = 3000–8000 nm (3.0–8.0 μm) Far-infrared (FIR) λ > 8000 nm (>8.0 μm)

Range definitions vary throughout the literature and should be understood as a general guideline.

Studies have shown that in this region of the spectrum there are multiple absorption bands due to the interaction of radiation with different gases present in the atmosphere (Table 3.2), which makes this spectral range, particularly NIR and MIR, attractive for various gas analyzers, including optical hygrometers (Smith 1993). In the more distant area there is an absorption band of H2O at 1594.8 cm−1 (6.3 μm). Important advantage of IR hygrometers is a low cross-sensitivity to the temperature fluctuations and variations of gas density and particle contribution (e.g., dust and particles).

It should be noted that little difference in the position of the absorption bands can be observed in various sources. As is known, pressure, temperature, and variance in the composition of the sample gas can affect the position of these bands. For example, the absolute signal for water at the same volume ratio at two different pressures will be different due to the pressure broadening.

It should be understood that in addition to discrete absorption bands of water vapor, there is also a so-called continuous absorption spectrum, for example, in the 200–1200 cm−1, which is also attributed to the water vapor.

Continuum absorption by water vapor is defined as any observed absorption by water vapor, not attributable to the Lorentz line contribution within 25 cm−1 of each line. It has been suggested that it results from the accumulated absorption of the distant wings of lines in the FIR. This absorption is caused by collision broadening between H2O molecules (called self-broadening) and between H2O and nonabsorbing molecules (N2) (called foreign broadening). The most recent work suggested that the large portion of the continuum might be due to collision-induced transitions and does not relate to the line wings.

The presence of such amount of the strips forces to approach carefully when selecting the spectral range in which the hygrometer can operate. Of course, this choice should take into account the technical capabilities such as the presence of appropriate source and the detector of radiation, the presence of other components that may also interact with the radiation in selected range, the presence of free water, and expected percent of water vapor in the sample. In this case, the selected wavelength must satisfy such requirements as follows:

• High selectivity with regard to water in reflection, transmission, or the absorption properties
• High contrast in optical properties between water and the other components of the measured material

Comparison of all these factors indicates that the preferred spectral range for IR optical hygrometers is around the strong absorption peaks of water at λ = 1.47 and 1.94 μm (Figure 3.12). These wavelengths are also well separated from the absorption peaks of CO2 and CO, which are often present in a gas mixture. For the humidity measurement, one can also use the absorption bands near 2.7 μm and in the region of 5–7 μm (Figure 3.13) (Silver and Stanton 1987). In a short interval from 5.7 to 6.6 μm water-vapor absorption is very intense. Of course, the wavelengths in real hygrometers may differ slightly from the specified values indicated in Table 3.6 that is associated sometimes with a limited selection of lasers with the required wavelength. But we must understand that the use of wavelengths, not coinciding with the absorption band of water vapor will be accompanied by a reduction of hygrometer sensitivity. As a result, when developing the hygrometers, it is commonly used radiation in the near IR spectral range between 1.3 and 2.7 μm. It is also important to note that at present, for given spectral range they were developed the diode lasers (DL) with unique properties such as very high spectral resolution and spectral power density, continuous tunability in combination with excellent technical properties, that is, rather low cost, very low size, weight and power consumption, long laser life time, and good beam quality. That is why over the last years IR hygrometers operated in this spectral range had huge influence on the atmospheric hygrometry. There is also a large set of highly sensitive receivers for this spectral range, capable of working without cooling, which greatly simplifies the work of hygrometers.

Figure 3.12   Infrared absorption spectra of the air in near IR range.

(Data extracted from Wernecke, R. and Wernecke, J., Industrial Moisture and Humidity measurement: A Practical Guide, Wiley-VCH, Weinheim, Germany, 2014.)

Figure 3.13   Infrared absorption spectra of a number of relevant gas species in the area between 2 and 6 μm. This is a simplified graphic presentation to show the importance of the individual absorption lines, as well as the shadowing effect of water vapor.

(With kind permission from Springer Science+Business Media: Solid State Gas Sensor, 2009, Springer, New York, Comini, E. [Eds.] et al.)

### Table 3.6   The Wavelengths of Monochromatic Radiation Used in Developing IR Hygrometers

Wavelength

References

6.5 μm

2.59–2.7 μm

1.854–1.90 μm

1.32–1.4 μm

With regard to the preferred absorption band in this spectral region, Buchholz et al. (2013) believed that the absorption line near 1.37 μm is the most optimal. According to Buchholz et al. (2013), the absorption line near 1.37 μm should be selected in order to minimize cross-sensitivity to other gases or neighboring H2O lines and the temperature dependence of the line area. Another advantage of this line is excellent availability of the high-quality telecom-grade fiber optical components and the fiber-coupled laser modules, which minimizes size, weight, and cost of the instrument and allows effective suppression of parasitic water absorption in the air absorption paths outside the intended measurement cell.

Figure 3.14   Principle architecture of a NDIR gas-sensing microsystem.

(With kind permission from Springer Science+Business Media: Solid State Gas Sensor, 2009, Springer, New York, Comini, E. [Eds.] et al.)

### Table 3.7   Types of Infrared Radiation Sources

Type

Method

Material

Wavelength (μm)

Remark

Resistor heating by current flow

Tungsten

Infrared bulb

1–2.5

Long wavelength region is cut off by external bulb (glass). Secondary radiation is emitted trough the tube

Nichrome Kanthal

Electric heater

2–5

Silicon carbide (siliconate)

Globar

1–50

Constant voltage, large current

Ceramic

Nernst glower

1–50

Preheating is needed

Secondary heating by other power source

Metal (stainless steel, etc.)

Sheath heater

4–10

Ceramic

IRS type lamp

4–25

1–20

Heating by gas burning

Heating by discharge

Carbon

Carbon arc lamp

2–25

Causes some environmental problems such as soot

Gas discharge

Mercury

Cesium

Xenon

Mercury lamp

Xenon lamp

0.8–2.5

Long wavelength region is cut off by external bulb (glass). Secondary radiation is emitted trough the tube

Stimulated emission

Laser reaction

Carbon dioxide

Gallium arsenic compounds

CO2 laser

InGaAs laser

PbSnTe laser

9–11

1.1–1.5

1.2–6 to 7

But it should be recognized that the thermal emitters (blackbody radiation sources) are not optimal for use in hygrometers and their parameters are significantly inferior to LEDs, and especially LDs (Wilson et al. 1995). First, thermal emitters need a high operating temperature (up to 2000K) to obtain large radiation power in the MIR range, and this can cause unwanted heating in a compact instrument. But the additional emitted radiation can be accompanied by degradation of system performance; it warms up detectors and electronics elements and generates parasitic components of the detector signal (Cozzani et al. 2007). Second, their emitting spectrum covers from visible to FIR. Therefore, they emit a too wide spectrum of light and they are not efficient from a power consumption point of view. Third, the extremely broadband radiation produced by a blackbody requires optical filtering for selection required spectral range, and this often removes most of the optical power. Also, narrow-bandpass filters sometimes transmit light at wavelengths outside the bandpass region that are still detected. Fourth, the emitting area of a blackbody source is usually many times larger than that of a semiconductor light source, so that obtaining a uniform and collimated light beam can be difficult. Finally, blackbody sources tend to have a slower speed of response than semiconductor devices, and this makes them less suitable for applications where high-speed intensity or frequency modulation is required (Wilson et al. 1995).

If you look at Figure 3.15, it is seen that the modern lasers that can be used for the development of optical hygrometers that span the NIR and MIR spectral regions, with wavelengths ranging from 0.6 to >2 μm for conventional DL and from 2 to >4 μm for recently developed QCLs (Hvozdara et al. 2002; Kosterev and Tittel 2002). QCLs are semiconductor lasers based on transitions in a multiple-quantum-well heterostructure (Faist et al. 1994). The emission wavelength depends mainly on the quantum well thickness rather than on the size of the band gap, as is the case with conventional DL. This has allowed, using the same base semiconductors (InGaAs and AlInAs grown on InP), the manufacture of lasers with an emission wavelengths ranging from 3.5 to 17 μm (Werle et al. 2002). Thus, QCLs can operate at wavelengths in the MIR range, which match well with the fundamental vibrational absorption bands of water vapor and other chemical species—in comparison to conventional diode sources, where the laser emission generally matches the weaker overtone bands (Grouiez et al. 2008). In addition, due to specific mechanism of recombination, QCLs can operate at near room temperature and produce more optical power (milliwatts) of radiation within a broad range of frequencies. For comparison, lead-salt DL (PbSnTe and PbSnSe) designed for the 5–8 μm wavelength region can operate only at liquid nitrogen temperature. This greatly limits their application in the development of hygrometers, as the hygrometer in this case should have a cryogenic refrigeration diode laser system. But in this case, hygrometer is bulky and cannot be used for stand-alone operation.

Figure 3.15   Approximate spectral windows of commercially available room-temperature diode laser sources.

Of course, LEDs and LDs also have disadvantages. They exhibit strong dependency of the power and spectrum characteristics on the temperature and their price significantly exceeds a thermal sources costs. But their usage considerably improves the performance of hygrometers and therefore, these disadvantages do not limit the application of LEDs and LDs in the hygrometers. The gas and solid-state lasers (Figure 3.16) can be used in the same region of the spectrum. But the size and the cost of these lasers are inferior to semiconductor lasers, which considerably limit their application in the development of portable devices.

Figure 3.16   Spectral windows of commercially available gas and solid-state lasers.

(Data extracted from http://www.hamamatsu.com.)

Signal can be measured using heterodyne or direct detection systems. According to Grund et al. (1995), these systems have the following differences:

• Diffraction-limited optics are usually used in heterodyne LIDARs, and the aperture and focusing of these optics should be at least approximately matched to the strength and range of the return; the limited field of view reduces background light, but the direct detection systems can make use of larger optics.
• The heterodyne signal-to-noise ratio (SNR) is approximately constant out to long range and can be optimized at long range by focusing, whereas the direct detection SNR falls rapidly with range; consequently, very good differential absorption LIDAR (DIAL) results can be obtained using direct detection at short ranges, but the long range results are impaired.

Thus somewhat smaller optics can be used for the heterodyne systems and the results from the smaller heterodyne system are superior to those from the direct detection LIDAR at long range but inferior at short range (Grund et al. 1995).

With regard to PDs, suitable for use in IR optical hygrometer, at present time there is a large variety of semiconductor detectors and thermal detectors, which are able to cover the entire spectral range, suitable for the development of optical hygrometers (Tables 3.8 and 3.9 and Figure 3.17). Figure 3.17 shows the normalized detectivity characteristics for different types of detectors. Quantum-type semiconductor detectors show excellent detectivity in the IR range and fast response, but they are strongly wavelength dependent (Figure 3.17) and, in addition, they generally need to be cooled for accurate measurement. In thermal detectors such as pyroelectric sensors, bolometers, and thermopiles the absorbed photon results in a temperature rise of the detector, which entails an alteration of its electrical properties.

Figure 3.17   Normalized detectivity characteristics for different type of detectors.

### Table 3.8   Thermal Detectors

Detector Types

Method of Operation

Bolometer

Metal

Semiconductor

Superconductor

Ferroelectric

Hot electron

Change in electrical conductivity

Thermocouple/Thermopile

Voltage generation, caused by change in temperature of the junction of two dissimilar materials

Pyroelectric

Changes in spontaneous electrical polarization

Golay cell/Gas microphone

Thermal expansion of a gas

### Table 3.9   Types of Infrared Detectors and Their Characteristics

Type

Detector

Spectral Response (μm)

Operating Temperature (K)

D* (cm·Hz1/2/W)

Thermal type

Thermocouple (Thermopile)

Bolometer

Pneumatic cell

Pyroelectric detector

Golay cell, condenser-microphone

PZT, TGS, LiTaO3

Dependens on window material

300

300

300

300

D* (λ, 10, 1) = 6·108

D* (λ, 10, 1) = 1·108

D* (λ, 10, 1) = 1·109

D* (λ, 10, 1) = 2·109

Quantum type

Intrinsic type

Photoconductivity type

PbS

PbSe

InSb

HgCdTe

1–3.6

1.5–5.8

2–6

2–16

300

300

213

77

D* (500,600,1) = 1·109

D* (500,600,1) = 1·108

D* (500,1200,1) = 2·109

D* (500,1000,1) = 2·1010

Photovoltaic type

Ge

InGaAs

Ex. InGaAs

InAs

InSb

HgCdTe

0.8–1.8

0.7–1.7

1.2–2.55

1–3.1

1–5.5

2–16

300

300

253

77

77

77

D* (λp) = 1·1011

D* (λp) = 5·1012

D* (λp) = 2·1011

D* (500,1200,1) = 2·109

D* (500,1200,1) = 2·1010

D* (500,1000,1) = 1·1010

Extrinsic type

Ge:Au

Ge:Hg

Ge:Cu

Ge:Zn

Si:Ga

Si:As

1–10

2–14

2–30

2–40

1–17

1–23

77

4.2

4.2

4.2

4.2

4.2

D* (500,900,1) = 1·1011

D* (500,900,1) = 8·109

D* (500,900,1) = 5·109

D* (500,900,1) = 5·109

D* (500,900,1) = 5·109

D* (500,900,1) = 5·109

Despite of its smaller detectivity compared to quantum detectors (Figure 3.17), thermal detectors are widely used due to its uncooled operation, their small dependency on the wavelength, and the flat response. At that thermopiles show better sensitivity than pyroelectric and bolometer detectors (Schilz 2000). Thermopile detectors fabricated by PerkinElmer are shown in Figure 3.18. Modern thermopiles fabricated in micromachining technology are extremely sensitive, with a fast response time due to its small size (Graf et al. 2007; Fonollosa et al. 2009). One should note that nowadays, silicon-based thermopiles are commonly used for IR detection in NDIR spectrometers.

