1. Introduction
Measurement of humidity fluctuations over the sea from a fixed (e.g., tower) or a mobile (e.g., ship) platform is very difficult due to the problem of contamination of the various instruments used (salt, particles, aerosol contamination) and the problems of reliability. The accurate and reliable measurement of humidity fluxes and humidity fluctuations is nonetheless a very important goal when studying ocean–atmosphere exchanges, and measurements of the mean specific air humidity ratio of water vapor mass have been attempted for many years.
Different techniques and devices for estimating humidity have been developed since the beginning of the 20th century, including:
humidity absorption by hygroscopic substances to estimate absolute humidity,
wet bulb thermometry (measurement of the temperature difference between a humidified thermometer and a “dry reference thermometer”) or dewpoint measurements (dew temperature measurement when a dew deposit appears on a cooled surface) to estimate relative humidity.
A slow measurement of humidity can now be carried out in the lower atmosphere (above the ocean) for humidity variations between 20% and 100%, with a relative accuracy between 5% and 10%. For a comprehensive analysis of the state of art at the end of the 1970s, see McKay (1978) or Coantic and Friehe (1980). Saturation phenomena (hysteresis) due to precipitation or response changes related to pollution by particles are now relatively well understood. This is not the case for fast-response humidity measurements at the sea surface, which remain a challenge today. Humidity is often large at the sea surface, where an important amount of particulate, especially salt, contamination is present.
For mean humidity measurement, a coarse time response of 10 s or more is acceptable, but for measuring fluxes, one needs a finer time response. When using, for example, the inertial dissipation method, a time response of less than 100 ms is required, which corresponds to a distance scale of about 2.5 m at a wind velocity of 25 m s−1. A good sensitivity in the humidity fluctuation is also required, and a noise level typically smaller than 20 mg m−3 is a reasonable goal (Auble and Meyers 1992).
The two following techniques are mainly used to estimate humidity fluctuations (see, e.g., the paper by Priestley and Hill (1985):
Lyman-alpha systems that involve ultraviolet radiation absorption in the water vapor absorption band (Tillman 1965). These systems are very sensitive to salt contamination. Moreover, numerous calibrations must be done due to the drift of the Lyman alpha source;
Two different kinds of infrared analyzers relying on the passage of air through a tube or in situ measurements (Leuning and Judd 1996), namely:
closed-path sensors such as the LI-COR, Inc. Lincoln, Nebraska (LI-COR) 62xx series analyzers, which use the absorption of infrared radiation by water vapor (Hay 1980). Using tube intakes is much less sensitive to optical contamination effects but is rather slow;
open path sensors, such as the system described by Auble and Meyers (1992), the OPHIR Corporation, Littleton, Colorado (OPHIR) IR-2000 infrared absorption hygrometer, or the LI-COR’s recent LI-7500 gas analyzer. Optical window contamination by salt and particles is a serious drawback.
Humidity fluctuations can also be deduced from the refractive index measured with a microwave refractometer using a resonant microwave cavity (Gilmer et al. 1965; Ottersten 1969a). Resonant cavity refractometers were developed in the 1960s for studying clear-air turbulence and radio propagation (see, e.g., Coantic and Leducq 1969) and for comparing refractive-index fluctuations with clear-air radar echoes (Ottersten 1969b). The main drawback of the microwave radiometer system was associated with volumic variations of the cavity and, as described by Priestley and Hill (1985), a frequency cut-off located too low in the inertial subrange. Therefore, though the technique was promising for more than 30 yr, it has not been extensively used. In this paper, we analyze how after a few modifications such an instrument can be used to derive the latent heat flux.
First, we briefly describe the background necessary to understand how to obtain humidity and humidity fluxes from the refractive index. In the next section, we give a technical description of the Centre d’Étude des Environnements Terrestre et Planétaires (CNRS), France (CETP) refractometer and explain how previous technical difficulties have been overcome. We then show how the system has been used in an experiment on an oceanographic ship. Finally, an analysis of data collected during the severe conditions of the Couplage avec l’ATmosphère en Conditions Hivernales (CATCH) experiment in the North Sea (Eymard et al. 1999) and the FETCH (Flux, Etat de la mer et Télédétection en condition de fetCH variable) experiment in the Mediterranean Sea (Hauser et al. 2000) is presented. The method chosen and results are discussed. We conclude with the great potential of this new microwave refractometer for deriving humidity and fluxes.
