The Sarima Model In The Monthly Average Temperature Time Series Applications

Time series analysis method has a good application in the financial, industrial, weather and other fields,which can by analysing probability theory analysis of the limited sample, establish a certain accuracy model. Time series analysis method is of strong maneuverability in the practical application.In the meteorological statistical data, the average monthly temperature time series is a special kind of time series, influenced by the solar radiation, monsoon climate and other factors, and the average monthly temperature data show a marked seasonal changes. With the climate background of global warming, extreme weather and climate events frequently occur, puting forward higher requirements for the meteorological work. The research helps to reveal the solar radiation and other factors affect the near surface temperature, which is of important significance for the prediction of the average monthly temperature. This article will explore that the SARIMA model analyzes Xingtai monthly average temperature data, interpretates the characteristics of the data and is used to predict the monthly average temperatureWe will first in the introduction, introduced the technology research about the present situation and development of weather forecast in our country, focusing on the effects of long-term weather process of physical factors and the methods of the long-term weather forecasting, as well as the content, characteristics and innovation of the research.In the second chapter, the time series analysis method was introduced in this paper, and the Box-Jenkins modeling scheme was introduced in detail.In the third chapter, firstly chose Xingtai1954-2009monthly average temperature series SARIMA, modeling, and then considering that appearences of the global warming in the last few years, according to the analysis results of the average annual temperature anomaly map, chose1992-2009monthly average temperature series in Xingtai again, modeling, and expanded the range of the application to SARIMA model. Finally, by the method of multiple regression analysis, from the statistics data of ground observation, used eight data as independent variables into SARIMA modeling, the variable introduced, effectively improve the prediction accuracy of monthly average temperature.In the fourth chapter, try using combined model of BP neural network and SARIMA model, by using BP neural network and SARIMA model to study changing trend of the weather system, in order to reveal effect of the changes of weather system on monthly average temperature.Our study shows the important application to SARIMA model in the mean monthly temperature time series, constributing greatly to the data analysis and forecasting in the monthly average temperature on other stations...