Figure 3.18   (a) Photo of PerkinElmer dual-thermopile detectors. The left side shows open detectors. Clearly the two thermopiles can be seen. The gray-colored squares in the middle of each sensor chip are the absorber areas, which collect the IR light to be measured. The small cube near to the lower chip is the thermistor, which senses the ambient temperature. The right side shows the detectors covered by a cap holding the two different IR-filters. . (b) Schematics of the detection unit in a TO8 package, with thermopile elements fabricated using micromachining technology, optical filters, and solder joints.

(From http://www.perkinelmer.com.)(Reprinted from Sens. Actuators A 149, Fonollosa, J. et al., Limits to the integration of filters and lenses on thermoelectric IR detectors by flip–chip techniques, 65–73, Copyright 2009, with permission from Elsevier.)

Filters are also important elements of the optical hygrometer for improving the selectivity of the specific optical narrowband filters. They are usually placed directly upon the IR detector set. A bandpass optical filter usually consists of a number of dielectric layers on a substrate. The thickness, the number and the material of deposited layers on the substrate determine the transmission characteristics of the filter (Rancourt 1996). Usually, silicon is used as a substrate to take advantage of its related technologies. Systems with narrowband but tunable optical filters are also designed (Santander et al. 2005).

As it is known, the sensitivity of optical instruments is dependent on the absorption path length. For example, according to the Lambert–Beer law, the transmission of the IR radiation through the absorbing gas is inversely proportional to the exponential function of gas concentration and the path length. This means that for achievement of high sensitivity in IR spectral range, the optical path should be long. Therefore, the measuring path in the IR optical hygrometers is normally greater than 1 m. It is understandable that so long optical path lengths are often undesirable or simply not possible in industrial applications. Therefore, instead of single transmission, compact cells with multiple reflections inside the cell are incorporated in the device (e.g., Herriot measurement cells) (Figure 3.19). At present, Herriot cells (Herriot et al. 1964; Herriot and Schulte 1965) have been widely applied in atmospheric monitoring and are commercially available with astigmatic optics providing more than 100 m of nonoverlapping optical path length in a physical cell less than 1 m in length. Liu (2012) in his experiments, has used a multipass absorption cell with physical length about 35 cm, which provided an optical absorption path length exceeded 30 m. Thus, using this approach, the optical path length remains constant, while the cell dimensions can be significantly reduced. For comparison, the measuring path in the UV optical hygrometer is typically a few centimeters in length.

Figure 3.19   Diagram showing the principle of Herriot cell operation.

Of course, one can use a simplified version shown in Figure 3.20, when open-path is used and the dual emission pass is achieved by using a separated reflector (Seidel et al. 2012). The same approach was used in developing the hygrometer by Wilson et al. (1995). However, such an approach can only be used for stationary devices. An additional difficulty of this approach is that to achieve the required accuracy it is necessary to ensure that any target (reflector) movement such as rotation, translation, or tilt of the target should have a negligible influence on the accuracy and, if possible, on the precision of the spectrometer. In addition, the performance of the spectrometer should not be disturbed if there is a small deviation from the precise perpendicular target positioning. There should not be any dependence of the exact laser position on the target, which is not only important for adjustment, but also for future scanning applications for spatially resolved measurements. Also, for high sensitivity it is still necessary to use sufficiently long optical path lengths. However, the more the optical path lengths are, the more difficult is to perform the above listed requirements. Seidel et al. (2012) in their hygrometer used the reflection tape being in a distance of 75 cm to 1 m.

Figure 3.20   Optical setup of the spectrometer with sending and detecting side (left) and separate scattering target (right). Here the light of a fiber-coupled DFB-laser at 1370 nm was collimated and directed through a 50/50 dielectric beam splitter onto the target, where it was partially reflected/scattered. The beam-splitter serves for directing the part of the reflected light onto a photodiode placed outside the incoming beam-path. A micro prismatic reflection tape was used as reflector (scattering target).

(With kind permission from Springer Science+Business Media: Appl. Phys. B, TDLAS–based open–path laser hygrometer using simple reflective foils as scattering targets, 109, 2012, 497–504, Seidel, A. et al.)

In order to reduce errors due to fluctuations of the source, temperature, and so on, a second cell is often utilized for reference (Silver and Hovde 1994; Amerov et al. 2007). The measurement signal and reference signal can then be combined to give a measurement value. As for errors and anomalies observed at measuring in the MIR region, they are discussed in detail by Chalmers (2002).

#### 3.4.2  Realization

Usually modern optical IR hygrometers are being developed using two approaches: (1) transmission of radiation at two wavelengths; one wavelength corresponds to the water absorption line and second wavelength is used as a reference (Foskett et al. 1953; Chen and Mitsuta 1967), and (2) TDLAS (Amerov et al. 2007; Soleyn 2009; Buchholz and Ebert 2013; Nwaboh et al. 2017). The advantages of these approaches were considered earlier in Section 3.3. Importantly, that multichannel approach can be also used in TDLAS hygrometers. For example, to achieve an unprecedented dynamic range of 1–40,000 ppm, Ebert and Buchholz (2016) have used in his hygrometer simultaneously two lasers at 1.4 and 2.6 μm for high/low H2O concentrations.

Successful representative of optical hydrometers, developed when using the first approach, is LI-7500A Open Path CO2/H2O Gas Analyzer, fabricated by LI-COR Inc. (http://www.licor.com) and shown in Figure 3.21. The LI-7500A analyzer is a high speed, high precision, NDIR gas analyzer that accurately measures densities of carbon dioxide and water vapor in the turbulent air structures. The LI-7500A sensor head has a 12.5 cm open optical path, with single pass optics and a large 8 mm diameter optical beam. Optical filters centered at 3.95 μm provide a reference signal for CO2 and water vapor. Absorption at wavelengths centered at 4.26 and 2.59 μm provide for measurements of CO2 (0–3000 ppm) and water vapour (0–60,000 ppm), respectively with accuracy ±1–2%. This design minimizes sensitivity to dirt and dust, which can be accumulated during normal operation. One should note that LI-COR Inc. designed also CO2/H2O Gas Analyzer, LI-7200, with enclosed optical cell (Figure 3.21b). This analyzer has the optical scheme similar to the optical scheme used in LI-750A.

Figure 3.21   CO2/H2O Gas Analyzers designed by LI-COR Inc.: (a) LI-7500A analyzer with open path and (b) LI-7200 with optical cell.

As regards the second approach, here we have a large number of successful projects (Amerov et al. 2007; Soleyn 2009; Buchholz and Ebert 2013; Nwaboh et al. 2017). For example, Nwaboh et al. (2017) suggested TDLAL hygrometer for absolute measurements of H2O in methane, ethane, propane, and a low CO2 natural gas. The sensor was operated with a 2.7 µm DFB laser, equipped with a high pressure single pass gas cell, and used to measure H2O amount of the substance fractions in the range of 0.31–25,000 µmol/mol. The operating total gas pressures were up to 5000 hPa. The relative reproducibility of H2O amount of substance fraction measurements at 87 µmol/mol was 0.26% (0.23 µmol/mol). The maximum precision of the sensor was determined using a H2O in methane mixture, and found to be 40 nmol/mol for a time resolution of 100 s. This corresponds to a normalized detection limit of 330 nmol mol−1·m Hz−1/2. The relative combined uncertainty of the H2O amount fraction measurements delivered by the sensor was 1.2%.

In the same spectral range the Pico-SDLA H2O (hereafter Pico-SDLA) hydrometer is working (Durry et al. 2008; Ghysels et al. 2016). Pico-SDLA hydrometer is a lightweight spectrometer which measures the water vapor using laser absorption spectroscopy. Hygrometer was equipped with the probe diode laser emitted at a wavelength of 2.63 μm and had a 1 m path length through ambient air. The water vapor absorption line was scanned by tuning the laser current and fixing the TEC temperature. As it was indicated before, the current modulation of the laser is the preferred method to scan the water vapor absorption line since the response time is much faster than for temperature modulation. After passing through the ambient-air sample, the laser beam was focused onto an indium arsenide detector using a sapphire lens. A feature of the hygrometer was that two different rotation–vibration absorption transitions of water vapor were probed during measurements. This transition was needed because of the large variation in mixing ratio occurring between the troposphere and the stratosphere. For measurements from the ground to around 200 hPa pressure level, the 413 ← 414H16 2O line at 3802.96561 cm−1 was used. While above 200 hPa pressure level, we use the 202 ← 101H16 2O line at 3801.41863 cm−1. During in-flight measurements, the switch from one line to the other is automatically driven. Both sets of line parameters were obtained from HITRAN 2012 database (Rothman et al. 2013). The mass of the Pico-SDLA was less than 9 kg, making it suitable as a payload for small stratospheric balloons (500 and 1500 m3). The hygrometer was able to measure the water vapor from the ground to altitudes of 35 km for concentrations ranging from 15,000 to less than 1 ppmv (Ghysels et al. 2016). The mixing ratio was extracted from the measured spectra using a nonlinear least squares fitting algorithm applied to the measured line shape. The authors of Ghysels et al. (2016) used the Beer–Lambert law to model the spectrum and used a Voigt profile (VP) to describe the molecular line shape.

Amerov et al. (2007) and AMETEK (http://www.ametekpi.com) have chosen the water vapor line at 1854 nm for the measurements. They proposed to use of an all-digital protocol for the modulation of the laser drive signal and the demodulation of the detector response. A reference cell used in parallel with the sample cell allowed to continuously validating the performance of the system. It was shown that the TDLAS hygrometer designed by Amerov et al. (2007) had a limit of detection of 5 ppmv for both the nitrogen and natural gas samples. Hygrometer can operate in the range of 5–2500 ppmv. The accuracy of the reading of hygrometer fabricated by AMETEK is <1 ppb by volume. Michel Instruments (Figure 3.22) reported that their hygrometers had limit detection ~1 ppmv. Liu (2012) has shown that by reducing the sample gas pressure, the spectral interference with background gas could be greatly reduced and effectively removed in real time, and therefore, enhanced specificity, improved accuracy, lower detection limit, faster response, and lower cost of ownership all become achievable. Liu (2012) demonstrated that while using this technique a detection limit approaching 10 ppbv, with accuracy around 5 ppbv, and the response time in minutes could be achieved. However, it should be noted that Liu (2012) in his experiments used a multipass absorption cell, large enough; its physical length was about 35 cm only but absorption path length was extended to approximately 30 m. The IR TDL absorption hygrometer, designed by Edwards et al. (2000), had the same accuracy, ~5 ppbv. For achievement such accuracy they used two-tone frequency modulation spectroscopy (TTFMS), a multipass absorption cell and a computer controlled diode laser operated at λ = 1.393 μm. Without using a multipass absorption cell, Hydrometers, as a rule, cannot provide such high sensitivity and accuracy. For example, 1.4 μm-TDLAS hygrometer with an internal optical cell with 1.5 m optical path length, elaborated by Buchholz and Ebert (2013) for airbone applications, had a precision of 33 ppbv.

Figure 3.22   The OptiPEAK TDL600 Tuneable Diode Laser Analyzer. The next generation TDLAS Analyzer for automatic online measurement of the moisture in variable compositions of natural gas and biomethane desihned by Michel Instruments. Range of 1–1000 ppmv. Limit detection ~1 ppmv.