2. Humidity measurement over the ocean
a. Background on humidity, refractive index, and fluxes
b. Relationship between index fluctuation and temperature and humidity fluctuation
δNQ/δNP = the ratio between the variation of N due to Q and the variation of N due to P,
δNQ/δNT = the ratio between the variation of N due to Q and the variation of N due to T.
3. Technical description of the CETP refractometer
a. Previous designs
The frequency measurement system was initially dependent on a sealed reference cavity and a frequency discriminator (Crain 1950) suitable for atmospheric measurements. Later, a voltage-controlled oscillator (VCO) was tuned to a submultiple of the sampling cavity resonance by means of a phase sensitive detector. The VCO frequency (≈100 MHz) was then counted (Vetter and Thomson 1971). A major simplification in refractometer design was carried out at Rutherford Appleton Laboratory, Chilton, United Kingdom. (R.A.L.) by eliminating the reference oscillator. A GaAs FET amplifier was simply connected in series with a sampling cavity (Yilmaz et al. 1983). A microwave frequency counter provided a time resolution of 10 ms.
We have tested this concept and unfortunately found that the system oscillated at a frequency that depended on the phase of the microwave amplifier transfer function. This phase is closely related to the amplifier saturation conditions. Therefore, it was not possible to ensure a low offset relative to zero phase, that is, to safely operate close to resonance.
b. The new design
We have brought together the best of the previous designs.
1) Sensor unit (Fig. 2)
A cavity made with the lowest expansion material was used. This solution has been previously experimented by Corning Glass (Gunderson and Smith 1968). We chose Zerodur-M glass ceramic from Schott Glaswerke (Mainz, Germany) because of its low temperature hysteresis (Haug et al. 1989). The machining and silver plus gold plating was made by Quartex.1 An already known and proven cavity geometry was used such as the one implemented by the R.A.L. The proportion of end surface area removed is close to 90%, as previously described (Thompson et al. 1959). The circuit comprises a VCO phase-locked on the resonance (Delahaye and Fournet–Fayas 1988).2 A mechanical design suitable for ground, ship, and airplane operation was employed.
2) Control unit
A microwave frequency counter was used, which was able to count a 9.4-GHz signal with a frequency resolution of 100 Hz and a time resolution of 10 ms and included a down converter in order to reduce the final counting range to 0–6 MHz. A high stability crystal oscillator was used to drive a 9.0–9.9-GHz frequency synthesizer with a 1-MHz step size. This synthesizer is useful to adjust the local oscillator frequency of the microwave frequency counter according to the dimensions of different cavity fabrication batches. It is also cheaper to add a synthesizer than to reduce machining tolerances.
In order to study and optimize the residual drift and to make the calibration easier, we have added the following:
two thermoregulated heating systems around the coaxial cables between the cavity and the electronics (Tcable) and close to the phase lock circuitry (TPLL),
an air temperature sensor (Tair) made with a 0.25 mm diameter type K thermocouple from Thermocoax (Suresnes, France) and an air pressure sensor (Pair) made with a diffused piezoresistive transducer from Endevco (San Juan Capistrano, California),
an additional temperature sensor made with a 1-mm diameter type K thermocouple successively mounted inside the glass ceramic material (Tglass) and on one of the coaxial to waveguide adapters (Ttrans),
analog and digital outputs on the control unit.
Two identical models have been built, one for airborne applications and one for shipborne use. The following specific improvements were carried out on the shipborne model:
In order to avoid condensation around the unventilated parts, dry air can be blown into the electronic box just behind the cavity from a pressure-regulated pump through a drilled SMA test connector.
The cleanliness of the cavity when the horizontal wind velocity is too low has lead to the addition of a suction fan in order to suck any particles through quickly. A thin stainless steel mesh attached at the front (Fig. 2) stops the larger ones.
