ABSTRACT
ABSTRACT: In this paper, the multivariate Copula based methodology is presented for dust storm frequency analysis in Inner Mongolia. On account of the three basic conditions for the occurrence of dust storms, strong wind, rich source of dust and instability of the atmosphere, Xianghuangqi was taken as the case, the joint probability distribution of 500hpa upper longitudinal circulation index, maximum wind speed and soil moisture were established to calculate the joint return period. From the fitting goodness evaluation, Frank Copula has been identified as the best fitted copula for three-dimensional model. Compared with the fitting effect for bivariate model, the fitting is better on high parts. For the visualization expression and fitting precision, the three-dimensional hazards of natural disasters is acceptable.