Electrical and optical scheme of hygrometer designed by Amerov et al. (2007) is shown in Figure 3.23. A DFB laser, which was used as the source of light, produced an optical power of approximately 3 mW, when operated at the target wavelength for the water–vapor measurement. Output from the laser was coupled into a single-mode fiber, which was connected to a fiber-optic splitter, used to divide the optical power in a 70/30 ratio and simultaneously connect the DFB laser to the sample and reference cells. Signals from the InGaAs 0.5 mm2 photodiode detectors were input to separate channels of the electronics unit. It was now possible to make simultaneous measurement of unknown samples and a known moisture reference, which was used to lock the output wavelength of the laser to the 1854 nm absorption line of water vapor. Further, the digital signal process methods employed in this system can successfully remove minor background interferences, caused by other species in the sample.

Figure 3.23   Electrical and optical scheme of TDLAS hygrometer operated at 1.854 μm.

(Data extracted from Soleyn, K., Proceedings of the 5th International Gas Analysis Symposium, GAS 2009, Rotterdam, the Netherlands, February 11–13, 2009.)

Figure 3.24 is a schematic that illustrates the components of a TDLAS hygrometer designed by Soleyn (2009). This hygrometer was adapted for its use in natural gas. The TDLAS hygrometer has very good correlation to the reference standard. The accuracy of the hygrometer exceeds ±4 ppmv in the range of 5–200 ppmv, and ±2% of reading in the range of 200–5000 ppmv. For use in natural gas the components of hygrometer, which can be wetted, are constructed of stainless steel with the exception of the optical window that consists of proprietary glass and the mirror that consists of proprietary polished metal alloy. Those components are selected for their resistance to corrosion and optical purity. It is known that the path length is related to a lower detection limit. Therefore, the path length in the hygrometer was optimized for a volume ratio of approximately 5–5000 ppmv. The photodiode and reference photodiode are housed in a hermetically sealed and dry enclosure. A platinum-resistance temperature detector (PRTD) measures the gas temperature and a silicon micromachined strain gauge pressure sensor measures the sample pressure. The temperature sensor is encased in a stainless steel sheath and the pressure sensor is also constructed of stainless steel with a hastelloy–wetted diaphragm. For field installation, a pipeline-insertion membrane filter and the pressure regulator will separate liquids (hydrocarbons, liquid water, and glycol carry over from the dehydration process) and drop them back into the pipeline and also reduce the pressure. A second pressure regulator will decrease the pressure close to atmospheric. The sampling system and absorption cell are installed in a stainless steel enclosure that is heated by a thermostatically controlled electrical resistance heater.

Figure 3.24   Diagram illustrating TDLAS hygrometer designed by Soleyn K.

(Data extracted from Soleyn, K., Proceedings of the 5th International Gas Analysis Symposium, GAS 2009, Rotterdam, the Netherlands, February 11–13, 2009.)

At present, many companies fabricate IR hygrometers and these hygrometers present on the market. For example, General Electric Company developed Aurora Moisture Analyzer to measure moisture in natural gas. Sampling system of this tool is shown in Figure 3.25. Aurora is equipped with a two-stage turnkey sampling system. An optional first stage consists of a membrane filter/regulator installed directly in the pipeline. It prevents any liquid (hydrocarbon, glycol, or water in liquid phase) from entering the sample line. The pipeline pressure is reduced by means of a regulator. As the gas enters the second stage it flows through a coalescing filter, and a pressure regulator further reduces the pressure. The flow rate is adjusted with a needle valve. Only clean low-pressure gas enters the absorption cell. An optional heater may be installed in the enclosure for application in cold climates. The heater also serves to keep the sample in the gas phase. The laser-based measurement system provides very fast response. The optical response is <2 s. The system is designed to operate continuously for many years with unsurpassed reliability. Factory service or calibration is recommended on a five-year interval.

Figure 3.25   Sampling system (a) and measurement cell (b) of Aurora Moisture Analyzer.

(From http://www.gesensinginspection.com)(Data extracted from Soleyn, K., Proceedings of the 5th International Gas Analysis Symposium, GAS 2009, Rotterdam, the Netherlands, February 11–13, 2009.)

Examples of other hygrometers are listed in Table 3.10. Most of these hygrometers use the TDLAS technology. As it is seen in the table, generally, these devices besides the water vapor control the content of CO2 in the atmosphere. For these purposes, it is commonly used a dual laser TDLAS setup which consists of two TDLs, tuned on the characteristic absorption lines of the analytes in the sample. Optical scheme of the 5100 HD analyzer fabricated by AMETEK Process Instruments and intended for detecting H2O and CO2 is shown in Figure 3.26.

Figure 3.26   Optical scheme of the 5100 HD analyzer fabricated by AMETEK Process Instruments. The Model 5100 HD can be configured with single/dual lasers and single/dual gas cells. In a two-laser two-cell configuration, two different gas streams can be simultaneously monitored. One specific application is the measurement of wet and dry natural gas streams for H2O levels. This dual cell configuration is superior to stream switching with a single gas cell which would require long equilibration periods when switching from wet to dry streams. Dual lasers systems can also be used for the monitoring of two gas species at the same time.

### Table 3.10   Examples of Commercial IR Hygrometers

Sensor Type

Tested Gases

Produces

Country

LI 6262 (closed-path)

H2O, CO2

LI-COR Inc.

USA

LI 7500 (open-path)

H2O, CO2

LI-COR Inc.

USA

IR-3000 moisture transmitter

H2O

MoistTech. Corp.

USA

OP-2 (open-path)

H2O, CO2

ATC Bioscientific Ltd.

UK

DF-700 moisture analyzer

H2O

Servomex

UK

Ophir IR-2000

(open-path, very large)

H2O

Ophir Corporation

USA

Gas Analyzer E-009

(open-path, very large)

H2O, CO2

Japan

5100/5100 HD analyzer

H2O, CO2

AMETEK Process Instruments

USA

OptiPEAK TDL600 analyser

H2O, gases

Michel Instruments

UK

GE’s Aurora analyzer

H2O

General Electric Company

USA

WVSS-II

H2O

SpectraSensors Inc.

USA

Optical hygrometers can operate at short-wave spectral region (1.0–0.7 μm). Studies have shown that in this spectral region there were also observed an absorption bands associated with the presence of water vapor. One of the examples of hygrometers working in this spectral region is a hydrometer, developed by Kebabian et al. (2002). Optical scheme of this hygrometer is shown in Figure 3.27. The hydrometer uses an argon emission line at 935.4 nm that can be Zeeman-split into two components. When the emission line is Zeeman-split by a longitudinal magnetic field of the proper strength, it is divided into one component that is only weakly absorbed, λ1, and a second component that is strongly absorbed by the water vapor line, λ2. This fact forms the basis for a differential absorption measurement. The major advantages of the gas discharge lamp, compared to the use of alternative light sources (e.g., NIR DL) are the stability of the spectral properties of the light source, which results in a relatively simple and inexpensive instrument. Moreover, the light source is reproducible at any time, based on its key physical dimensions (capillary diameter and length), gas fill pressure, and magnetic field. It also offers the advantages in its combination of time response and the ability to function properly in the presence of liquid water that cannot ordinarily be achieved by other types of sensors including the chilled mirror hygrometers and capacitance-based sensors. Kebabian et al. (2002) shows that this hygrometer is acceptable for in situ measurements of fine structure (over tens of meters distance) in the water vapor concentration profile both in free air and within and in the vicinity of clouds that could not be obtained using the commonly utilized technologies.

Figure 3.27   Schematic of optical water vapor hygrometer operated at 935.4 nm with all its optical elements.

(From Kebabian, P. et al., J. Geophys. Res. Atmosph., 107, 4670, 2002. With permission of American Geophysical Union.)

An improved model of optical hygrometer, working in this range, has been developed by Matthey et al. (2006). They used the same spectral range, 935 nm. However, as a source of radiation, they have used low-power continuous-wave DL. In addition, they used four single-mode frequency-stabilized optical signals. Three lasers were locked to three water–vapor absorption lines of different strengths (strong [935.4283 nm], medium [9935.3049 nm], and weak [935.6501 nm]), whereas the fourth lied outside any absorption line (935.600 nm). On-line stabilization was performed by wavelength-modulation spectroscopy using a compact water-vapor reference cells. An offset-locking technique implemented around an electrical filter was applied for the stabilization of the off-line slave laser to an online master laser. Matthey et al. (2006) believe that such configuration is promising for the in situ water vapor monitoring in the atmosphere at various altitudes. Strongly absorbing water vapor lines can be used for higher altitudes, whereas weakly absorbing lines can be served for lower altitudes, which show high water-vapor concentrations. The simultaneous use of four different wavelengths (three on-line and one off-line) enables measurement of H2O profiles across the entire altitude ranging from the lower stratosphere to the boundary layer. The specific wavelengths must be selected depending on the atmospheric process under study, the desired altitude resolution and coverage and the required measurement precision. In particular, such approach can be promising for development of LIDAR systems discussed in Chapter 8 (Wulfmeyer and Walther 2001).

One should note that besides ground-based instruments there are TDL spectrometers, developed for aircraft application (Table 3.11).

### Table 3.11   Tuneable Diode Laser (TDL) Spectrometers Designed for Aircraft Application

Device

Разработчик

References

JPL open-path tuneable diode laser spectrometers

Jet Propulsion Laboratory

The LaRC/ARC diode laser hygrometer (DLH)

NASAs Langley and Ames Research Centers (LaRC and ARC)

#### 3.5  UV Optical Hygrometers

UV spectral range is also of interest for optical hygrometers development. In this spectral range there are several emission lines, which are of particular significance in the humidity measurement due to their high selectivity to water. These are, for example, the Lyman–alpha line and the line of krypton (123.6 nm). Lyman–alpha radiation λ(L α) is emitted by hydrogen atoms H(2P–2S) at a narrow line in the far UV portion of the spectrum at λ = 121.56 nm (Figure 3.28). It is produced by an electrical discharge in hydrogen. The wavelength of the Lyman–alpha line coincides with a deep absorption minimum in transmission spectra of air due to the presence of water vapor.

Figure 3.28   Output spectrum of the light source between 115 and 135 nm, which corresponds to the response interval of the NO ionization chamber. The spectrum was recorded with a PMT Hamamatsu 1259 (CsI photocathode and MgF2 window). The resolution is 0.15 nm. Measurements were carried out in a 4.8-cm long absorption cell with dry air. At Lyman–alpha the peak signal is 6 × 105 counts/s.

(From A. Zuber and G. Witt, Appl. Opt., 26, 3083–3089, 1987. With permission of Optical Society of America.)

In principle, any wavelength satisfying the requirement, that there be large difference between the absorption cross section of the water vapor and the other constituents, can be used for humidity control. The advantage of using hydrogen Lyman–alpha line is that at this wavelength the difference in the cross sections amounts is almost three orders of magnitude. For example, the dominant absorber of radiation in the vacuum UV region (below 185 nm) is molecular oxygen with a pressure-dependent absorption cross section at λ = 121.6 nm of ~2.3 × 10−20 at STP (Kley 1984), whereas the absorption cross section for water vapor at the same wavelength is ~1.57–1.59 × 10−17 cm2 (Kley 1984). These data show that while for Lyman–alpha light water vapor the absorption is very strong, the oxygen absorption is uniquely low, and most other common gases are relatively transparent, for example, nitrogen. In addition, oxygen effects can be removed using measurements of temperature and pressure to calculate the fractional oxygen density. Ozone absorption does represent a true interferent but only becomes significant at extremely high altitudes (stratosphere) where the natural ozone concentrations routinely reach significant levels. Thus, even the trace amounts of water vapor in a sample of the air dominate the absorption of this radiation. Therefore, the most commonly used source of UV radiation in hygrometers is the emission of hydrogen gas which includes the Lyman–alpha line at 121.6 nm (Buck 1976a, 1985; Beaton and Spowart 2012). In addition, the Lyman–alpha light sources have high output and spectral purity (Zuber and Witt 1987).