3) Dynamic response to the airflow orientation
Using the refractometer for flux measurements on board a ship, the transfer function of the cavity and its angular response were analyzed, since it is not always possible to keep the sensor unit directed upwind. The effect of the presence of the different elements will be discussed below.
The distance response of the instrument was established to be of about Lr = 1 m, basedon experimental and theoretical studies (Gilmer et al. 1965), due to the shape and dimension of the system. Thus, the wavelengths smaller than Lr (and therefore frequencies higher than Fr = U/Lr, where U is the relative wind speed) are not considered.
The air velocity inside the refractometer is altered by the presence of the protection mesh at the upwind end and by the suction fan at the downwind end. In order to calibrate these effects, laboratory experiments were conducted in a wind tunnel at Institut de Recherche sur les Phénomènes Hors Équilibre (CNRS), France (IRPHE). The apparatus was put in the middle of the wind facility, and the air velocity was measured using small hot wire anemometers (DISA 55 M 01: 1 mm × 3 μm) inside the cavity, Uint, and outside the refractometer, U. Wind velocities from 0 to 15 m s−1 were applied. Several configurations were tested, with and without the protection mesh and the suction fan. Figure 3 shows the variation of Uint with U, with the mesh and the fan installed. At low wind speed (U < 5 m s−1), the suction fan induces an increase of the flow, and Uint is greater than U. At higher wind speed, the protection grid creates a discharge downstream of the refractometer entrance, inside the cavity, and the fan is not sufficient enough to evacuate the flow; therefore, Uint is lower than U.
The calibration curve Uint = f(U) is also shown in Fig. 3. The variable Uint depends on the horizontal incidence angle α (the ship is not always pointed into the wind direction, and, for rough sea states, α may vary significantly). In the laboratory wind tunnel, wind measurements were conducted with horizontal incidence angles ranging from 0° to 90°. Figure 4 shows the variation of Uint with alpha. There is clearly an effect at high winds and large incidence angle. However, the calibration curve is valid within 10% accuracy for incidence angles lower than 40°. In actual conditions, data with alpha greater than a threshold value of 30° were not considered. Although Uint is not used in the following computations, it appears that the flow is not disturbed in the angular ranges mentioned above, in spite of the obstacles. Thus, what is measured is indeed N, except for excessive angles. There is no index variation when the fan is turned off.
c. Performance of the new design
1) Dynamic performance
Flux measurements mainly require good dynamic performance. If the sample frequency is 50 Hz and the cut-off frequency of the antialiasing filter of the acquisition unit is 25 Hz, the response time of the instrument must be close to 15 ms. In the laboratory, we have adjusted the time constants of the phase locked loop in order to get this value for the rise and fall time, when a small ball is thrown through the cavity (Fig. 5). As the sample volume is not precisely defined on each open side of the cavity, this measurement is not accurate, but it gives an upper limit.
The airborne model flying at 100 m s−1 was used to check the high frequency response by direct comparison with a Lyman-alpha hygrometer (Nacass et al. 1995). On entry into a cloud top, we have observed similar output signals with a delay corresponding to the distance between the locations of both instruments (Fig. 6). The power spectrum from 0.01 to 12.5 Hz is the same for both, and no discrepancy is observed from the f−5/3 law during clear air conditions.
2) Static performance
The expansion coefficient of the cavity (≈10−6 K−1) is higher than expected. For applications that need an absolute value of humidity, this aspect is still under development. In the case of flux measurements, it is only required to check the very small influence of drift variations over the power spectrum. In a calibration chamber, we changed the temperature of dry nitrogen around and inside the cavity from 20° to 0°C and measured the difference ΔN between the experimental value of N and the theoretical one. We observed that ΔN varies roughly from 0 to −20 but with a delay of 5 min (Fig. 7). This long delay may be due to the low thermal conductivity of glass ceramics. This means that higher frequency variations of temperature are integrated and thus the drift variations induce negligible errors in the power spectrum over 0.01 Hz.
