Abstract: The paper introduces disaster entropy based on the theory of information entropy and assesses the uncertainty on the disaster occurrence. The theory of information entropy is used for making certain of the uncertainty of the information datum, just like the theory of information entropy, the disaster entropy is used for making certain the importance sectors on disaster incidents and build the calculation model about the quantity of disaster information. Based on many years statistical datum on sand disaster in five different regions (that is Jinhe, Dabanchen, Hami, Trupan, Shanshan) in Xinjiang the paper calculates the frequency of sand disaster and the disaster entropy and make some analysis on the results. It shows that the value of the total disaster entropy in Hami is the biggest and that in Dabanchen is the smallest, the difference in the entropy value between Hami and Dabanchen could reach 7. 73.This result also shows that the disaster sectors in Hami region are not only bigger than that of other regions in number but also much great than in other regions in intensity. It makes certain the degree of disaster intensity in quantity and shows whether sectors resulting to disaster is few or not in quality by applying the model of disaster entropy. As a conclusion, the paper also discusses something about the usage on the disaster entropy. The paper put forward zero-disaster phenomenon and conclude this phenomenon into the calculation on the disaster entropy. As a parameter function, the choice about W (Pi) is important. At present, the definition of this parameter function adopts two solutions：the first is provided by experts on disaster while the second is obtained by the division on statistical datum.

Key words: Disaster Entropy, Sand Disaster, Xinjiang