中文摘要：

参考温度变化过程的固有特性,利用1951年至2010年重庆市的温度数据,采用BP神经网络对温度进行预测估计,并采用蒙特卡洛方法对天气衍生品进行定价.仿真结果表明,该方法有效且相对误差较小,对开发重庆市的天气衍生品市场有一定的实用价值.

英文摘要：

Firstly,based on temperature inherent physical characteristic,taking temperature data of Chongqing from 1951 year to 2010 year as data set,BP artificial neural network model was used to estimate and forecast the temperature.Secondly,weather derivatives pricing was implemented by Monte Carlo method.The results of empirical simulation and model verification show that the model has relative minor error.It has important significance to the development of weather derivatives market in Chongqing.