Lyman–alpha lamp is a direct current (DC)-powered hydrogen discharge tube with a temperature-controlled hydrogen reservoir made from a mixture of uranium and uranium hydride (Dieke and Cunningham 1952). Powdered uranium and a certain quantity of uranium hydride are contained in a side arm. The side arm can be heated and maintained at a constant temperature. Thus, the uranium hydride serves as a source of H2, whereas the uranium simultaneously acts as an effective chemical absorber of gaseous contaminants that arise during the discharge. Such configuration of the lamp permits a low operating H2 pressure for good spectral purity of the Lyman–alpha emission line, whereas the hydride restores the H2 lost to the electrodes so as to give an acceptable lifetime. One should note that the Lyman–alpha hygrometer was developed at the beginning of the 1970s almost in parallel in the United States, the Soviet Union, and the former GDR (Buck 1973; Kretschmer and Karpovitsch 1973; Martini et al. 1973); and the American instrument was commercially produced by AIR Inc., Boulder CO. Other examples are listed in Table 3.12. Some of devices indicated in this table are no longer manufactured.

### Table 3.12   Commercial UV Hygrometers

Sensor Type

Produces

Country

L-5V

Buck Research

USA

Lyman–alpha hygrometer

MUERIJ METEO

NL

Lyman–alpha hygrometer

Wittich and Visser

NL

AIR-LA-1 Lyman–alpha hygrometer

AIR

Now: Vaisala

USA

Finland

KH2O–Krypton hygrometer

Campbell Sci.

USA

Source: Foken, T. Micro–Meteorology. Springer, Berlin, Germany, 2008.

The embodiment of the UV hygrometer, using absorption at λ(L α) = 121.6 nm is shown in Figure 3.29. As in the IR hygrometers, the measurement with UV radiation is based on the detection of the transmitted or absorbed fraction of light after propagation through the sample. UV hygrometer shown in Figure 3.30 was developed by Beaton and Spowart (2012) for the airborne measurement of atmospheric water vapor. It is established that a significant fraction of radiation is absorbed over a few millimeters path length under normal conditions, which helps to minimize the adsorption surface and increase the exchange rate. On account of the extremely wide variation of the water vapor concentration between the lower and upper troposphere, Beaton and Spowart (2012) decided that it was necessary to adjust the sample path length to optimize the signal dynamic range for the humidity and the air density of primary interest. It is known that if the path length is too short, the mixing ratio noise will be excessive at low humidity. If the path length is too long, there will be reduced sensitivity at high humidity resulting from non-Lyman–alpha light dominating in the signal. Therefore, the construction used in this hygrometer allowed varying the sample path length from 2.4 mm up to a useful maximum of 10 mm by placing spacers between the sample housing and the detector housing. The sample volume is a short cylinder 19 mm in diameter, giving a volume of 0.7 cm3 when the path length is set at its minimum 2.4 mm.

Figure 3.29   The UV hygrometer mounted to a C-130 aperture plate designed for aircraft applications. A radio-frequency (RF)-excited Lyman–alpha lamp (HHeLM-L) from Resonance, Ltd., which contains a temperature-controlled hydrogen sponge for improved spectral purity along with long life was used. The power supply box (15 cm × 15 cm × 5 cm) is not shown. The ruler is 30 cm long. The total cost was approximately U.S.\$15,000 dollars (USD).

(From Beaton, S.P. and Spowart, M., J. Atmos. Oceanic. Technol. 29, 1295–1303, 2012. With permission of American Meteorological Society.)

Figure 3.30   Components of a Lyman–alpha hygrometer.

In the first UV Lyman–alpha light hygrometers usually a nitric oxide ionization cell served as the detector. A photoionization chamber is a simple device containing an ionizing gas and two electrodes. Incoming radiation ionizes the gas (nitric oxide in this application), and an electric field maintained between the electrodes induces the electron and ion drift. The resultant current is proportional to the incident light intensity. Radiation enters the detector through another magnesium fluoride window aligned with the source across the sampling volume. The combination of the magnesium fluoride windows and nitric oxide as the ionization medium limit the response of the detector to incident wave lengths between 132 and 115 nm. This effectively filters almost everything except the Lyman–alpha line. Beaton and Spowart (2012) have shown that a solar-blind diamond photocathode vacuum photodiode (R6800U-26, Hamamatsu) with a spectral response of 115–220 nm could be also used. The light level is high enough that photomultiplier tubes are not needed, and the photodiodes are more compact and have lower microphonics and magnetic field sensitivity than photomultiplier tubes. In this case a highly stable high-voltage power supply is not needed, and in fact the photodiodes are rather insensitive to power supply voltage once it exceeds 10 V. In addition, the vacuum photodiodes have an unlimited shelf life and are mechanically and electrically robust. Their drawbacks compared to the NO ionization cell are lower quantum efficiency, ~10% versus ~60% for the NO cell (Carver and Mitchell 1964), and a smaller active area. There are also devices where a photomultiplier connected to a low‐noise electrometer is used as detector of Lyman–alpha radiation (Mestayer et al. 1986, 1987).

Simulations have shown that the Lyman–alpha absorption hygrometer can offer a very fast response (~5 ms). It is great advantage of UV hygrometers: the fast response makes it a suitable instrument for ultrafast hygrometry (sample rates of 10–100 Hz), the water vapor flux measurement, or micrometeorological measurements inside and outside clouds. Of course, the time of response indicated before is theoretical limit. However, even real instruments have response time smaller than 1 s. It was also found that this device can measure the water vapor densities of 0.1–25 g/m3 with a relative precision of 0.2% and an accuracy of 5%.

It is important for applications that in contrast to IR, the optical pathway required for measurement in UV range is much shorter that required for IR light. Therefore, the UV optical hygrometers can be directly embedded in the pipes where the humidity control is needed. For example, Figure 3.31 depicts a measurement device integrated into a gas flow. The source and detector are aligned perpendicularly to the direction of gas flow. The distance between the source and the detector L is variable and should be suited to the expected water content of the gas to be measured. L can be calculated using the Lambert–Beer law, and is a function of the absolute water content and the effective absorption cross section. In real devices an optical path length can vary from several millimeters to 20 mm or more. The incident beam enters the sample chamber through the entrance window and penetrates the gas to be measured. The typical window materials are the quartz glass and magnesium fluoride.

Figure 3.31   UV hygrometer incorporated in the gas pipeline.

(Idea from Wernecke, R. and Wernecke, J., Industrial Moisture and Humidity measurement: A Practical Guide, Wiley-VCH, Weinheim, Germany, 2014. With permission.)

However, the Lyman–alpha absorption hygrometer is a secondary measurement device and therefore it is not possible to use the single-beam Lyman–alpha absorption hygrometer alone to measure the absolute water vapor density (Schanot 1987). This means that this device must be regularly calibrated, and for this purpose an additional absolute water vapor sensor, such as a chilled mirror hygrometer, is needed. Therefore, data received using such devices are typically corrected by the means of data from a second, slow response hygrometer, and application of a suitable algorithm. Various correction schemes have been developed to make the data useful. But, of course, this situation is not comfortable for users. The Lyman–alpha light source aging, temperature drift, and optical window contamination are other factors that prevent a stable predictable calibration. One should note that the devices working in the UV range are affected more than those in the IR range. This means that the calibration characteristics of UV devices are subjected to large changes during the application time. Therefore, the Lyman–alpha absorption hygrometer usually is used together with more stable absolute hygrometer such as a dew point hygrometer. When a reference hygrometer is not available, a variable path length self-calibration technique in the case of a constant absolute humidity can be used. A first Lyman–alpha hygrometer with variable path length was proposed by Buck (1976a) and an updated version was proposed by Foken et al. (1995). In addition, for quantitative analysis of the concentration C of the absorbing species in single beam devices both I and I o must be measured simultaneously. Temporal changes in the output from the lamp seriously affect the accuracy of the measurement at small optical depths as well (Zuber and Witt 1987). Other complications are the possible presence of alternate absorbers with pressure- or temperature-dependent absorption cross sections as well as the requirement of a parallel beam which cannot be met without collimating optics. Zuber and Witt (1987) and Weinheimer and Schwiesow (1992) have shown that many indicated above problems could be resolved using a two-beam optical scheme (Figure 3.32). In this hygrometer, the absorption is determined along two separate paths containing the sample and the dry reference air, respectively, hence eliminating undesired absorption by O2 and other gases presented in the air. At the same time, the differential absorption eliminates problems related to the intensity variations in the light source output. The dynamic range of the instrument can be tuned by diluting the sample with reference air. The sample and the reference air are admitted to their respective chambers with the aid of a gas handling system. Weinheimer and Schwiesow (1992) also proposed to use the second wavelength that is not strongly absorbed by water vapor to control the degradation of windows. Weinheimer and Schwiesow (1992) believe that solids or liquids in the path have broadband absorption features and are essentially opaque at UV wavelengths, so absorption at both primary and secondary wavelengths should be approximately the same for closely spaced cells and wavelengths.

Figure 3.32   Schematic drawing of the optical scheme of two-beam UV hygrometer, which include light source (the hydrogen discharge lamp emitting the Lyman–alpha [121.6 nm]), absorption paths for sample and reference air, and no filled ionization chamber detectors. The output from the two detectors is passed to electronics units.

(Idea from Zuber, A. and Witt, G., Appl. Opt., 26, 3083–3089, 1987; Weinheimer, A.J. and Schwiesow, R.L., J. Atmos. Oceanic Technol., 9, 407–419, 1992.)

Further critical disadvantages of a Lyman–alpha absorption hygrometer are a long-term instability of the UV hydrogen lamp, long-term intensity fluctuations, and degradation of the MgF2 windows via solarization, which requires a frequent calibration. It is in many cases hampered the use of UV hygrometers and forced many producers to abandon a production of Lyman–alpha absorption hygrometers. In addition, these lamps were mainly handmade. Therefore, there is the requirement for careful selection of the light source before installing it in the gauge, since for the lamp there is a considerable scatter of the illumination parameters in relation to their careful preparation. Correspondingly, these problems affect the cost of a gauge and the hygrometer that is markedly higher than the cost of routine humidity meters. Beaton and Spowarti (2012) noted that a nitric oxide ionization cells were also unstable and they could fail in use or even while in the storage.

Experience has shown that krypton lamps are more stable, long-lived and inexpensive sources and their production is much easier. Therefore, during the past decades the UV krypton hygrometers have also been developed. First UV hygrometer using a krypton lamp was developed in 1985 (Campbell and Tanner 1985). The benefits of this instrument were a longer lifetime and easier production. But the absorption band in krypton hygrometers is not directly located in the Lyman–alpha band. Emission from the krypton tube exhibits a major band at 123.58 nm and a minor band at 116.49 nm.

Modern krypton hygrometers fabricated by Campbell Scientific Inc. is shown in Figure 3.33. This device (KH20 krypton hygrometer) is designed for measurement of rapid fluctuations in atmospheric water vapor (Foken and Falke 2012). High frequency response is suitable for eddy-covariance applications (up to 100 Hz). In the KH20, the source is a low-pressure krypton glow tube. As it was indicated before, emission from the krypton tube exhibits a major band at 123.58 nm and a minor band at 116.49 nm (Budovich et al. 2013). Radiation at 123.58 nm is strongly attenuated by the water vapor whereas the absorption by other gases in the optical path is relatively weak at this wavelength. However, at 123.6 nm the water vapor absorption strength is down by a factor of almost 2 in comparison with absorption at Lyman–alpha line, whereas the oxygen absorption is stronger by a factor of about 40 (Tillman 1965; Yoshino et al. 1996), making the oxygen interference much more severe and variable with air density, a particular problem for measurements from aircraft. Radiation at the shorter wavelength (116.49 nm) is also attenuated by the water vapor and oxygen molecules. This means that a much larger correction for oxygen is required than for Lyman–alpha device. In addition, a variable path length self-calibration technique cannot be applied in the krypton hygrometer, because this device works with two absorption lines, each with two absorbers. All this reduces the accuracy of measurement by krypton hygrometers. However, Campbell and Tanner (1985) determined that for their Kr lamp, this line (116.49 nm) contributes only 3% to the total lamp output. Since it is very near the nominal 115 nm cutoff for the magnesium fluoride windows, they concluded that it is greatly attenuated. Thus, the usage of MgF2 windows fitted to the source and detector tubes should help to improve the accuracy of measurements. But it is necessary to note that magnesium fluoride is hygroscopic material. This material changes its transfer characteristics in humid environments by interaction of atmospheric constituents with UV photons. Windows had to be cleaned manually more or less frequent. Another disadvantage is the wide range, nonlinear output signal of both Lyman–alpha and krypton-based systems. This either requires a special analog signal processing or the use of wide range A/D-converters for accurate calculations. In addition, we should keep in mind that due to simplicity of configuration and its signal offset drift these hygrometers cannot be used for determination of absolute concentrations. A characteristic of the KH20 is that the source tube window experiences scaling when operating, especially in humid environments. This will arise through the disassociation of atmospheric constituents by UV photons and can be removed simply by wiping the windows.

