USING SCINTILLATION AT 2 WAVELENGTHS TO MEASURE PATH-AVERAGED HEAT FLUXES IN FREE-CONVECTION

Local free convection scaling is one of the obvious triumphs of boundary-layer similarity theory. In free convection, there is no dynamic velocity scale; the sensible and latent heat fluxes, therefore, scale directly with the temperature and humidity structure parameters C(t)2 and C(q)2. By using scintillation to measure the refractive index structure parameter C(n)2 at two electromagnetic (EM) wavelengths, we can obtain C(t)2 and C(q)2 and, thus, in effect, measure path-averaged values of the sensible and latent heat fluxes. Here I describe this so-called two-wavelength method for free convection, derive quantitative guidelines for optimizing the method, and evaluate its potential accuracy. I show that the two-wavelength method works best when one EM wavelength is in the visible or infrared region and the other is in the millimeter or radio region. When the Bowen ratio is between -5 and -0.1 or between 0.1 and 5, the expected accuracy of the measured fluxes is +/- 10-20% - typical of what is possible with eddy-correlation measurements. With the two-wavelength method, however, the fluxes represent spatial averages.