Capturing the spatio-extreme dependency, the joint probability that an extreme disaster happens among several locations, is of high interest to policy makers and environmental risk managers since plans of resource allocation for disasters are all based on this information. Max-stable Process, a spatial application of Extreme Value Theory, is frequently used for this propose. However, estimating Max-stable Process is extremely difficult due to its clumsy structure and much computational power is involved. For our research, we would like to introduce a new estimation method of the Max-stable process so that the statistical efficiency and computational efficiency can be balanced. Several environmental applications are demonstrated to show the effectiveness of my method.