Figure 3.33   KH20 krypton hygrometers developed by Campbell Scientific: calibration range—1.7–19.5 g/m3 (nominal), frequency response—100 Hz; operating temperature range—30 to +50°C; weight—6.8 kg.

No doubts that other wavelengths in the UV range can be also used for hygrometer design. For example, the emission line (λ = 184.9499 nm) of the hydrogen low pressure lamp can be used for these purposes (Wernecke and Wernecke 2014). However, the sensitivity and the accuracy of these devices will be much worse than in the devices considered previously. Gersh and Matthew (1988) for their research have used the radiation at wavelengths of 177 and 205 nm. Xenon filled, fused silica window flashlamp with filters was used as the source of UV radiation, while the CsTe photocathodes, which are solar blind, were used as detectors. They used this hygrometer to monitor and control the operation of the industrial drying chamber. The basic concept of this UV hygrometer was to measure the differential absorption by water vapor of UV radiation in two spectral bands and to relate this measurement to the absolute humidity of the air in the measurement volume. In the first absorption spectral band (λ = 177 nm), the UV radiation is strongly attenuated by the water vapor, whereas in the second (λ = 205 nm) reference band, the UV radiation is not affected by the presence of water vapor. Then, the ratio of the absorption to the reference band intensities is a reflection of the mass of water vapor per unit volume in the measurement region (which is the definition of the absolute humidity). No doubts this approach is correct, but the selected spectral range does not provide the necessary sensitivity. A complication arises from the fact that atmospheric oxygen also absorbs the UV radiation in the same spectral region as water vapor.

With regard to other sources of light in the desired spectral range, that unfortunately in this particular spectral range the set of sources is very limited (Heering 2004). LEDs can be manufactured to emit light in the UV range, although practical LED arrays are very limited below 365 nm. LED efficiency at 365 nm is about 5%–8%, whereas the efficiency at 395 nm is closer to 20%, and the power outputs at these longer UV wavelengths are also better. Only Taoyuan Electron Lmt. (HK) suggests the UVC Flashlight LED operated at 280 nm (http://www.ledwv.com). But this source has small power ~25 mW, stringent requirements for the operation and short service life. The UVLEDs should be used within three months. A similar situation is observed for UV DL. The Ce:LiSAF and Ce:LiCAF DL (Lawrence Livermore National University) can generate radiation in the range from 280 to 316 nm. Lasermate Group, Inc. provides a diode-pumped solid-state lasers (DPSS) working in the range from 261 to 2200 nm. Laser 2000 produces QL262 laser with wavelength 262 nm (http://www.laser2000.co.uk/). Pulsed laser systems passively Q-switched laser, emitting wavelengths of 213 nm was designed by CryLaS GmbH (http://www.crylas.de/). Only gas lasers can generate light with wavelength smaller than 200 nm: the UV argon-fluoride (ArF) excimer lasers operate at 193 nm, and Ar2* excimer laser can operate at 126 nm. However, even in this case, the emission spectrum is not optimal for the hydrometer construction (Parkinson and Yoshino 2003). Furthermore, the use of gas lasers does not contribute to lower prices and sizes of developed hygrometers.

The use of semiconductor radiation detectors in the UV hygrometers is also limited because of low sensitivity in the current spectral range (Seib and Aukerman 1973; Razeghi and Rogalski 1996). Typical spectral characteristics of such photodetectors are shown in Figure 3.34.

Figure 3.34   Examples of Si semiconductor photodetectors optimized for operation in the UV spectral range.

#### 3.6  Fluorescent-Based Hygrometers

As was showed earlier, the water vapor plays a unique role in atmospheric processes as a key chemical and radiative component. Water vapor is present in all layers of the atmosphere including upper layers, such as the troposphere, stratosphere, and mesosphere, so it is important to control the concentration of water vapor in these layers. Using for these purposes most of the methods described before, as well as most of the solid humidity sensors, which will be considered in Vol. 2 “Electronic and Electrical Humidity Sensors” of our issue, is hampered because of the low water vapor concentrations at these altitudes and strong variations in temperature when the altitude changes. At certain altitudes, a strong temperature jumps are possible. Currently, a limited number of techniques can be used for in situ measurements of the upper troposphere and stratosphere humidity. Among them, the fluorescent technique, offering high accuracy and fast response, has proved reliable (Meyer et al. 2015). Devices using this technique worked well when used on the board balloons, high-altitude aircrafts and rockets.

#### 3.6.1  Principles of Fluorescent-Based Technique

Fluorescence is a radiation that is emitted after a chemical species absorb a radiation of another wavelength. For excitation of fluorescence, the UV radiation is being used. It results from the instantaneous deactivation that takes place after the species go to a higher energy state upon the absorption of incident radiation. The excited species has a lifetime of about 1–100 ps, and returns to its ground (unexcited) state through the emission of radiation:

3.4 ()

where:

• A* is an excited state of a substance A
• h is Planck’s constant
• ν is the frequency of the photon

The emission occurs in all directions, and is of a longer wavelength than the incident radiation. The wavelength of the emitted radiation is influenced by the chemical structure of the fluorescent species. The fluorescence quantum yield gives the efficiency of the fluorescence process. It is defined as the ratio of the number of photons emitted to the number of photons absorbed:

3.5 ()

The radiant intensity of the emission conveys information on the concentration of the fluorescent species. For a weakly absorbing system (i.e., A < 0.05), the fluorescence intensity (I F) is linearly related to the concentration (C) of the fluorescent species, as expressed in the following equation:

3.6 ()

where:

• I 0 is the intensity of the incident radiation
• k F is a constant that is dependent on the absorption characteristics of the fluorescent species, the quantum efficiency of the fluorescence, and the configuration of the instrumentation system

In relation to water vapor, the theory of fluorescent-based technique has been developed by Kley and Stone (1978) and Bertaux and Delannoy (1978). Later refinements were made by Keramitsoglou et al. (2002). The fluorescent method of water vapor detection is based on the photodissociation of H2O at wavelengths below 137 nm and the subsequent fluorescent relaxation of the exited OH* radical produced (Yushkov et al. 1995; Kley et al. 2000). For Layman–alpha dissociation of water vapor, the process can be expressed as

3.7 ()

with a quantum yield far less than 1. About 10% of the absorbed photons result in excitation of the OH fragment to the A 2Σ+ electronic state. The electronically excited OH* radical either fluoresces at 310 nm

3.8 ()

or is quenching by collisions with air molecules

3.9 ()

Fluorescence is seen in both (0→0) and (1→0) bands but the strongest fluorescence is from highly excited rotational levels (N = 20–22) of the (0→0) band.

A photomultiplier with an interference filter measures the intensity of fluorescence which is signature of the parent molecules of H2O. The intensity of fluorescence is obtained as

3.10 ()

where:

• L is the length of the absorption between the lamp and interaction region
• [OH*], [H2O], [air], [O2] are the number densities of OH*(A 2Σ+), H2O, air and oxygen respectively
• A is Einstein transition probability
• F λ is photon flux of the light source
• and are cross sections of water vapor and oxygen for Layman–alpha, respectively
• φ is the quantum yield of photo dissociation
• kq quenching coefficient

Thus, by measuring the fluorescence radiation the H2O abundance can be determined.

Of course, for correct measurement of the photon flux, the absorption by oxygen has to be taken into account. However, at air pressure below 10−1 hPa the quenching by the air and oxygen absorption are negligible (Kley and Stone 1978) and the fluorescence intensity becomes

3.11 ()

In the other limiting case P air > 10 hPa, that is, in the atmosphere up to 20–35 km, kq *[air]>> A and hence

3.12 ()

The factor C in Equation 3.12 summarizes molecular coefficients, known from the literature, as well as instrument specific quantities. For example, for C calculations one can use the values of the following parameters: Einstein coefficient for reaction Equation 3.8, A = 1.26·106 s−1 (Crosley and Lenge 1975); and kp = 2.3·10−11 cm−3 s−1 (Kley and Stone 1978). If C is a constant, the number of detected fluorescence photons is proportional to the H2O mixing ratio [H2O]/[air] for measurements in the troposphere and lower stratosphere. Then the water vapor mixing ratio (μ) may be expressed (following Kley and Stone 1978) as

3.13 ()

For measurements at higher altitudes, Equation 3.11 has to be used to obtain correct water vapor mixing ratios. Thus, under condition with negligible oxygen absorption, the fluorescence gives a direct measurement of the atmospheric water vapor mixing ratio in this case (Yushkov et al. 1995). Of course, we have to keep in mind that in reality, C is a function of the photo dissociation rate of the reaction Equation 3.7, and thus depends on the photon flux in the fluorescence volume, which in turn depends on the variations of the lamp intensity (Zoger et al. 1999).

#### 3.6.2  Examples of Fluorescent-Based Hygrometers Realization

The first fluorescent-based hygrometers suitable for research of atmosphere have been developed by Kley and Stone in 1978. However, the actual flight instrument for use in the stratosphere was described and characterized in 1979 (Kley et al. 1979). Later on fluorescent-based hygrometers, developed in other laboratories, have appeared ( Goutail and Pommereau 1987; Weinstock et al. 1990, 1994; Kretova et al. 1991; Yushkov et al. 1995; Zoger et al. 1999). The most advanced Lyman–alpha fluorescence hygrometers have been developed in the laboratories at National Oceanic and Atmospheric Administration (NOAA) in Boulder, at Harvard and in Julich (Germany).

All existing fluorescent-based hygrometers are built on a similar principle, but constructively they may differ essentially. Their versions are shown in Figure 3.35. Typically, the difference in design is associated with different designations of developed devices and different approach to calibrating fluorescent-based hygrometers. For example, the devices designed for aircraft applications, in contrast to the devices intended for balloon measurements (Yushkov et al. 1995), are not limited in size and weight, which allows entering additional elements, improving their parameters (Goutail and Pommereau 1987; Zoger et al. 1999). Typically, such devices have a measuring chambers and additional receivers for controlling the intensity of the UV lamp radiation (Figure 3.35). The presence of closed measuring chamber essentially extends the opportunity to use fluorescent-based hygrometers, as it allows using them during the day.

Figure 3.35   (a) Sketch of the fluorescence cell and the major components as Lyman–alpha radiation source, photomultiplier tube, vacuum UV (VUV) detectors and the mirror drive in the hygrometer designed by Zoger et al. (1999). (b) Conceptual diagram illustrating the principal mechanical and optical components of the airborne Fast In-situ Stratospheric Hygrometer (FISH). The size of the cell is 0.3 L in total. As the lamp is not monochromatic, the number of lamp background counts also has to be taken into account. Therefore a swiveling mirror is implemented between the lamp and the measuring cell. During one measurement cycle the mirror is placed in three different positions to determine the total fluorescence rate Ng (mirror position 1), the background rate Nu (mirror position 3) and the lamp intensity I0 (mirror position 2). Thus, the mirror drive allows monitoring the intensity of the Lyman–alpha radiation (I0) and a determination of the background and dark count rates. Ng is the number of detected fluorescence photons proportional to the water vapor mixing ratio. Thus, calibration factors can be determined which are nearly independent of changes of the output of the radiation source. The measurements using this hygrometer can be accomplished with a precision <0.2 ppmv at 1 s integration time. ((a) From Zoger, M. et al., J. Geophys. Res., 104, 1807–1816, 1999. With permission. (b) from Meyer, J. et al., Atmos. Chem. Phys. 15, 8521–8538, 2015. Published by the European Geosciences Union as open access.)

Hygrometer shown in Figure 3.36 does not have such a possibility, because the measurements take place in the open space. This means that the readings are sensitive to the ambient solar radiation, which usually exceeds the fluorescence radiation to be detected.

Figure 3.36   (a) Optical scheme and (b) electronics block diagram of fluorescent-based compact hygrometer designed by Yushkov et al. (1995) to carry out the high-resolution night-time upper tropospheric and stratospheric water vapor balloon measurements. To increase the detection efficiency of the useful signal and to reduce the size of the device, the radiation source and the optical detection system are positioned coaxially. The lenses are made from U-Viol glassUS-49. The front lens has a diameter 50 mm and is sealed to the lamp body. The focal length is 10 mm which leads to a distance of 24 mm between the windows of vacuum ultraviolet (VUV) lamp and the analyzed volume. Hygrometer did not use special measurement cell. The measurements are carried out in the open space in the immediate vicinity of the objective. Therefore, the device does not need a suction system to sample the air to be analyzed. The total uncertainty of the measurement is less than 10% at the stratospheric mixing ratios greater than 3 ppmv increasing to about 20% at mixing ratios less than 3 ppmv. A description of improved version of hydrometer one can find in (Lykov et al. 2011, 2012). Compact and light hygrometer has a mass ~0.5 kg. The absence of the receiver, controlling the intensity of the lamp radiation, the authors explain by the high stability of the radiation intensity of lamps used in hygrometer.

