Please use this identifier to cite or link to this item: https://hdl.handle.net/11681/9027
Title: Spectral measurements in a disturbed boundary layer over snow
Authors: Andreas, Edgar L.
Keywords: Atmosphere
Boundary layer
Turbulence
Snow
Issue Date: Nov-1987
Publisher: Cold Regions Research and Engineering Laboratory (U.S.)
Engineer Research and Development Center (U.S.)
Series/Report no.: CRREL report ; 87-21.
Description: CRREL Report
From the Introduction: Atmospheric turbulence is broad-banded. In the surface layer, fluctuations in the velocity components and in the scalar atmospheric constituents occur over a continuum of frequencies from less than 1/hour to well over 100 Hz. Computing spectra and cospectra from time series of these variables separates the contributions to the total signal variance (or covariance in the case of cospectra) by frequency. Thus, spectra and cospectra provide a picture of the turbulence process. When topographic inhomogeneities perturb the atmospheric boundary layer, such representations are especially useful for examining how and at what scale the topography influences the turbulence processes. Velocity and temperature spectra and cospectra have been measured often over horizontally homogeneous land surfaces (e.g., Kaimal et al. 1972, 1976, Wyngaard and Cote 1972, Tsvang et al. 1985). Humidity spectra have been measured less frequently over land (Miyake and McBean 1970, McBean 1971, Priestley and Hill 1985), because weak humidity fluctuations make the measurements difficult. Over the ocean, where fluctuations in humidity can be large, spectra of the velocity components and of the temperature and humidity have all been measured extensively (e.g., Pond et al. 1971, Leavitt 1975, Schmitt et al. 1979, Nicholls and Readings 1981, Smith and Anderson 1984). Over snow-covered surfaces, on the other hand, spectral measurements have been rare. In fact, most spectral measurements over snow have been made over snow-covered sea ice. Ranke and Smith (1971, 1973) and Andreas and Paulson (1979) reported velocity spectra and cospectra measured over compact Arctic sea ice; Smith et al. (1970) reported similar measurements in the Gulf of St. Lawrence. Teunissen (1980) measured spectra of the three velocity components over snow-covered ground in farming country near Toronto, Canada. Smith (1972) reported additional velocity spectra from the Gulf of St. Lawrence and included a few temperature spectra and cospectra. But only Thorpe et al. (1973) have measured both temperature and humidity spectra simultaneously over a snow-covered surface—again sea ice. Hicks and Martin (1972), MacKay and Thurtell (1978) and Yelagina et al. (1978) measured velocity, temperature and humidity fluctuations over snow-covered Lake Mendota and over snow fields in Ontario, Canada, and near Leningrad, espectively, but none reported individual spectra or cospectra, just covariance (integrals of the cospectra). The measurements that I report here are thus, evidently, the first extensive set of turbulence spectra and cospectra measured over snow-covered ground. Although the experimental site was fairly homogeneous for a couple hundred meters around my instruments, hills typically 100 m high bordered the site on two sides. Hence, all the measured longitudinal (u) and vertical (w) velocity spectra and the temperature (t) and humidity (q) spectra show inertial or inertial-convective subranges typical of the horizontally homogeneous near-field but contain excess energy at low frequency because of far-field topographic disturbances. The w-t, w-q and t-q cospectra generally behave in the inertial-convective subrange as cospectra collected over a horizontally homogeneous surface; the u-w cospectra usually have erratic sign throughout their frequency range, again suggesting just how much the topography affects the velocity field. The phase and coherence spectra corroborate the topographic effects by implying a dichotomy between low-frequency and high-frequency transfer processes. Because I have simultaneous measurements of the t and q spectra and the t-q cospectrum, it was possible to make theoretical computations of refractive index spectra for light of visible and millimeter wavelengths. Both sets of refractive index spectra have inertial-convective subranges. Hence, lastly, I compare refractive index structure parameters (Ct) obtained from the spectra for visible wavelengths with simultaneous scintillometer measurements of the same quantity.
URI: http://hdl.handle.net/11681/9027
Appears in Collections:CRREL Report

Files in This Item:
File Description SizeFormat 
CRREL-87-21.pdf972.52 kBAdobe PDFThumbnail
View/Open