Click for next page ( 142


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 141
instrumentation Development Needs for Use of Mass-Balance Technique BY D. LENSCHOW In Part I of this report, several of the investigations of the proposed research program require the measure- ment of fluxes of chemicals. Such measurements are required in the mass-balance or box-budget experi- ments proposed under the general heading "Noncyclic Transformation and Removal," and also for field stud- ies of dry deposition and for advanced experiments un- der the heading "Sources of Atmospheric Chemicals, Biological and Abiological." In~this brief section, we outline the principles of such measurements and the general requirements they place on instrumentation. The budget for the mean concentration of a species s can be written as i dxi in s s (8.1) where the overbar denotes an average over a time or horizontal distance long enough to obtain a stable esti- mate, and the prime denotes a departure from the mean. For aircraft flux measurements, the time or dis- tance over which a mean is calculated is typically at least 5 min or 25 km. This equation states that the time rate of change, plus the mean advection of s, plus the vertical turbulent flux divergence of s is equal to the mean inter- nal sources Ps and sinks Ls of s. Integrating vertically and applying (8.1) to a convective boundary layer that is vertically well mixed to a depth h and capped with a well- defined inversion layer, 141 8 h at + OCR for page 141
142 to direct eddy correlation measurements (Hicks et al., 1980~. These techniques include (1) box methods, where an enclosure is placed over a surface of interest and the time rate of change of the species concentration is measured; and (2) profile or gradient methods, where differences in concentration at several levels close to the surface can be related to flux at the surface. These meth- ods require only mean concentration measurements. However, the box method assumes that the enclosure does not affect the surface flux, or at least that such effects can be corrected, and the profile method requires accurate measurements of concentration differences (on the order of 1 percent of the mean concentration or better), as well as supporting micrometeorological mea- surements to determine the surface stress and stability. Further details on use of different platforms and sensors for measurements of fluxes and other micrometeorolog- ical variables are presented by Dobson et al. (1980) and in a special issue of Atmosph~r~c Technology. ~ Absolute accuracy for turbulence measurements is not important, as long as the sensor output does not drift significantly during the period of a flux-measuring run (typically about 10 min), and the sensor gain is known accurately. This is because a flux measurement is calcu- lated from departures from a mean value; the mean of each quantity is not used in the calculation of a vertical turbulent flux. However, the fluctuations must be re- solved. In a convective boundary layer where the fluctu- ations are generated by a flux at the surface, the mea- sured resolution of a trace species should be at least about 10 percent of its surface flux divided by the con . . vect~ve ve oc~ty, w* = [(glT) (W'8~)oh]~/37 where (w'b')O is the surface virtual temperature flux and g/T is the buoyancy parameter (gravitational accelera- tion divided by the mean temperature). Typically, w* is of the order of 1 m/s. In order to resolve the eddies important for vertical transport through most of the convective boundary layer, a sensor must resolve wavelengths at least as small as 30 m and preferably 5 m. Close to the surface, the requirements are even more stringent since there the eddies scale with height above the ground. At 10-m height, for example, wavelengths as small as 3 m must be resolved accurately to measure all the significant con- tributions to the turbulent flux (Kaimal et al., 19723. For a stably stratified boundary layer (e. g., the clear-air "Instruments and Techniques for Probing the Atmospheric Boundary Layer." Atmospheric Technology No. 7, 1975, edited by D. H. Lenschow. Available from National Center for Atmospheric Research, PO. Box 3000, Boulder, CO 80307-3000. PART II ASSESSMENTS OF CURRENT UNDERSTANDING nocturnal boundary layer over land), somewhat smaller spatial resolution is required to resolve the turbulent fluctuations. A sample rate of 50 per second would be desirable if aircraft measurements are proposed for this situation. For measurements above the boundary layer, in the free troposphere and lower stratosphere, a slower sam- pling rate seems adequate. Possibly two per second would be satisfactory. If at all possible, however, a higher rate (say, 5 to 10 per second) should be used. Whatever rate is used, it should be remembered that the Nyquist frequency (i.e., the maximum frequency that can be resolved by a spectral decomposition of the data) is half the sampling rate. Thus a 20-per-second sampling rate will resolve 10-Hz spectral variables. At an aircraft speed of 100 m/s, this is equivalent to a 1 O-m wavelength. Another factor to remember is that a first-order in- strument time constant must be multiplied by ~2,r before taking the reciprocal in order to estimate the response of the instrument in the frequency domain. Thus, if one has a first-order time constant of 0.1 s, the amplitude of its variance is reduced to 72 percent of the variance of the input signal at a frequency of 1 Hz. For flux calculations, the phase angle between the input and the output is also important. At 1 Hz, the output signal lags the input by 32 in the above example. This lag reduces the contribution to the eddy flux at 1 Hz to 85 percent of the input signal. This phase-angle require- ment means that sampling of variables must be simulta- neous, or at well-defined time intervals so that correc- lions can be made, if necessary, to match the phase angles of variables before calculating fluxes. For flux measurements, measurement noise that is not correlated with the vertical velocity does not contrib- ute to the vertical flux. The noise may, however, necessi- tate a longer averaging time. This is important to keep in mind for measurements of trace gases that involve counting a limited number of photons. In this case, the noise is likely to have a flat "white noise" spectrum with a Poisson distribution, while the signal spectrum typi- cally decreases with frequency with a-5/3 slope. Thus the noise may surpass the signal above some frequency that depends on the magnitude of both the signal and the noise. By evaluating the left side of (8.1), it is possible to estimate the net internal production or loss of s (Ps and Ls) This has been done to estimate the net photochemi- cal production of O3 in the boundary layer (Lenschow et al., 1981). In addition, the species flux at low levels in the boundary layer is a direct measurement of surface deposition (or emission) if chemical reactions are not significant on a time scale of a few hundred seconds or less. Even if the lowest measurement level is several tens