(Idea from Yushkov, V. et al., SPIE Proc., 2506, 783–794, 1995.)

In this regard, present device can only be used at night time. However, the devices with an open structure have another significant advantage. They avoid contamination by desorption of H2O from the walls, which is very important when measuring in the atmosphere with a very low concentration of water vapor.

For excitation of fluorescence, one can use the hydrogen or krypton vacuum UV lamps described earlier (Varier 1967; Buck 1976b; Zoger et al. 1999). As it was mentioned previously, the radiation of the hydrogen lamp contains a Layman line (121.6 nm). An oxygen spectral window at this line provides the most effective generation of excited hydroxyl radicals within the volume where fluorescence is being registered. The radiation of krypton vacuum UV lamp does not possess this property. Hydrogen vacuum UV lamps may have different design, which is associated with the use of various sources of hydrogen. The most common sources of hydrogen are the uranium hydride and a mixture of hydrogen and helium.

According to Varier (1967) and Buck (1976b), the intensity of a 121.6 nm line in glow-discharge lamps using a mixture of 2%–25% hydrogen and helium or argon grows through decreasing recombination of hydrogen atoms on the walls of the bulb. Such lamps are easier to manufacture and are more ecologically safe compared to those containing uranium hydride and uranium (Hutcheson 1972; Keramitsoglou et al. 2002). It was established that the uranium hydride could be replaced by the hydride of ZrCo, which possesses the same thermodynamic properties (Lykov et al. 2012). This replacement allows eliminating the thermostatic control of the lamp, which helps to reduce the size and weight of the hygrometer.

But, it was established that in the spectra of hydrogen UV lamps within 200–360 nm range an intensive hydrogen continuum is present, which overrides the spectrum of hydroxyl fluorescence and may cause a considerable noise during laboratory calibration of hygrometers in a closed chamber. Therefore, in optical fluorescence hygrometers, which use a hydrogen glow-discharge lamp, a 270–320 nm radiation should be rejected with a special window-filter. For such filter, MgF2 is usually used, spectral characteristics of which are shown in Figure 3.37.

Figure 3.37   Typical spectra of the transmittance of monocrystalline magnesium fluoride window of hydrogen lamp without filter (1) and with filter (2) for selective absorption at 300 nm.

(Data extracted from Yushkov, V. et al., SPIE Proc., 2506, 783–794, 1995.)

To improve sensitivity, an interference filter centered at 318 nm, is also desirable. It allows selecting the spectral region, coincident with the emission from the upper rotational levels of the (0→0) band of the AX system of OH. With regard to the sensitivity measurement of the photofragment fluorescence, it is usually carried out by sun-blind photomultiplier (PM).

Usually for this purposes a HAMAMATSU R647-P photomultiplier is being used. In some devices, in addition to indicated above elements, one (Goutail and Pommereau 1987) or two (Zoger et al. 1999) additional detectors may also be present. They are applied to control the intensity of radiation, used for fluorescence excitation. Typically, such detectors are the NO filled ionization chamber detectors. The energy required to ionize NO corresponds to a threshold wavelength of 134 nm. Therefore, the NO filled ionization cells are a highly selective radiation sensor for the vacuum UV (VUV) spectral region (Samson 1967). An iodine ionization cell that is sensitive from 115–135 nm can also be used (Kelly et al. 1989). The electronic system used in hygrometer designed for the lamp modulation and synchronous demodulation of the signal received. This technique improves noise-to-signal ratio in more than 100 times.

Studies have shown that the fluorescent-based hygrometers have high sensitivity. Besides, they are compact enough for the balloon and aircraft measurements (Table 3.13). Rocket-borne measurements in the mesosphere using this technique have also been reported (Khaplanov et al. 1996). In addition, the Lyman–alpha fluorescence technique can achieve a large dynamic range for measurements from the middle and upper troposphere at about 1000 ppmv into the dry stratosphere with only 2–5 ppmv, where changes on the order of 0.1 ppmv can be detected with a relative uncertainly of ±5%. At that high speed of response, when integration time is on the order of 1 s, enables the measurement of small-scale features in the atmosphere. However, the H2O measurement range of these devices is limited to pressures lower than 300 hPa because of strong Lyman–alpha absorption in the lower troposphere. Other disadvantages of these hygrometers are the instability of emission hydrogen UV lamp, a relatively short service life, and the lack of commercially produced samples with reproducible parameters.

### Table 3.13   Examples of Lyman–Alpha Photofragment Fluorescence Hygrometers Designed and Fabricated for Atmosphere Monitoring Using Balloons and Aircrafts

Hygrometers

References

NOAA Aeronomy instrument (U.S.)

Harvard Water Vapor (HWV) instrument (U.S.)

Central Aerological Observatory (CAO) instrument (Russia)

FISH (Fast in situ Stratospheric hygrometer) instrument (Germany)

UK Met Office instrument (UK)

Numerous studies have confirmed that the optical hygrometers have several important advantages, which are not found in other hygrometers. At first, the measurement of water vapor using this optical (or spectroscopic) technique has received very favorable response from the industry because these devices employ a noncontact sensor technology for the detection of water vapor in various environments including natural gas: the sample never comes into direct contact with the sensor element. This noncontact approach significantly reduces maintenance requirements of the instruments and reduces the overall costs of maintaining the equipment. Moreover, in principle, optical methods for humidity measurement can be used in extreme environmental conditions (high pressure, high temperature, aggressive gases, etc.), and the lack of delay of the measurement. The speed of response is limited only by the speed of the indicator or recorder. The time-constant of an optical hygrometer is typically just a few milliseconds. Therefore, these instruments are very useful for tracking fast humidity fluctuations, since the method does not require the detector to achieve vapor-pressure equilibrium with the sample. This makes the method ideal for use in measuring humidity from an airplane or under any circumstance where rapid changes of humidity can occur.

In addition, optical method has specific sampling; it is possible to sample instantaneously any desired path length of atmosphere from a few millimeters to thousands of meters (Figure 3.38), and in so doing arrive at an integrated value of the absolute humidity in the path in question. This is co-called as open-path measurement. If the path length is appreciable, the value obtained is more representative than that obtained by spot sampling (Foskett et al. 1953). Of course, extractive or cell-based IR measurements, commonly used in the previously considered optical hygrometers, have the obvious advantage over the open-path setup that the physical conditions (concentration, pressure, temperature, absorbing path length, etc.) of the measurement can be readily controlled or changed at will (Table 3.14). This provides better conditions for prolonged measurements when high precision of determination is needed, for example, for fundamental studies in the gas spectroscopy or analytical method development. Open-path measurements, on the contrary, are more prone to technical difficulties connected to changing atmospheric (meteorological) conditions leading to uncertainties, but the in situ nature of such studies offer distinctive advantages as well.

Figure 3.38   Experimental setup for remote atmosphere monitoring with retroreflector using open-path measurement.

(From Korotcenkov, G. (Ed.), Electrochemical and Optical Sensors, Momentum Press, New York, pp. 311–476. With permission.)

### Table 3.14   Comparison of Extractive and Open-Path Techniques

Extractive (Closed Cell)

Open-Path

• Path-length adjustable, independent of plume size
• Simplicity, repeatability
• Lower detection limits
• Stable conditions, longer measurement times
• Temperature and pressure control
• Quantitative reference and background spectra collected more easily

• In situ measurement, non-invasive sampling
• Detection of very polar, reactive, and labile compounds
• Continuous monitoring
• Line-integrated (averaged)
• concentrations: area sources of pollution can be examined
• Measurement at dangerous sites

• Time delay between sampling and measurement
• Very polar, reactive compounds difficult to sample (may lead to errors in the analysis)
• Memory effects, losses on the cell wall

• Path-length limited by the size of plume
• Field adjustment needed at every site
• Weather dependence
• No temperature and pressure control
• Decreasing S/N for longer paths
• Calibration spectra over the same path as the sample cannot be collected
• Experiments are hardly repeatable

Source: Meyers, R.A. (Ed.), Encyclopedia of Analytical Chemistry, Wiley, Chichester, UK, 2000.

The advantages of optical hygrometer can also include their ability to maintain high sensitivity even at low concentrations of water vapor, where the psychrometer, hair hygrometer, electrical hygrometer, and even the dew point apparatus all become less sensitive. Moreover, optical hygrometer will operate as well below the freezing point and above it. In addition, this instrument has negligible drift and can generally operate over a wide humidity range. Optical hygrometers can be used for the measurement of high humidity, as well as for very low concentrations in the range of ppmv. Finally, the method in no way alters the sample concentration by either adding or subtracting water or changing the state of any part of the sample as occurs in the psychrometric or dew point methods.

We need to note that all mentioned above devices have advantages and also shortcomings. Optical techniques often require a high effort in equipment and analysis.

These are not diminutive. The size of measurement cell with required optical path length is a limiting factor in the development of IR hygrometers. Attempts to reduce the size of IR hygrometers are accompanied by a restriction of the measurement range and the sensitivity decrease. Miniaturization of UV hygrometers is impossible due to the size of existing UV light sources and the detectors of radiation in this area. Besides, optical hygrometers are expensive in comparison with other devices. They may also suffer from contamination. In addition, regular calibration is required for many traditional instruments, which is both inconvenient and expensive. Usually for this purposes they are used other hygrometers such as dew or frost point hygrometers. Therefore, according to WMO reports (1992, 2011), optical hygrometers are more suitable for measuring changes in the vapor concentration rather than absolute levels.

Who has the desire to learn more about optical sensors and optical spectroscopy can refer to reviews and books published in this field (Wolfbeis 1991, 1992, 2006, 2008; Ligler and Rowe Taitt 2002; Narayanaswamy and Wolfbeis 2004; Orellana and Moreno-Bondi 2005; Orellana 2006; Tkachenko 2006; McDonagh et al. 2008; Korotcenkov et al. 2011, and many other).