The dynamic performance corresponds to the slope of the transfer function (18). As N0 is negligible in comparison with 106, this slope is independent of N0 drifts. The static performance is linked to the term (106 + 2N0) and is directly related to the drift of N0. As this drift has a time constant of several minutes, the refractometer is well suited to applications requiring only variation measurements in a range starting at 0.01 Hz, such as flux determination using the Inertial Dissipative Method (IDM). High frequencies larger than 0.01 Hz are particularly adapted to IDM, since they are much less subject to distortion and ship motion effects than the EDM.
4. Experimental implementation
a. The installation of the refractometer
The shipborne refractometer has been successively installed on two different ships during two international campaigns. The CATCH experiment was part of the Fronts and Atlantic Storm-Track Experiment (FASTEX) program carried out in January and February 1997 in the North Atlantic (Joly et al. 1997; Eymard et al. 1999). The study of energy exchanges between the ocean and the atmosphere was performed from the R/V Le Suroit. The FETCH campaign took place in March and April 1998 in the Mediterranean Sea (Dupuis et al. 1999; Hauser et al. 2000). One of its objectives was the measurement and the parameterization of the turbulent fluxes at the interface between the ocean and the atmosphere. The two CETP refractometers were used, one on the Merlin IV aircraft and the other on the R/V L’Atalante.
On a large ship, a good accuracy for atmospheric measurements and especially humidity can be obtained only at the bow at some height above the deck (Blanc 1986; Nacass 1999). Therefore, atmospheric sensors have been mounted on the same 12-m mast on both ships (see Fig. 8 for the implementation on board L’Atalante). Three cables link the refractometer sensor and the control unit: a multiplepair cable for the power supply of the sensor and the fan and housekeeping data, a coaxial cable for the 9.4-GHz signal, and an incompressible plastic tube delivering low pressure dry air to the sensor. It would have been preferable to put the control unit close to the sensor, but a waterproof box able to evacuate heat produced by the frequency synthesizer was difficult to design; thus, long cables were used despite the fact that coaxial cable is somewhat expensive.
b. The surrounding instruments
The following atmospheric and radiation sensors were also installed at the top of the mast and at the same level as the Shipborne Refractometer Sensor (SRS):
a three-axis Ultrasonic Anemometer Sensor (UAS)“research vertical” from Gill Instruments Ltd. The UAS setup in FETCH (Fig. 9) is different from the setup in CATCH. We changed it because of some perturbation on vertical air velocity due to the support. The acquisition of the analog data outputs of the refractometer is carried out by this unit too;
two identical ventilated Young radiation shields with temperature and relative humidity sensors HMP 233 from Vaisala;
two sets of anemometers “wind monitor” from Young for wind speed and horizontal wind direction;
two digital barometers from Atmospheric Instrumentation Research and from Vaisala;
a global Solar Irradiance Sensor (SIS) CM3 and a far infrared irradiance sensor CG1 from Kipp and Zonen mounted on a universal joint;
an optical rain detector ORG 115 from Scientific Technology.
The above sensors are complemented for time and positioning with a GPS receiver SV6 from Trimble. Some sensors are located underwater including the following:
2 sensors that give the direction and speed relative to the surface in order to calculate the absolute wind parameters: a gyroscope Brown SGB 1000 installed on the gravity center shown in Fig. 7 and an electromagnetic two-axis knotmeter from Alma;
the sea surface temperature, which is measured at 5 m deep with a platinum hull thermometer fitted inside the sea water aspiration at the stem. The temperature close to the surface is then deduced by means of a model.
5. Data processing and analysis
a. Flux computation methodology
b. First results
c. Discussion
The refractometer is a very promising sensor for measuring latent heat flux over the ocean because of the very high quality of its high-frequency response, which is important using the IDM and because the sensor has been found to be very reliable and to allow continuous long-term measurements in extreme meteorological conditions such those encountered during the CATCH experiment. The instrument has some limitations however. The most severe limitation of the instrument is related to the perturbation by liquid water inside the cavity in certain meteorological conditions such as drizzle, rain, or even air saturated by water vapor. In these cases, the measurements of refractive index are no longer reliable and must be rejected. However, this limitation is common to most fast humidity sensors, and as mentioned in section 2, a system of air suction has been used to allow the refractometer to work very rapidly after the rain or drizzle has stopped. As far as sea spray is concerned, a rough estimate of the number of particles at the level of 17-m height can be estimated (following Makin 1998), and the temperature warming due to particle evaporation and of the consecutive flux perturbation in the cavity can be computed. Using a drastic hypothesis of a multiplicative effect of 100 in the probable number of particles due to ship dynamic effect gives (for a wind speed of 15 m s−1 and a friction velocity of 1 m s−1) a warming of 2.5 × 10−8 °C corresponding to a heat flux contribution of 4.1 × 10−4 W m−2, which is negligible. Though the sensor is theoretically not sensitive to such small contamination, some laboratory tests and intercomparison between other flux estimates computed with other instruments should be made to confirm these results.
