Time Series Forecasting of Temperatures using SARIMA: An Example from Nanjing

2018 5th International Conference on Advanced Composite Materials and Manufacturing Engineering

Chen, Peng^1 ; Niu, Aichen^1 ; Liu, Duanyang^2 ; Jiang, Wei^3 ; Ma, Bin^1

Jiangsu Meteorological Information Center, Nanjing, China^1

Jiangsu Meteorological Observatory, Nanjing, China^2

Jiangsu Meteorological Climate Center, China^3

关键词: Forecasting accuracy;    Mean temperature;    Model Selection;    Nanjing , China;    Research approach;    Seasonal autoregressive integrated moving averages;    Time series forecasting;    Time-series modelling;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/394/5/052024/pdf
DOI  :  10.1088/1757-899X/394/5/052024

来源: IOP

【 摘 要 】

Time series modelling and forecasting - a method that predicts future values by analysing past values - plays an important role in many practical fields. In this paper, we analyse the monthly mean temperature in Nanjing, China, from 1951 to 2017, using SARIMA (Seasonal Autoregressive Integrated Moving Average) techniques. Data from 1951 to 2014 are used as the training set, while data from 2015 to 2017 are used as the testing set. A detailed explanation of model selection and forecasting accuracy is presented. The results show that the proposed research approach obtains good forecasting accuracy.