No abstract

Abstract. A site at Owens Dry Lake was observed for more than 4 years. The site was a vegetation-free saline playa where the surface formed "ephemeral crusts," crusts that form after rainfall. Sometimes these crusts were destroyed and often a layer of particles on the crust would engage in vigorous aeolian activity. Three "phases" of active sand drifting are defined as almost no movement (extreme supply limitation), loose particles on crust with some degree of sand drift (moderate supply limitation), and unlimited source movement corresponding to a destroyed surface crust (unlimited supply). These "phases" occurred 45, 49, and 6% of the time, respectively. The accumulation of loose particles on the crust was mostly the result of in situ formation. Crusted sediments with loose particles on top can exhibit mass flux rates about the same as for noncrusted sediments. Crusted sediments limit or eliminate sand drift in two conditions: for rough crusts that effect a sufficiently high threshold friction velocity (above the wind friction velocity) and for limited amounts of loose particles on the crust where particle supply is less than would be transported in normal saltation for a thick sandy surface. These "supply-limited" cases are similar to wind erosion of limited spilled material on a hard concrete surface. We quantified "supply limitation" by defining a "potential" or "supply unlimited" sand drift function Q = A G where A represents supply limitation that decreases as the particle source is depleted. Here Q is the mass of sand transported through a surface perpendicular to the ground and to the wind and having unit width during time period t, and G = f 2 2 dt for u, > u,t. G is integrated for the same time period t as for Q, u, is the friction velocity of the wind, and u,t is the threshold friction velocity of the wind. Hard crusts (usually formed in the summer) tended to show almost no change of threshold friction velocity with time and often gave total protection from wind erosion. Rough crusts provided sufficient protection expressed as high threshold friction velocities. For these high threshold friction velocities, aeolian activity was greatly reduced or practically prevented. The softest crusts, usually formed in the winter, provided much less protection and sometimes were destroyed by the wind. Following this destruction the "potential" or "supply unlimited" sand drift would be observed. Background and ObjectiveThe movement of coarse material by the wind is widespread on Owens (dry) Lake in east central California. A discussion of sedimentary environments and dust generation processes at Owens (dry) Lake is given by Gill [1995] and Cahill et al. [1996]. However, the entire surface of the dry lake bed is not a homogeneous source of dust. B. Cox (personal communication, Bishop, California, 1998) describes the lake surface as having four distinct surface types. The first surface type is a

A joint United States/Russian/French collaborative experiment was undertaken in March 1993 and March 1996. Projects LODE I and II (Lake Owens Dust Experiments) took place on the anthropogenically desertified playa (dry lakebed) and surrounding regions of Owens Lake, in east-central California. One of the five parts of Project LODE was to determine relationships between optical depth and flux of dust emitted from the dry lake. Project LODE II included subsequent dust plume measurements and size distributions obtained through April 1996, to further refine the flux measurements for distinct mineral aerosol source regions at Owens Lake.Size distributions of dust aerosol were determined and aerosol optical depths were calculated from sunphotometer solar extinction measurements taken downwind in plumes coming from the emissive areas of Owens Lake. This source was visually observed for 10 measured dust storms. The plume mass was calculated to be 1Á5 Â 10 9 g using ground-based measurements and !1Á6 Â 10 9 g from satellite data. Project LODE II results were found to be consistent with LODE I results for the south end of the playa, but flux values were found to be reduced for the northeastern portion of the playa by comparison. Vertical flux values estimated by sunphotometry were found to be consistent with values estimated via a micrometeorological method.

No abstract