As the refractometer is not a direct measurement of air humidity, average values of air temperature, humidity, and pressure as well as estimates of sensible heat flux are needed to provide latent heat fluxes [see Eq. (26)]. As shown in Fig. 1, regarding the standard Bowen ratio over the sea (≈0.2), the air refractive index flux depends mainly on latent heat fluxes. In almost all cases, bulk estimates of sensible heat flux can be used, but attention should be paid to particular situations with very moist air and/or high air–sea temperature differences. While processing CATCH and FETCH data, we found some samples with high Bowen ratios leading to very comparable contributions of heat and evaporation fluxes in the refractive flux. Figure 14 shows the major differences between these two experiments. Figure 14a shows that the two campaigns are characterized by different Bowen ratio. If a classical value of 0.2 is found to be representative for FETCH, a much higher value of 0.5 is obtained for CATCH. Moreover, although it is of smaller importance, a much higher variability of the ratio of δNQ/δNT is observed during CATCH than during FETCH (panel 14b). Consequently, during FETCH, a very small percentage of samples shows a significant impact of the sensible heat flux in the refractive index flux (less than a few percent of samples have |δNQ〈wq〉/δNT〈wt〉| less than 5), but this is no longer true for CATCH, where good quality sensible heat fluxes should also be computed for estimating latent heat flux from the air refractive index flux (about 30% of samples have |δNQ〈wq〉/δNT〈wt〉| less than 5) as shown in panel 14c. In these cases, one needs a more accurate estimate of sensible heat flux than using a simple bulk estimate. Unfortunately, as explained in detail in Eymard et al. (1999), during CATCH, the spectra of sonic temperature were found to be of very poor quality and therefore could not used here, contrary to FETCH. Therefore, the latent heat flux estimated during 30% of CATCH period is probably less accurate.
6. Conclusions
We have described a refractometer device, technically improved to analyze refractive index intensity and fluctuations over the sea. This system can be calibrated with a precision of six refractive index units and has been utilized with success to estimate shipborne humidity fluctuations and latent heat flux at sea during the very severe winter conditions of the CATCH 97 experiment in the North Atlantic Sea, with strong wind and sea state (particles projection), and during FETCH 98 in the Mediterranean Sea, with strong wind and salt deposition. The refractive index can be considered as a “passive scalar.” Its fluctuation density spectrum exhibits a −5/3 spectrum range between 0.1 and 10 Hz, providing a very high quality inertial subrange. The refractive index is primarily a function of humidity, less importantly of temperature and pressure. The last term is always negligible. A method is proposed to calculate the latent heat flux based on refractive index flux and an estimate of the sensible heat flux. The latter is generally found to be small enough to justify the use of a bulk parameterization. However, in some meteorological conditions, such as in 30% of CATCH cases, its contribution can be significant. The refractometer does not need calibration for refractive index variations. It is very resistant to salt and other contamination and looks very promising for future routine use to measure humidity fluxes in combination with a sonic anemometer:
The latent heat fluxes derived from CATCH and FETCH by the inertial dissipation method compare well with bulk estimates, and the system has been found to be very robust in high sea states. The good results obtained during these two experiments have allowed the refractometer system to be implemented in subsequent experiments involving French research vessels for obtaining real-time fluxes.
Acknowledgments
The authors thank CNET, INSU (CNRS), DRET, SHOM, IFREMER, European Community, and Meteo France who, during the CATCH and FETCH experiments, have given a support and the opportunity to test the shipborne refractometer. We particularly acknowledge the crews of R/V Le Suroit and R/V L’Atalante for their outstanding efforts.