#### References

Allen M.G. (1998) Diode laser absorption sensors for gas–dynamic and combustion flows. Meas. Sci. Technol. 9, 545–562.
Amerov A. , Maskas M. , Meyer W. , Fiore R. , Tran K. (2007) New process gas analyzer for the measurement of water vapor concentration. In: Proceedings of ISA 52nd Symposium of Instrumentation, Systems, and Automation Society, Houston, TX, pp. 63–74, April 15–19, 2007.
Amerov A. , Fiore R. , Langridge S. (2009) Process gas analyzer for the measurement of water and carbon dioxide concentrations. In: Proceedings of the Annual ISA Analysis Division Symposium, Vol. 478, pp. 184–197.
Bange J. , Espositi M. , Lenschow D.H. , Brown P.R.A. , Dreiling V. , Giez A. et al. (2013) Measurement of aircraft state and thermodynamic and dynamic variables. In: Wendisch M. , Brenguier J.-L. (Eds.), Airborne Measurements for Environmental Research. Methods and Instruments, Wiley-VCH, Weinheim, Germany, Chapter 2.
Bass M. , van Stryland E.M. (Eds.) (2002) Fiber Optics Handbook, McGraw-Hill, New York.
Bauer D. , Heeger M. , Gebhard M. , Benecke W. (1996) Design and fabrication of a thermal infrared emitter. Sens. Actuators A 55 (1), 57–63.
Beaton S.P. , Spowart M. (2012) UV absorption hygrometer for fast–response airborne water vapor measurements. J. Atmos. Oceanic. Technol. 29, 1295–1303.
Bertaux J.-L. , Delannoy A. (1978) Vertical distribution of H2O in the stratosphere as determined by UV fluorescence in-situ measurements. Geophys. Res. Lett. 5, 1017–1020.
Buchholz B. , Ebert V. (2013) Field tests of a new, extractive, airborne 1.4 μm–TDLAS hygrometer (SEALDH–I) on a Learjet 35A. Geophys. Res. Abstracts 15, EGU2013–4569–1.
Buchholz B. , Kühnreich B. , Smit H.G.J. , Ebert V. (2013) Validation of an extractive, airborne, compact TDL spectrometer for atmospheric humidity sensing by blind intercomparison. Appl. Phys. B 110, 249–262.
Buck A.L. (1973) Development of an improved Lyman–alpha hygrometer. Atmos. Technol. 2, 213–240.
Buck A.L. (1976a) The variable–path Lyman–alpha hygrometer and its operating characteristics. Bull. Amer. Meteor. Soc. 57, 1113–1118.
Buck A.L. (1976b) Lyman-α radiation source with high spectral purity. Appl. Opt. 16, 2634–2638.
Buck A.L. (1985) The Lyman-alpha absorption hygrometer. In: Proceedings, Moisture and Humidity Symposium, Washington, D.C. , Instrument Society of America, Research Triangle Park, NC, p. 411.
Budovich V.L. , Polotnyuk E.B. , Gerasimov N. , Krylov B.E. (2013) Characteristics of a glow–discharge krypton lamp in the vacuum ultraviolet. J. Opt. Technol. 80 (11), 691–694.
Campbell G.S. , Tanner B.D. (1985) A krypton hygrometer for measurement of atmospheric water vapor concentration. In: Moisture and Humidity: Measurement and Control in Science and Industry. Instrument Society of America, Research Triangle Park, Proceedings of the 1985 International Symposium on Moisture and Humidity, Instrument Society of America, pp. 609–614.
Carver J.H. , Mitchell P. (1964) Ionization chambers for the vacuum ultra–violet. J. Sci. Instrum. 41, 555–557.
Cerni T.A. (1994) An infrared hygrometer for atmospheric research and routine monitoring. J. Atmos. Oceanic Technol. 11, 445–462.
Chalmers J.M. (2002) Mid-infrared spectroscopy: anomalies, artifacts and common errors. In: Chalmers J.M. , Griffiths P.R. (Eds.), Handbook of Vibrational Spectroscopy, John Wiley & Sons, Chichester, UK, pp. 2327–2347.
Chen H.-S. , Mitsuta Y. (1967) An infrared absorption hyhrometer and its application to the study of the water vapor flux near the ground. Special, Contribution. Geophysical Institute, Kyoto University, No. 7, pp. 83–94.
Cho J. , Newell R.E. , Sachse G.W. (2000) Anomolous scaling of mesoscale tropospheric humidity fluctuations. Geophys. Res. Lett. 27, 377–380.
Comini E. , Faglia G. , Sberveglieri G. (Eds.) (2009) Solid State Gas Sensor. Springer, New York.
Cozzani E. , Summonte C. , Belsito L. , Cardinali G.C. , Roncaglia A. (2007) Design study of micromachined thermal emitters for NDIR gas sensing in the 9–12 μm wavelength range. IEEE Sens. 1–3, 181–184.
Crosley D.R. , Lengel R.K. (1975) Relative transition probabilities and the electronic transition moment in the A-X system of OH. J. Quant. Spectrosc. Radiat. Tran. 15, 579–591.
Dieke G.H. , Cunningham S.P. (1952) A new type of hydrogen discharge tube. J. Opt. Soc. Amer. 42, 187–189.
Durry G. , Amarouche N. , Joly L. , Liu X. , Parvitte B. , Zeninari, V. (2008) Laser diode spectroscopy of H2O at 2.63 μm for atmospheric applications. Appl. Phys. B 90, 573–580.
Ebert V. , Buchholz B. (2016) Absolute high–speed laser hygrometry: From aircraft applications to primary validation. In: Proceedings of the Conference on Lasers and Electro–Optics (CLEO:2016): Science and Innovations, OSA, San Jose, CA, STh4J.1, June 5–10, 2016.
Edwards C.S. , Barwood G.P. , Gill P. , Stevens M. , Schirmer B. , Benyon R. , Mackrod P. (2000) An IR tunable diode laser absorption spectrometer for trace humidity measurements at atmospheric pressure. In: Proceedings of 2000 IEEE Sensor Conference, Aventura, FL, pp. II3–II4, July 24–28, 2000.
Faist J. , Capasso F. , Sivco D.L. , Sirtori C. , Hutchinson A.L. , Cho A.Y. (1994) Quantum cascade laser. Science 264, 553–555.
Foken T. (2008) Micro–Meteorology. Springer, Berlin, Germany.
Foken T. , Falke H. (2012) Technical Note: Calibration instrument for the krypton hygrometer KH20. Atmos. Meas. Tech. Discuss. 5, 1695–1715.
Foken T. , Dlugi R. , Kramm G. (1995) On the determination of dry deposition and emission of gaseous compounds at the biosphere–atmosphere interface. Meteorol. Z. 4, 91–118.
Fonollosa J. , Carmona M. , Santander J. , Fonseca L. , Moreno M. , Marco S. (2009) Limits to the integration of filters and lenses on thermoelectric IR detectors by flip–chip techniques. Sens. Actuators A 149, 65–73.
Foskett L.W. , Foster N.B. , Thickstun W.R. , Wood R.C. (1953) Infrared absorption hygrometer. Monthly Weather Rev. 81 (9), 267–277.
Fowle E.E. (1912) The spectroscopic determination of aqueous vapor. Astrophys. J. 35, 149–162.
Gat J.R. (1996) Oxygen and hydrogen isotopes in the hydrological cycle. Ann. Rev. Earth Planet. Sci. 24, 225–262.
Gersh M.E. Matthew M.W. (1988) VaporSense contamination-resistant high temperature UV hygrometer. In: SPIE Vol. 961 Industrial Optical Sensing, pp. 52–67.
Ghysels M. , Riviere E.D. , Khaykin S. , Stoeffler C. , Amarouche N. , Pommereau J.-P. , Held G. , Durry G. (2016) Intercomparison of in situ water vapor balloon-borne measurements from Pico-SDLA H2O and FLASH-B in the tropical UTLS. Atmos. Meas. Tech. 9, 1207–1219.
Goutail F. , Pommereau J.-P. (1987) Stratospheric water vapor in situ measurements from infra-red montgolfier. Adv. Space Res. 7 (7), 111–114.
Graf A. , Arndt M. , Sauer M. , Gerlach G. (2007) Review of micromachined thermopiles for infrared detection. Meas. Sci. Technol. 18, R59–R75.
Green A.E. , Kohsiek W. (1995) A fast response, open path, infrared hygrometer, using semiconductor source. Bound. Layer Meteor. 74, 353–370.
Grouiez B. , Parvitte B. , Joly I. , Courtois D. , Zeninari V. (2008) Comparison of a quantum cascade laser used in both CW and pulsed modes: Application to the study of SO2 lines around 9 μm. Appl. Phys. B 90, 177–186.
Grund C.J. , Hardesty R.M. , Rye B.J. (1995) Feasibility of tropospheric water vapor profiling using infrared heterodyne differential absorption Lidar. In: Proceedings of 5th Atmospheric Radiation Measurement (ARM) Science Team Meeting, San Diego, CA, March 19–23. http://www.arm.gov/publications/proceedings/conf05/
Heering W. (2004) UV sources: Basics, properties and applications. IUVA News 6 (4), 7–13.
Herriot D. , Kogelnik H. , Kompfner R. (1964) Off-axis paths in spherical mirror interferometers. Appl. Opt. 3, 523–526.
Herriot D.R. , Schulte H.J. (1965) Folded optical delay lines. Appl. Opt. 4, 883–889.
Hintsa E.J. , Weinstock E.M. , Anderson J.G. , May R.D. , Hurst D. (1999) On the accuracy of in situ water vapor measurements in the troposphere and lower stratosphere with the Harvard Lyman-α hygrometer. J. Geophys. Res. 104, 8183–8189.
Hvozdara L. , Pennington N. , Kraft M. , Karlowatz M. , and Miziakoff B. (2002) Quantum cascade lasers for midinfrared spectroscopy. Vib. Spectrosc. 30, 53–58.
Hutcheson E. (1972) Monitoring the thickness of thin MgF and LiF films on alluminium by reflectant measurements using 1216 Å line of hydrogen. Appl. Opt. 11 (7), 1590–1595.
Kebabian P. , Kolb C.E. , Freedman A. (2002) Spectroscopic water vapor sensor for rapid response measurements of humidity in the troposphere. J. Geophys. Res. Atmosph. 107 (23), 4670.
Kelly K.K. , Tuck A.F. , Murphy D.M. et al. (1989) Dehydration in the lower Antarctic stratosphere during late winter and early spring 1987. J. Geophys. Res 94 (D9), 11317–11357.
Keramitsoglou I. , Harries J.E. , Colling D.J. , Barker R.A. , Foot J.S. (2002) A study of the theory and operation of a resonance fluorescence water vapour sensor for upper tropospheric humidity measurements. Meteorol. Appl. 9, 443–453.
Khaplanov M. , Gumbel J. , Wilhelm N. , Witt G. (1996) Hygrosonde-A direct measurement of water vapor in the stratosphere and mesosphere. Geophys. Res. Lett. 23, 1645–1648.
Kley D. , Stone E. (1978) Measurements of water vapor in the stratosphere by photo dissociation with Ly-α (1216 Å) light. Rev. Sci. Instrum. 49 (6), 691–697.
Kley D. , Stone E.J. , Henderson W.R. , Drummond J.W. , Harrop W.J. , Schmeltekopf A.L. , Thompson T.L. , Winkler R.H. (1979) In situ measurements of the mixing ratio of water vapor in the stratosphere. J. Atmos. Sci. 36, 2513–2524.
Kley D. (1984) Ly(α) absorption cross–sections of H2O and O2 . J. Atmos. Chem. 2 (2), 203–210.
Kley D. , Russell III J.M. , Phillips C. (Eds.) (2000) SPARC assessment of upper tropospheric and stratospheric water vapour. World Climate Research Programme, SPARC Report No.2.
Korotcenkov G. , Cho B.K. , Sevilla III F. , Narayanaswamy, R. (2011) Optical and fiber optic chemical sensors. In: Chemical Sensors: Comprehensive Sensor Technologies. Vol. 5: Electrochemical and Optical Sensors, Korotcenkov G. (Ed.), Momentum Press, New York, pp. 311–476.
Kosterev A.A. , Tittel F.K. (2002) Chemical sensors based on quantum cascade lasers. IEEE J. Quantum Electron. 38, 582–590.
Kretova M. , Khaplanov M. , Yushkov V. (1991) Resonance-fluorescence hygrometer for humidity measurement in the stratosphere. Prib. Tekh. Eksp. 3, 195–198 (in Russian).
Kretschmer S.I. , Karpovitsch J.V. (1973) Sensitive ultraviolet hygrometer. Izv. AN SSSR, Fiz. Atm. i Okeana 9, 642–645 (in Russian).
Ligler F.S. , Rowe-Taitt C.A. (2002) Optical Biosensors—Present and Future. Elsevier, New York.
Linnerud I. , Kaspersen P. , Jaeger T. (1998) Gas monitoring in the process industry using diode laser spectrosocopy. Appl. Phys. B 67, 297–305.
Liu F. (2012) A new method for measuring trace moisture in natural gas. In: Proceedings of ISA 57th Analysis Division Symposium, Anaheim, CA. Sessions 7–12.
Lübken F.-J. , Dingler F. , von Lucke H. , Anders J. , Riedel W.J. , Wolf H. (1999) MASERATI: A rocketborne tunable diode laser absorption spectrometer. J. Anders, Appl. Opt. 38, 5338–5349.
Lykov A. , Yushkov V. , Khaykin S. , Astakhov V. , Budovich V. (2011) New version of balloon hygrometer for in situ water vapor measurements in the upper troposphere and lower stratosphere (flash-BM). In: Proceedings of 20th ESA Symposium on European Rocket and Balloon Programmes and Related Research, Hyère, France, ESA SP-700, May 22–26, 2011.