OCR for page 141
INSTRUMENTATION DEVELOPMENT NEEDS of meters above the surface, the surface flux may be estimated by extrapolating to the surface a flux profile obtained from measurements at several levels through the boundary layer, since the flux profile is linear for a conserved species in a horizontally homogeneous well- mixed boundary layer. Because of the usefulness of resolving turbulent fluc- tuations of trace species in the boundary layer, it is im- portant to develop this capability for a larger number of species. In this way, measurements of chemical and pho- tochemical production and loss, surface sources and sinks, and transport through and across the top of the boundary layer can be obtained for direct comparison with model predictions, or as fundamental data in them- selves. In evaluating the mean concentration budget, the horizontal mean advection term is obtained by measur- ing the horizontal gradient of a species s. The required accuracy of this measurement can be estimated by as- suming that the removal at the surface is equal to the horizontal advection term. Accuracy of surface removal can be specified in terms of a deposition velocity, which is defined as va' = - (w's')O/s. Equating the surface removal to the advective term in (8.2), one obtains s ~ ( uh ) (8.3) where L is the horizontal distance across which the dif- ference bs-s2 -so is measured. As an example to illustrate the magnitude of the horizontal changes ex- pected in an aircraft experiment, let L = 105 m, u = 5 m/ s, and h = 103 m. Thus as =20tsm-~]v~. (8.4) For many species, a reasonable accuracy goal for mea- suring vet is 5 x 10-4 m/s. Thus bs/s = 1 percent. In many cases, L can be increased by as much as a factor of 10. Therefore, bs/s = 1 to 10 percent. A potential alternative to eddy correlation flux mea- surements is implementation of the eddy accumulation technique (Desjardins, 1977~. In this technique, mean concentrations of trace species in two gas samples are measured. The rate of flow ofthe sampler is controlled to be proportional to the magnitude of the vertical air ve- locity. One sample is obtained from upward moving air, and the other is obtained from downward moving air. The difference in concentrations between the two air- streams is proportional to the vertical flux. The main advantage of this technique is that fast-response concen- tration measurements are not required for flux mea- surements; instead, fast-response, accurate, and sensi- tive air flow control is necessary. However, very accurate mean concentration differences are required. These lat 143 ter two requirements may preclude application of this technique for many species, particularly if their removal rate is small. Surface-tower techniques provide an important com- plement to the aircraft methods described above. Whereas aircraft provide direct measurements of spatial averages of dry deposition fluxes, tower instrumenta- tion provides a more detailed investigation of the factors that control these fluxes on a time-evolving basis. A comprehensive study would necessarily involve both techniques. Instrumentation developed to meet the re- quirements for aircraft eddy flux applications will also satisfy the requirements for tower operation. In some cases, however, the requirement for rapid response can be relaxed slightly, and sometimes it can be replaced by a demand for extremely accurate difference measure- ments. This is the case if the desire is for instruments suitable for measurement of concentration gradient in- stead of covariance. One of the important applications of direct flux mea- surements is to provide detailed knowledge of deposition velocities for various species, and the variables that de- termine them. These deposition velocities can then be used in numerical models to parameterize surface fluxes. Field verification of these modeling studies re- quires concentration data from a network of surface observation sites. Simple, but reliable sampling meth- ods need to be developed for this purpose. Methods analogous to high-volume filtration for airborne parti- cles appear to offer special promise. Such methods are already in operation in some networks (e.g., in Canada and Scandinavia), and the methods need to be im- proved to permit routine and inexpensive operation on a global basis. BIBLIOGRAPHY Desjardins, R. E., 1977. Energy buclget by an eddy correlation method.J. Appl. Meteorol. 16:248-250. Dobson, F., L. Hasse, and R. Davis, 1980. Air-Sea Interaction Instru- mentsar~dMethods. Plenum, New York, 801 pp. Hicks, B. B., M. L. Wesely, and J. L. Durham, 1980. Critique of methods to measure dry deposition: Workshop summary. EPA- 600/9-80-050. Environmental Protection Agency, Washington, D.C., 83 pp. Kaimal,~J. C., J. C. Wyngaard, Y. Izumi, and O. R. Cote, 1972. Spectral characteristics of surface-layer turbulence. Quart. J. Roy. Meteorol. Soc. 98:563-589. Lenschow, D. H., R. Pearson, Jr., and B. B. Stankov, 1981. Esti- mating the ozone budget in the boundary layer by use of aircraft measurements of ozone eddy flux and mean concentration. J. Geophys. Res. 86:7291-7297. Lenschow, D. H., R. Pearson, Jr., and B. B. Stankov, 1982. Mea- surements of ozone vertical flux to ocean and forest. I. Geophys. Res. 87:8833-8837. Lilly, D. K., 1968. Models of cloud-topped mixed layers under a stronginversion. Quart. J. Roy. Meteorol. Soc. 94:292-309.