REFERENCES
Auble, D. L., and T. P. Meyers, 1992: An open path, fast response infrared absorption gas analyzer for H2O and CO2. Bound.-Layer Meteor.,59, 243–256.
Birnbaum, G., 1950: A recording microwave refractometer. Rev. Sci. Instrum.,21, 169–176.
Blanc, T. V., 1986: The effect of inaccuracies in weather-ship data on bulk-derived estimates of flux, stability and sea-surface roughness. J. Atmos. Oceanic Technol.,3, 12–26.
Coantic, M., and D. Leducq, 1969: Turbulent fluctuations of humidity and their measurement. Radio Sci.,4, 1169–1174.
——, and C. A. Friehe, 1980: Slow response humidity sensors. Air-Sea Interaction: Instruments and Methods, F. Dobson, L. Hasse, and R. Davis, Eds., Plenum Press, 399–411.
Crain, C. M., 1948: The dielectric constant of several gases at a wave-length of 3.2 centimeters. Phys. Rev.,74, 691–693.
——, 1950: Apparatus for recording fluctuations of the refractive index of the atmosphere at 3.2 centimeters wave-length. Rev. Sci. Instrum.,21, 456–457.
DeCosmo, J., K. B. Katsaros, S. D. Smith, R. J. Anderson, W. A. Oost, K. Bumke, and H. Chadwick, 1996: Air–sea exchange of sensible heat and water vapor: The HEXOS results. J. Geophys. Res.,101, 12 001–12 016.
Delahaye, J. Y., and C. Fournet-Fayas, 1988: Appareil de mesure de la fréquence de résonance d’une cavité hyperfréquence. French patent 88-10127, France (Centre National d’Études des Télécommunications, Issy-les-Moulineaux, France).
Dupuis, H., P. K. Taylor, A. Weill, and K. B. Katsaros, 1997: The inertial dissipation method applied to derive momentum fluxes over the ocean during the SOFIA/ASTEX and SEMAPHORE experiments with low to moderate wind speeds. J. Geophys. Res.,102 (C9), 21 115–21 129.
——, C. Guérin, A. Weill, and D. Hauser, 1999: Heat flux estimates by the inertial dissipation method during the FETCH experiment. The Wind Driven Air-Sea Interface: Electromagnetic and Acoustic Sensing, Wave Dynamics and Turbulent Fluxes, M. L. Banner, Ed., 297–304.
Eymard, L., and Coauthors, 1999: Surface fluxes in the North Atlantic current during CATCH/FASTEX. Quart. J. Roy. Meteor. Soc.,125, 3563–3599.
Gilmer, R. O., R. E. McGavin, and B. R. Bean, 1965: Response of NBS microwave refractometer cavities to atmospheric variations. Radio Sci.,69D, 1213–1217.
Gossard, E. E., 1960: Power spectra of temperature, humidity and refractive index from aircraft and tethered balloon measurement. IEEE Trans. Antennas Propag.,3, 186–201.
Gunderson, L. C., and D. M. Smith, 1968: A highly temperature-stable microwave resonator. Appl. Opt.,7, 805–812.
Haug, R., A. Klaas, W. Pannhorst, and E. Rodek, 1989: Length variation in Zerodur M in the temperature range from −60°C to +100°C. Appl. Opt.,28, 4052–4054.
Hauser, D., H. Dupuis, X. Durrieu de Madron, C. Estournel, C. Flamand, J. Pelon, P. Queffelou, and J. M. Lefèvre, 2000: La Campagne FETCH; étude des échanges océan/atmosphère dans le Golfe du Lion. La Météor.,8, 14–31.
Hay, D. R., 1980: Fast-response humidity sensors. Air-Sea Interaction: Instruments and Methods, F. Dobson, L. Hasse, and R. Davis, Eds., Plenum Press, 413–432.
Joly, A., and Coauthors, 1997: The Fronts and Atlantic Storm-Track Experiment (FASTEX): Scientific objectives and experimental design. Bull. Amer. Meteor. Soc.,78, 1917–1940.
