Soil moisture was measured gravimetrically at random selected locations within 50 m of the equipment installation to a depth of 2.5 cm mentioned Dust samples from the Bagnold sampler were dried at 100 C. earlier. for 24 hours prior to mechanical sieving through a sonic sifter to estimate fractions above and below 53 um diameter. RESULTS The reduced data for precipitation, soil moisture, windspeed, and temperature are presented in Table 1. Bagnold data, summarized in Tables 2 and 3, were expressed as dust flux (ug/em2/day) and Pu concentrations (pCi/g dry dust); the values represent estimated integral averages for each time interval. Table 4 also includes the fraction of the dust particles collected with diameters greater than 53 ym to compare with soil sampling results at the site (Nyhan et al., 1976). Dust flux was also estimated separately for ground creep events (C_), saltation pro- c cesses (Cl); and for total flux (C,). Multiple and simple linear and curvilinear regression analyses were applied to the dust flux data assuming a soil erosion relationship stating that the local wind erosion factor varies directly with the cube of the wind velocity, and inversely with the cube of the soil moisture content (Chepil and Woodruff, 1963). Additionally, the model includes an effect of air temperature near the ground on dust flux based on the relationship between heat flux from soil to air and soil moisture content via evapotranspiration (Geiger, 1965; Change et aZ., 1965; Baver, 1964). Heat flux from soil to air was assumed to be proportional to the fourth power of the absolute temperature of air near the ground because of its relationship to soil temperature at the surface of the ground, which is so dependent. Precipitation was assumed to affect dust flux through its effect on soil moisture status, soil crustal formation and dissolution, raindrop impact energy, and sheet water erosion in the lower collectors (Baver et al., 1972). Our initial hypothesis was that dust flux would be related to the parameters described above by a relationship of the form (Baver et al., 1972; Wischmeier et al., 1958): ~3,.3 Cog p _7 DyPT 4, MU (1) where C. st dust flux for specified collectors (ground creep (c), 7“ saltation (s), and total (t)) ug/em2/day (Table 2) by = proportionality constant (regression coefficient) 683