Research Of Meteorological Factors Impact On PM10Concentration From Straw Burning Source

Massive crop straws are frequently burning in annual harvest time, these actionswill release a great quantity inhalable particles （PM10） to cause serious air pullution,that will affect people the normal order. Research the influence of meteorologicalfactors and PM10concentration from straw burning source, which will provide basicscience support to solve the problems about straw burning and air pollution.Anhui province is the study area in this paper. According the prefectural data ofcrop yield and MODIS fire points in2012to analyse straw burning time variation andspatial distribution, then build the fire-point-modified burning ratio model. Using thismodel to calculate PM10emissions inventory from wheat straw burning. Among thePM10concentration, wind speed and precipitation, the interplay rules are analyzedwith SPSS software, and building multiple regression modle about PM10concentration. The main conclusions as follows:（1） Straw total resources is39.315million tons in Anhui province in2012bygrass valley ratio method, staple crop straw resources as follows: paddy（15.217Mt）,wheat（16.652Mt）, corn（5.370Mt）, rape（2.076Mt）. Regional straw resources: HuaibeiArea> Jianghuai Area> Jiangnan Area. There are five cities own rich straw resources,including Fuyang, Luan, Bozhou, Chuzhou and Suzhou.（2） Separating the MODIS fire point data by date and county with ArcGISsoftware tools, which contain Extract, Export, Merge, Attribute, Buffer. The total offire points is4517, the time of intensive points distribution are early May, early andmiddle June, late July and late November. The fire points maximum region is HuaiheArea in June and the fire point density is0.051per square kilometre. The regressionequation of county wheat planting area and fire points is Y=0.975X₁-0.418andthe regression equation of county wheat yield and fire points is Y=1.517X-2+1.642.（3） Calculating each county wheat yield fire point burning ratio derives fromB_m=B_t×N_i×（1.517W_m+1.642）^-1, then according to the specialized range torevise county burning ratio. Calculating the fire point modified burning ratio with theformula,which is B_c=(∑B_mi×W_mi)(∑W_mi)^-1.The compute PM10total emissionis32,000tons for wheat straw burning in Anhui province in2012with emission factormethod. The county PM10emission increases progressively from south to north. Bothof the high county PM10emission intensity and farmers per capita emission mostlyappeared at the eastren of Huaibei Area. （4） According to analysis with SPSS software, the mean PM10concentration of17monitoring points is between0.052to0.098mg/m³in Anhui province in2012,Hefei concentration is the highest city and Chizhou is the lowest one. Regional meanPM10concentration: Jianghuai Area> Huaibei Area>Jiangnan Area. Categorizing17monitoring points into as5classes with Hierarchical Clustering method, and thisresult is better to distinguish city PM10pollution. Choosing the PM10concentrationdata contain early January, April, July and middle September for counting citybackground concentration, and the result is between0.053to0.100mg/m³. Thecorrelation coefficient is0.457both of PM10concentration amplitude and numberof fire points in200km radius buffer, and correlativity at confidence level of0.05.（5） Most city PM10concentration and wind speed&precipitation are correlatednegatively. PM10concentration notches up small increase when wind speed <1m/s ordaily precipitation <1mm, and PM10concentration reduces markedly when windspeed>3m/s or daily precipitation>20mm.The mean PM10concentration of next dayis lower than rain day, this result shows the precipitation decontamination fuctionmarkedly. The best decontamination efficient of1mm precipitation is-0.00148mg·m^-3·mm^-1in Hefei,and the efficient is better than normal time in straw burningtime.（6） The PM10concentration multiple regression modle is set up, which includedependent variable of daily PM10concentration for wind speed, precipitation,hesternal PM10concentration as indepengent variables. The PM10concentrationmodel isY=0.068-0.013X₁-0.001X₂+0.542X3. According to independentsample test, the model fitted PM10concentration value can effectively predict trend,but the peak fitted value is lower than measured value and has the hysteretic quality.So the model needs further examination and revise in future.