The assumptions for equations (1) and (2) other than already expressed are that (a) steady-state conditions exist for time-averaged quantities, (b) horizontal gradients are insignificant so that all fluxes are vertical, and (c) the equations are applied in a zone known as the constant flux layer beneath the surface boundary layer, which grows nearly linearly with distance downwind from a surface property discontinuity. MEASUREMENT PRINCIPLES Relatively new developments in instrumentation have increased the feasibility of making dust-flux measurements by either (1) the eddy-correlation method or (2) the flux-gradtent method. The first is technically more difficult. Both methods were simultaneously demonstrated, perhaps for the first time, by Shinn et al., 1976. Gillette et al., 1972; Gillette, 1976; and coworkers have used the second method extensively for several years. The eddy-correlation method requires high-speed data acquisition and/or processing and is at present stilt in the prototype phase because of the requirement for a speed of response better than one second. The integrating nephelometer used by Shinn (loc. ett.) is unsuitable at low mass loadings and is marginal at high winds. Recently, Husar and Macias (1976) developed a diffusion charging, fast-response aerosol] detector which has been feasibility demonstrated for particle flux using the eddy-correlation method at Argonne National Laboratory, Wesley et ai., 1976. With these new developments, perhaps resources will be applied to meet our needs, which is to have an adequate tool for examining processes that rate limit Pu transport. E s z The flux-gradient method contains more critical assumptions than the first method and is more bulky and less useful as a probe for contaminated areas of limited extent. It is, however, more adaptable to long-term monitoring. Measurements of the gradient have to be performed in a relatively shallow zone beneath the surface boundary layer. The concentration gradient of dust being rapidly suspended (upward dust flux) has been found to conform to a power-law (Fig. 2). That is ya A a? so that ax dz Px, 3 A,P constants x | | , iL | \ a 66 1954 4 1955 Kansas and 4 Colorado e 1974, Nevada | . x 1974, Texas 0.3 In strong wind speed cases, it was estimated (Shinn et al., 1976) that the power P=-0.25, but much more data should be examined, In measurements of concentration gradients, more than two detectors are required (preferably more than three) and the detectors must be matched in response and calibration. Optical particle detectors are best; gravimetric methods are least satisfactory. ! J \a o 1974, Nevada 2 (4) _ \ . (5) In summary, both of the approaches to measure dust-flux have drawbacks, they are about equal in cost, and perhaps it is more satisfactory to operate both simultaneously. om \4 r 2 _ \ \° \ \ bo 0.3 1 \ 2 x] Fig. 2. Typical dust gradients with height during strong suspension events. Solid tine is a power law with exponent -0.25; dashed line 1s power law with exponent -0.35. 175