Kistler, A. L., 1959: Fluctuation measurements in a supersonic turbulent boundary layer. Phys. Fluids,2, 290–296.
Large, W. G., and S. Pond, 1981: Open ocean momentum flux measurements in moderate to strong winds. J. Phys. Oceanogr.,11, 324–336.
——, and ——, 1982: Sensible and latent heat flux measurement over the ocean. J. Phys. Oceanogr.,12, 464–482.
Larsen, S. E., J. B. Edson, C. W. Fairall, and P. G. Mestayer, 1993: Measurement of temperature spectra by a sonic anemometer. J. Atmos. Oceanic Technol.,10, 345–354.
Leuning, R., and M. J. Judd, 1996: The relative merits of open- and closed-path analysers for measurement of eddy fluxes. Global Change Biol.,2, 241–253.
Makin, V. K., 1998: Air-sea exchange of heat in the presence of wind waves and spray. J. Geophys. Res.,103 (C1), 1137–1152.
McKay, D. J., 1978: A sad look at commercial humidity sensors for meteorological applications. Preprints, Fourth Symp. on Meteorological Observations and Instrumentation, Denver, CO, Amer. Meteor. Soc., 7–14.
Nacass, P., 1999: Utilisation de la modélisation des écoulements aérodynamiques pour la mesure atmosphérique à bord du navire océanographique L’Atalante. Tech. Rep. Météo-France/CNRM/CAM 22, 132 pp. [Available from Centre d’Aviation Météorologique, Brétigny, France.].
——, C. Abonnel, H. Bellec, M. Morera, and J. Y. Delahaye, 1995:Comparative measurements of in-flight humidity sensors on the Meteo-France Merlin-IV. Preprints, Ninth Symp. on Meteorological Observations and Instrumentation, Charlotte, NC, Amer. Meteor. Soc., 430–433.
Ottersten, H., 1969a: Atmospheric structure and radar backscattering in clear air. Radio Sci.,4, 1179–1193.
——, 1969b: Radar backscattering for the turbulent clear atmosphere. Radio Sci.,4, 1251–1255.
Priestley, J. T., and R. J. Hill, 1985: Measuring high-frequency humidity, temperature and radio refractive index in the surface layer. J. Atmos. Oceanic Technol.,2, 233–251.
Rabin, R. M., and R. J. Doviak, 1989: Meteorological and astronomical influences on radar reflectivity in the convective boundary layer. J. Appl. Meteor.,28, 1226–1235.
Smith, D. S., and Coauthors, 1988: Sea surface wind stress and drag coefficients: The HEXOS results. Bound.-Layer Meteor.,60, 109–142.
Smith, E. K., and S. Weintraub, 1953: The constants in the equation for atmospheric refractive index at radio frequencies. Proc. IRE,41, 1035–1037.
Straiton, A. W., 1964: Measurement of the radio refractive index of the atmosphere. Advances in Radio Research, J. A. Saxton, Ed., Academic Press, 2–52.
Stull, R. B., 1988: An Introduction to Boundary Layer Meteorology. Kluwer Academic Publishers, 666 pp.
Thayer, G. D., 1974: An improved equation for the radio refractive index of air. Radio Sci.,9, 803–807.
Thompson, M. C., Jr., F. E. Freethey, and D. M. Waters, 1959: End plate modification of X-Band TE011 cavity resonators. IRE Trans. Microwave Theory Tech.,7, 388–389.
Tillman, J. E., 1965: Water vapor density measurements utilizing the absorption of vacuum ultraviolet and infrared radiation. Humidity and Moisture, A. Wexler, Ed., Vol. 1, Reinhold, 428–443.
Vetter, M. J., and M. C. Thompson Jr., 1971: Direct-reading microwave refractometer with quartz-crystal reference. IEEE Trans. Instrum. Meas.,20, 58.
Yelland, M. J., and P. K. Taylor, 1996: Wind stress measurements from the open ocean. J. Phys. Oceanogr.,26, 541–558.
Yilmaz, U. M., G. R. Kennedy, and M. P. M. Hall, 1983: A GaAs FET microwave refractometer for tropospheric studies. Proc. AGARD Conf., CP-346, Norway, 9.1–9.5.
