Title: Analysis of time series variations of temperature and its forecast in the northeastern Bangladesh

Authors: Ahmad Hasan Nury; Khairul Hasan; Md. Jahir Bin Alam; Md. Mahedi Hasan

Addresses: Faculty of Civil and Environmental Engineering, Shahjalal University of Science and Technology, Sylhet, Bangladesh ' Faculty of Civil and Environmental Engineering, Shahjalal University of Science and Technology, Sylhet, Bangladesh ' Faculty of Civil and Environmental Engineering, Shahjalal University of Science and Technology, Sylhet, Bangladesh ' Faculty of Statistics, Bangladesh University of Professionals, Dhaka, Bangladesh

Abstract: Time series analysis of temperature data can be a valuable tool to investigate variability pattern as well as to predict short and long-term changes. Here, linear trends showed that the maximum temperature is increasing by 2.97°C and 0.59°C per hundred years, and the minimum, by 2.17°C and 2.73°C per hundred years at the Sylhet and Sreemangal stations. Anomaly of these regions also showed increasing temperature. Seasonal autoregressive integrated moving average (SARIMA) model was fitted for temperature time series with its traditional three steps: identification, diagnosis and forecasting respectively. For monthly maximum and minimum temperature at Sylhet and Sreemangal stations, respective SARIMA models were (3, 1, 3) (1, 1, 1)12, (2, 1, 3) (0, 1, 1)12, (3, 1, 1) (1, 1, 1)12 and (2, 1, 1) (1, 1, 1)12. This will help the policy makers to understand the nature and scale of possible temperature changes in northeastern Bangladesh.

Keywords: seasonal autoregressive integrated moving average; SARIMA; linear trend; anomaly; temperature; time series; maximum temperature; minimum temperature; variability pattern; predict short and long-term change; Bangladesh.

DOI: 10.1504/IJGW.2017.086283

International Journal of Global Warming, 2017 Vol.13 No.2, pp.157 - 182

Available online: 17 Aug 2017 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article