Search

Whether grazed semi-arid grassland can be regarded as a net-sink or net-source of aeolian dust is difficult to detect since deposition and emission processes are in gradual transition. In grassland, dust arrives from sources far away or is directly emitted and immediately part of the suspension load. The processes of dust emission or deposition can be mainly identified by vertical concentration measurements close to the ground and close to the sources. Often the concentration differences at the used measuring heights are too small to derive the direction of vertical fluxes precisely, especially at dust storms of light-to-medium intensities. For this reason, particle mass (PM, PM) and particle number concentrations were measured simultaneously at ground level and at heights of up to 80 m using a dust monitoring system lifted by a kite. The measured data of dust concentration gradients are used to calculate the vertical fluxes with the necessary certainty to derive the direction of the dust fluxes. Additionally, dust arriving from different directions and crossing different land use/land cover patterns was used to interpret the interaction of source/sink relationships along the transport direction upstream of the measuring point. Results show that dust measurements above grassland have to span greater height differences to derive vertical fluxes certainly, i.e., concentration gradients (Δ c/Δ z) of PM and PM provide an indication of the direction of the dust flux. Furthermore, the results indicate a close relationship between the particle size compositions of the dust and land use/cover patterns. Arable land in the surroundings causes an emission flux (vertical upward), especially of the coarser fractions >8 µm, whereas grassland causes deposition fluxes of size classes >2 µm during a dust storm of low intensity. Sources from far away are reflected by higher concentrations of finer particles, measured during a super-regional dust storm. Both, dust concentration gradients and particle size composition reflect the potential source areas and deposition processes adequately with regard to different land use/cover patterns up to 50 km away. [ABSTRACT FROM AUTHOR]

Aims: The aims of this study were to identify areas of wind erosion and dust deposition and to quantify the effects of different grazing intensities on soil redistribution rates in grasslands based on the Cs technique. Because the method uses a reference inventory as threshold for erosion or deposition, the classification of any other site as source or sink for dust depends on the accurate selection of this reference site. Methods: Measurements of Cs inventories and depth distributions were carried out at pasture sites with predominant species of Stipa grandis and Leymus chinensis which are grazed with different intensities. Additional measurements were made at arable land, plant-covered sand dunes and alluvial plains. Wind-induced soil erosion and dust deposition rates were calculated from Cs inventories by means of the 'Profile-Distribution' and the 'Mass Balance II' models. Results: The selection of the reference site was based on fluid dynamical and process-determining parameters. The chosen site should meet the following four conditions: (i) located at a summit position with obviously low deposition rates, (ii) sufficient vegetation cover to prevent wind erosion, (iii) plane to exclude water erosion and (iv) in the wind/dust shadow of a higher elevation. The measured reference inventory of Cs was 1967(±102) Bqm located at a summit position of moderately grazed Leymus chinensis steppe. The Cs inventories at other sites ranged from 1330 Bqm at heavily grazed sites to 5119 Bqm at river deposits, representing annual average soil losses of up to 130 tkm and deposits of up to 540 tkm, respectively. The calculated annual averages of dust depositions at ungrazed Leymus chinensis sites were related to the dust storm frequencies of the last 50 years resulting in a description of the temporal variability of annual dust depositions from about 154 tkm in the 1960s to 26 tkm at recent times. Based on this quantification already 80% of the total dust depositions can be related to the 20 years between the 1960s and the end of the 1970s and only 20% to the time between 1980 and 2001. Conclusions: Cs technique is a promising method to assess the effect of grazing intensity and land use types on the spatial variability of wind-induced soil and dust redistribution processes in semi-arid grasslands. However, considerable efforts are needed to identify a reliable reference site, because erosion and deposition induced by wind may occur at the same places. The combination of the dust deposition rates derived from Cs profile data with the dust storm frequencies is helpful for a better reconstruction of the temporal variability of dust deposition and wind erosion in this region. The calculated recent deposition rates of about 20 tkm are in good agreement with data of other authors. [ABSTRACT FROM AUTHOR]