Variation of δNQ/δNP and δNQ/δNT with the Bowen ratio for RH = 50% and 100%
Citation: Journal of Atmospheric and Oceanic Technology 18, 3; 10.1175/1520-0426(2001)018<0459:ANSMRF>2.0.CO;2
The 3D representation of the shipborne refractometer sensor
Citation: Journal of Atmospheric and Oceanic Technology 18, 3; 10.1175/1520-0426(2001)018<0459:ANSMRF>2.0.CO;2
Variation of the air velocity Uint inside the cavity when a wind velocity U is applied (horizontal wind incidence α = 0), measured in a wind tunnel at IRPHE
Citation: Journal of Atmospheric and Oceanic Technology 18, 3; 10.1175/1520-0426(2001)018<0459:ANSMRF>2.0.CO;2
Variation of air velocity Uint inside the cavity with the horizontal wind incidence angle α, measured in a wind tunnel at IRPHE
Citation: Journal of Atmospheric and Oceanic Technology 18, 3; 10.1175/1520-0426(2001)018<0459:ANSMRF>2.0.CO;2
Variation of the VCO input voltage when a small ball is thrown through the cavity
Citation: Journal of Atmospheric and Oceanic Technology 18, 3; 10.1175/1520-0426(2001)018<0459:ANSMRF>2.0.CO;2
Comparison between time series measured by a Lyman–alpha hygrometer and the refractometer on entry into a cloud top
Citation: Journal of Atmospheric and Oceanic Technology 18, 3; 10.1175/1520-0426(2001)018<0459:ANSMRF>2.0.CO;2
Comparison of temporal variations of the refractometer error with the temperature during a sudden cooling of the sensor when ventilated with dry nitrogen
Citation: Journal of Atmospheric and Oceanic Technology 18, 3; 10.1175/1520-0426(2001)018<0459:ANSMRF>2.0.CO;2
Sketch of the arrangement of the atmospheric instrumentation during the FETCH experiment
Citation: Journal of Atmospheric and Oceanic Technology 18, 3; 10.1175/1520-0426(2001)018<0459:ANSMRF>2.0.CO;2
Atmospheric sensors on top of the 12-m mast during tests before the FETCH experiment: Ultrasonic Anemometer Sensor (UAS), Shipborne Refractometer Sensor (SRS), and Solar Irradiance Sensor (SIS) (photograph by G. Lachaud)
Citation: Journal of Atmospheric and Oceanic Technology 18, 3; 10.1175/1520-0426(2001)018<0459:ANSMRF>2.0.CO;2
Block diagram showing the various sensors and acquisition units of the atmospheric instrumentation on the R/V L’Atalante during the FETCH experiment
Citation: Journal of Atmospheric and Oceanic Technology 18, 3; 10.1175/1520-0426(2001)018<0459:ANSMRF>2.0.CO;2
Example of normalized spectra of air refractive index as a function of (a) frequency or (b) wavelength. The spectra correspond to a large range of wind speed and air temperature minus sea temperature or humidity difference
Citation: Journal of Atmospheric and Oceanic Technology 18, 3; 10.1175/1520-0426(2001)018<0459:ANSMRF>2.0.CO;2
(a), (b), and (c) Time series of latent heat flux (solid line: bulk empirical formulae; circles: refractometer measurements as described in section 5.1), wind speed, and air plus sea surface temperature, respectively. These data concern the CATCH/FASTEX campaign
Citation: Journal of Atmospheric and Oceanic Technology 18, 3; 10.1175/1520-0426(2001)018<0459:ANSMRF>2.0.CO;2
Same as Fig. 12 but for the FETCH campaign
Citation: Journal of Atmospheric and Oceanic Technology 18, 3; 10.1175/1520-0426(2001)018<0459:ANSMRF>2.0.CO;2
Histograms of (a) Bowen ratio, (b) |δNQ/δNT|, and (c) |δNQ〈wq〉/δNT〈wt〉| for the CATCH and FETCH campaigns
Citation: Journal of Atmospheric and Oceanic Technology 18, 3; 10.1175/1520-0426(2001)018<0459:ANSMRF>2.0.CO;2
