Quantitative Analyses Of Dust Over Northeast Asian For Spring 2004 To 2005

In this paper, the quantitatively features of Northeast Asian dust events in spring are studied firstly. Using conventional surface meteorological records for March, April and May of 2004 and 2005, the distribution of dust-event frequencies and intensity are examined. The classification method based on observed dust concentration is studied. The results show that Takla Makan desert and Korean Peninsula were mainly occurred dust-in-suspension with the dust concentration in the range of 1000⁶000μg m^-3;Northeast China bored on Mongolia and southeast Mongolia were mainly occurred dust storms or severe dust storms and seldom occurred dust-in-suspension ,with the dust concentration in the range of 3000 9000μg m^-3;Badain Juran Desert, Tengger, Mu Us Deserts and the north part of Niaoning , Jilin were mainly occurred blowing dust and seldom occurred the other types of dust events, with the dust concentration in the range of 1000³000μgm^-3.The numerical model predicted dust concentration is quantitatively verified base on the TSP and PM₁₀ converted from observed horizontal visibility in the spring (March to May) of 2005. It is shown that numerical model have the better predict ability on the starting time and intention of dust events in the model domain and the single point station. After that, the numerical model predicted dust concentration is classified base on three kinds dust weather (severe dust storm, dust storm and blowing dust). Find the follow results: when severe dust storms occur, the predicted dust concentration is in the range of 10⁵¹0⁶μg m^-3;when dust storms occur, the predicted dust concentration is in the range of 10⁴10⁵μg m^-3;when dust-in -suspen- sion occur, the predicted dust concentration is in the range of 10³¹0Vgm^-3.We also study the relationship between the predicted dust concentration and the dust events classification. These experiments are the first step to provide some ideas about the verification of dust model and to make the model results much more direct and easily applied to the operational forecast to a certain extent.