Climate change is not a hot point worldwide, for it will generate various influences on the development of a nation, including social, economic and political problems. And for developing countries, the risk is relative high. In this essay, we build three models to identify whether a nation is fragile or not, evaluate the impact of climate change, and predict when and how a nation will be under high fragility. Meanwhile, we choose several nations both in and out of the top 90 nations with high fragility listed by Fragile State Index to test these models. Evaluation Model: In this model, we adopt AHP (Analytic Hierarchy Process) to build a model which can determine whether a country is fragile and measure the impact of climate change. We select 17 indicators as decisive variables to generate SPI, a parameter we use to illustrate the degree of fragility of. Then we choose four nations with moderate fragility to calculate the threshold of SPI, and find the threshold is 2.5987. Thus we conclude that the nation with SPI under the threshold is stable, otherwise, it is fragile. Validation Model: In order to validate the reliability of the evaluation model, we build a new model based on Entropy Weight Method(EWM) to test it. In this model, we define an parameter CPI to demonstrate the degree of vulnerability of a nation. And if the value of CPI of is above the threshold, we consider it is at high risk of fragility, otherwise, it is under stability. Adopting the same four nations to get the threshold, and get its value 25.0000. After many tests, we find the results are almost the same as the first model, which reveals the reliability of the model. Prediction Model: In this segment, we build a model through Grey Relation Analysis(GRA) to predict when a nation will be of fragility if it conducts improper development policies, and then we choose Argentina to carry out the prediction. Model implication in Syria: In this section, we conduct our model in Syria to find out how climate change has affected the fragility of the nation. Syria experiences a severe drought in recent years, so its social development has been changed a lot due to the climate shock. And we find Syria is now of high fragility according to the result of the model. Then we give advice based on its current situation to diminish the vulnerability. Model implication in Argentina: Here we adopt the prediction model to predict the trend of fragility in Argentina. Firstly, we collect its previous data to carry out the prediction model and find the result is of high reliable. Then we conduct the model to predict when this nation will become fragile, and the result shows that the time point is between 2027 and 2028. Fitting model in Japan :we can get Japan in 2017 intervention costs about 6469 Billion dollars. At the end of the essay, we analyze the sensitivity, and find these models we have built are reliable.