Dustfall Source Areas Tracing During Dust Storms

Jiuquan is located in the inland basins of Northwest China,its major feature is dry and cold climate,scarce precipitation,bare ground and the prone dust weather.As a major industrial city in Hexi corridor,Jiuquan not only attacked by dust storm throughout the year,but also affected by human activities such as industry and transportation.Mineral dust aerosol often mixed with anthropogenic aerosol particles,they both interact with each other and increase the complexity of particle properties,eventually give a greater impact on the ecological environment and human health.So mineral dust source region identification and apportionment combined with enrichment analysis of human activities such industry and transportation have a far-reaching significance to improve the air quality of the industrial town in Hexi Corridor and construct a better western homeland,so it is urgent to carry out this work.This study combined HYPLIST-4 model,MODIS color satellite image and Fingerprinting method to identify the nature source of five dust storm/floating dust weather occurred in 2016 in Jiuquan.Besides,we used enrichment factor method,dust “oxide” model to determine anthropogenic pollutants degree.In this study,we analyzed 29 kinds of elements,including 4 kinds of nonmetal elements and the 25 kinds of metal elements.This paper selects the Al element as the reference element in enrichment factor analysis,we find dust aerosol in Jiuquan is polluted more seriously in winter and spring,this may be related to winter heating and inversion layer formation.According to the calculation results,we divided all the elements into 3 categories:(a)Zn and P are moderately enriched by human activities such as waste incineration,metal metallurgy,mining,coal mining.(b)Ba,Cr,Cu,Mn,Ca,Fe,V,Nd,Y,Na,Ni mainly come from nature,and the feature of most elements is stable,only slightly enriched by man-made pollution.(c)the EF values of the other 13 elements are less than 1,which are derived from the crust source only.In general,thedust fall dust samples were mainly contributed by natural sources and affected by a small amount of human activities.The calculation results of dust “oxide” model show that earth is the main contributor of dust fall,the contribution varied from 75.33% to82.58% and the average value was 78.50%,the contribution of industrial pollution sources varied from 1.35% to 3.09%,the average value was only 1.97%.Except the mineral source and industrial source,other human activities such as transportation,waste incineration contribute about 19.53% elements to atmosphere.In 5 dust storm/floating dust weather,the contribution of mineral dust was higher in summer and autumn,while the proportion of contribution rate was higher in autumn and winter.Through the HYSPLIT trajectory model and MODIS satellite images we found that the potential sources of dust storm and floating dust weather in Jiuquan are mainly distributed in the western,eastern and local areas.The probability of observation point received air mass from northwest,west,southeast and northeast respectively were 53%,20%,20% and 7%.Based on the analysis results above,the distinctive characteristics of the soil in different regions and the distribution of the sampling points we divided two source sets in our study.After K-W test and stepwise discriminant analysis,we selected 8 kinds elements including Ca,Ni,Y,Si,Cr,Mg,Ba,Sr et al to composite fingerprint factor in source set 1 and the correct classification rate can reach 82.80%,while the source set 2 did not pass the test.The results of multivariate mixed model show that in the dust storm/floating dust weather happened in February 27,March 20 and July 20,all of the mineral dust were totally contributed by local source area,northwest and northeast sources contributed nothing.In April 30 events,mineral dust were contributed by northwest and local sources,the contribution were 23.24% and 76.76% respectively.In the event happened in August2,the mineral dust was also contributed by the northwest source and the local source,the contribution reached 84.77% and 15.23% respectively.REM value in all calculation are less than 15%,which proves that the results of the mixed model are reliable.