Lykov A.D. , Astakhov V.I. , Korshunov L.I. , Yushkov V.A. , Budovich V.L. , Budovich D.V. , Dubakin A.D. (2012) Using vacuum ultraviolet sources in the design of a fluorescence hygrometer. J. Opt. Technol. 79 (8), 515–520.
Martini L. , Stark B. , Hunsalz G. (1973) Elektronisches Lyman–Alpha–feuchtigkeitsmessgerat. Meteorol. Z. 23, 313–322.
Matthey R. , Schilt S. , Werner D. , Affolderbach C. , Thevenaz L. , Mileti G. (2006) Diode laser frequency stabilisation for water-vapour differential absorption sensing. Appl. Phys. B 85 (2), 477–485.
May R.D. (1998) Open–path, near–infrared tunable diode laser spectrometer for atmospheric measurements of H2O. J. Geophys. Res. 103 (D15), 19161–19172.
McDonagh C. , Burke C.S. , MacCraith B.D. (2008) Optical chemical sensors. Chem. Rev. 108, 400–422.
McManus J. , Kebabian P. , Zahniser M. (1995) Astigmatic mirror multiple pass absorption cells for long pathlength spectroscopy. Appl. Opt. 34, 3336–3348.
Meyers R.A. (Ed.) (2000) Encyclopedia of Analytical Chemistry. Wiley, Chichester, UK.
Meyer J. , Rolf C. , Schiller C. et al. (2015) Two decades of water vapor measurements with the FISH fluorescence hygrometer: A review. Atmos. Chem. Phys. 15, 8521–8538.
Mestayer P. , Rebattet C. , Goutail F. (1986) Improved Lyman‐alpha hygrometer for small‐scale atmospheric turbulence measurements. Part I: Miniaturizing the sampling volume. Rev. Sci. Instrum. 57, 20–25.
Mestayer P. , Rebattet C. , Goutail F. (1987) Improved Lyman‐alpha hygrometer for small‐scale atmospheric turbulence measurements. Part II: The O2 filtering technique. Rev. Sci. Instrum. 58, 2165.
Mirumyants S.O. , Maksimyuk V.S. (2002) A portable optical gas analyzer for remote measurement of the methane concentration in closed spaces and rooms. J. Opt. Technol. 69 (12), 900–903.
Narayanaswamy R. and Wolfbeis O.S. (Eds.) (2004) Optical Sensors—Industrial, Environmental and Diagnostic Applications, Springer Series on Chemical Sensors and Biosensors, Vol. 1. Springer-Verlag, Berlin, Germany.
Nwaboh J.A. , Pratzler S. , Werhahn O. , Ebert V. (2017) Tunable diode laser absorption spectroscopy sensor for calibration free humidity measurements in pure methane and low CO2 natural gas. Appl. Spectrosc. 71 (5), 888–900.
Orellana G. (2006) Fluorescence-based sensors. In: Baldini F. , Chester A.N. , Homola J. , Martellucci S. (Eds.), Optical Chemical Sensors. Springer-Verlag, Dordrecht, the Netherlands, pp. 99–116.
Orellana G. , Moreno-Bondi M.C. (Eds.) (2005) Frontiers in Chemical Sensors: Novel Principles and Techniques, Springer Series on Chemical Sensors and Biosensors, Vol. 3. Springer-Verlag, Berlin, Germany.
Parkinson W.H. , Yoshino K. (2003) Absorption cross–section measurements of water vapor in the wavelength region 181–199 nm. Chem. Phys. 294, 31–35.
Rancourt R.D. (1996) Optical Thin Films: User Handbook, SPIE Press, Bellingham, Washington, DC.
Razeghi M. , Rogalski A. (1996) Semiconductor ultraviolet detectors. J. Appl. Phys. 79 (10), 7433–7473.
Rothman L.S. , Gordon I.E. , Babikov Y. et al. (2013) The HITRAN 2012 molecular spectroscopic database. J. Quant. Spectrosc. Rad. Trans. 130, 4–50.
Samson J.A.R. (1967) Techniques of Vacuum Ultraviolet Spectroscopy. John Wiley & Sons, New York.
Santander J. , Sabate N. , Rubio R. , Calaza C. , Fonseca L. , Gracia I. , Cane C. , Moreno M. , Marco S. (2005) Mirror electrostatic actuation of a medium infrared tuneable Fabry–Perot interferometer based on a surface micromachining process. Sens. Actuators A 123–124, 584–589.
Schanot A.J. (1987) An evaluation of the uses and limitations of Lyman-alpha hygrometer as an operational airborne humidity sensor. In: Proceedings of 6th Symposium Meteorological Observations and Instrumentation, American Meteorological Society, New Orleans, LA, pp. 257–260.
Schilz J. (2000) Thermophysica minima: Applications of Thermoelectric Infrared Sensors (Thermopiles): Gas Detection by Infrared absorption; NDIR, PerkinElmer Optoelectronics.
Seib D.H. , Aukerman L.W. (1973) Photodetectors for the 0.1 to 1.0 μm spectral region. In: Morton L. (Ed.) Advances in Electronics and Electron Physics, Vol. 34, Academic Press, New York, pp. 95–221.
Seidel A. , Wagner S. , Ebert V. (2012) TDLAS–based open–path laser hygrometer using simple reflective foils as scattering targets. Appl. Phys. B 109, 497–504.
Sigrist M.W. (Ed.) (1994) Air Monitoring by Spectrometric Techniques. John Wiley & Sons, New York.
Silver J.A. , Stanton A.C. (1987) Airborne measurements of humidity using a single–mode Pb–salt diode laser. Appl. Opt. 26, 2558–2566.
Silver J.A. , Hovde D.C. (1994) Near–infrared diode laser airborne hygrometer. Rev. Sci. Instrum. 65 (6), 1691–1694.
Smith F.G. (Ed.) (1993) Atmospheric Propagation of Radiation, volume 2 of The Infrared & Electro–Optical Systems Handbook ( J.S. Accetta , D.L. Shumaker Eds.), Infrared Information Analysis Center, Ann Arbor, MI & SPIE Optical Engineering Press, Bellingham, Washington, DC.
Smith W.J. (2007) Modern Optical Engineering, 4th ed., McGraw-Hill, New York.
Soleyn K. (2009) Development of a tunable diode laser absorption spectroscopy moisture analyzer for natural gas. In: Proceedings of the 5th International Gas Analysis Symposium, GAS 2009, Rotterdam, the Netherlands, February 11–13.
Spannhake J. , Schulz O. , Doll T. (2005) Design, development and operational concept of an advanced MEMS IR source for miniaturized gas sensor systems. In: Proceedings of 2005 IEEE Sensor Conference, Irvine, CA, pp. 762–765, October 30–November 3, 2005.
Tillman J.E. (1965) Water vapor density measurements utilizing the absorption of vacuum ultraviolet and infrared radiation. In: Ruskin R.E. (Ed.) Humidity and Moisture, Vol. 1, Principles and Methods of Measuring Humidity in Gases, Reinhold, New York, pp. 428– 443.
Tkachenko N.V. (2006) Optical Spectroscopy. Methods and Instrumentations. Elsevier, Amsterdam, the Netherlands.
Varier G. (1967) Experiment on the interaction of Ly-α radiation and atomic hydrogen. Appl. Opt. 6 (1), 167–171.
Vay S. , Anderson B.E. , Sachse G.W. et al. (1998) DC-8-based observations of aircraft CO, CH4, N2O, and H2O(g) emission indices during SUCCESS, Geophys. Res. Lett. 25, 1717–1720.
Vay S.A. , Anderson B.E. , Jensen E.J. et al. (2000) Tropospheric water vapor measurements over North Atlantic during subsonic assessment ozone and nitrogen oxide experiment (SONEX). J. Geophys. Res. 105, 3745–3756.
Wang W.E. , Michel A.P.M. , Wang L. , Tsai T. , Baeck M.L. , Smith J.A. , Wysocki G. (2014) A quantum cascade laser-based water vapor isotope analyzer for environmental monitoring. Rev. Sci. Instrum. 85, 093103.
Webster C.R. , Menzies R.T. , Hinkley E.D. (1988) Infrared laser absorption: Theory and applications. In: Measures R.M. (Ed.), Laser Remote Chemical Analysis, John Wiley & Sons, New York, Chapter 3.
Weinheimer A.J. , Schwiesow R.L. (1992) A two–path, two–wavelength ultraviolet hygrometer. J. Atmos. Oceanic Technol. 9, 407–419.
Weinstock E. , Schwab J. , Nee J.B. , Schwab M. , Anderdson J. (1990) A cryogenically cooled photofragment fluorescence instrument for measuring stratospheric water vapor. Rev. Sci. Instrum. 61 (5), 1413–1432.
Weinstock E.M. , Hintsa E.J. , Dessler A.E. , Oliver J.F. , Hazen N.L. (1994) New fast response photofragment fluorescence hygrometer for use on the NASA ER‐2 and the Perseus remotely piloted aircraft. Rev. Sci. Instrum. 65, 3544–3554.
Wen X.-F. , Sun X.-M. , Zhang S.-C. , Yu G.-R. , Sargent S.D. , Lee X. (2008) Continuous measurement of water vapor D/H and 18O/16O isotope ratios in the atmosphere. J. Hydrol. 349, 489–500.
Wen X.-F. , Lee X. , Sun X.-M. , Wang J.-L. , Tang Y.-K. , Li S.-G. , Yu G.-R. (2012) Intercomparison of four commercial analyzers for water vapor isotope measurement. J. Atmos. Oceanic Technol. 29, 235–247.
Werle P. (1996) Spectroscopic trace gas analysis using semiconductor diode lasers. Spectrochim. Acta A 52 (8), 805–822.
Werle P. (1998) A review of recent advances in semiconductor laser based gas monitors. Spectrochim. Acta A 54 (2), 197–236.
Werle P. , Slemr F. , Maurer K. , Kormann R. , Mucke R. , and Janker B. (2002) Near- and mid-infrared laser-optical sensors for gas analysis. Opt. Lasers Eng. 37, 101–114.
Werle P.W. (2004) Diode–laser sensors for in-situ gas analysis. In: Hering P. , Lay J.P. , Stry S. (Eds.) Laser in Environmental and Life Sciences, Springer-Verlag, Berlin, Germany, pp. 223–243.
Wernecke R. , Wernecke J. (2014) Industrial Moisture and Humidity measurement: A Practical Guide. Wiley-VCH, Weinheim, Germany.
White J.U. (1942) Long optical paths of large aperture. J. Opt. Soc. Am. 32, 285–288.
White S.E. , Cataluna M.A. (2015) Unlocking Spectral Versatility from Broadly−Tunable Quantum−Dot Lasers. Photonics 2, 719–744.
Wiederhold P.R. (1997) Water Vapor Measurement, Methods and Instrumentation. Marcel Dekker, New York.
Wilson A.C. , Barnes T.H. , Seakins P.J. , Rolfe T.G. , Meyer E.J. (1995) A low–cost, high–speed, near–infrared hygrometer. Rev. Sci. Instrum. 66 (12), 5618–5624.
WMO (1992) World Meteorological Organization: Measurement of Temperature and Humidity ( Wylie R.G. and Lalas T. Eds.). (Technical Note No. 194), (WMO–No. 75), Geneva, Switzerland.
WMO (2011) World Meteorological Organization: Technical Regulations. Volume I – General Meteorological Standards and recommended Practices, (WMO–No. 49), Geneva, Switzerland.
Wolfbeis O.S. (1991) Fiber Optic Chemical Sensors and Biosensors, Vol. 1. CRC Press, Boca Raton, FL.
Wolfbeis O.S. (1992) Fiber Optic Chemical Sensors and Biosensors, Vol. 2. CRC Press, Boca Raton, FL.
Wolfbeis O.S. (2006) Fiber-optic chemical sensors and biosensors. Anal. Chem. 78, 3859–3874.
Wolfbeis O.S. (2008) Fiber-optic chemical sensors and biosensors. Anal. Chem. 80, 4269–4283.
Wulfmeyer V. , Walther C. (2001) Future performance of ground-based and airborne water-vapor differential absorption lidar. I. Overview and theory. Appl. Opt. 40, 5304–5320.
Wunderle K. , Rascher U. , Pieruschka R. , Schurr U. , Ebert V. (2015) A new spatially scanning 2.7 μm laser hygrometer and new small–scale wind tunnel for direct analysis of the H2O boundary layer structure at single plant leaves. Appl. Phys. B 118, 11–21.
Yakir D. , Sternberg L.D.S.L. (2000) The use of stable isotopes to study ecosystem gas exchange. Oecologia 123, 297–311.
Yoshino K. , Esmond J.R. , Parkinson W.H. , Ito K. , Matsui T. (1996) Absorption cross section measurements of water vapor in the wavelength region 120 to 188 nm. Chem. Phys. 211, 387–391.
Yushkov V. , Lukjanov A. , Merkulov S. , Khaplanov M. , Shyshatzkaya L. , Gumbel J. (1995) Optical fluorescent hygrometer for water vapor low concentration measurements. SPIE Proc. 2506, 783–794.
Zoger M. , Afchine A. , Eicke N. et al. (1999) Fast in situ stratospheric hygrometers: A new family of balloon-borne and airborne Lyman photofragment fluorescence hygrometers. J. Geophys. Res. 104 (D1), 1807–1816.
Zuber A. , Witt G. (1987) Optical hygrometer using differential absorption of hydrogen Lyman–alpha radiation. Appl. Opt. 26 (15) 3083–3089.

## Use of cookies on this website

We are using cookies to provide statistics that help us give you the best experience of our site. You can find out more in our Privacy Policy. By continuing to use the site you are agreeing to our use of cookies.