Agricultural Research Management Information System

Home
Research Summary
Search
About Us
- ARMIS
- Brochure
Contact Us
Report
User Request
Data Input
Help
- Operation Manual
  - PDF
  - Video
- Program Area & Commodity
We have reached 37600 number of research entries at this moment.

- Logout

Research Detail

Home
Research
Detail

Trend and Variability Analysis for Forecasting of Temperature in Bangladesh

J. A. Syeda
Department of StatisticsHajee Mohammad Danesh Science and Technology University, Dinajpur

Abstract/ Executive Summary:

An attempt was made to investigate the trend and variability pattern for decadal, annual and seasonal (three crop seasons) average dry bulb temperature (ADBT) for the six divisional stations of Bangladesh namely Dhaka, Khulna, Rajshahi, Barisal, Sylhet and Chittagong. The monthly ADBT for 2009-2012 is forecasted using the univariate Box-Jenkin’s ARIMA (autoregressive integrated moving average) modelling technique. The rates of linear trend for annual average dry bulb temperature (ADBT) were found negative for Rajshahi and Barisal but positive for Dhaka, Khulna, Sylhet and Chittagong. The rates for Seasonal ADBT were positive for all the three seasons for Dhaka and Chittagong, but negative for all the three seasons for Barisal while positive for Kharif and Rabi seasons and negative for Prekharif seasons for Khulna and Sylhet. The rates were positive for Kharif season but negative for Prekharif and Rabi seasons for Rajshahi.

Keywords: Forecasting, Temperature, Kharif and Rabi seasons

Location: The monthly average dry bulb temperature (ADBT) in degree centigrade for Dhaka, Rajshahi, Khulna,Barisal, Sylhet and Chittagong

Start Date: 00-01-1948

End Date: 00-12-2008

Program Area: Risk Management in Agriculture

Non-commodity: Adoption and mitigation technology

Objectives:

1. To investigate the trend and variability pattern for decadal, annual and seasonal (three crop seasons) average dry bulb temperature (ADBT) for the six divisional stations of Bangladesh namely Dhaka, Khulna, Rajshahi, Barisal, Sylhet and Chittagong.

Methodology:

Data were taken from Bangladesh Meteorological Department, Dhaka. The monthly ADBT in degree centigrade were taken for Dhaka, Rajshahi, Khulna, Barisal, Sylhet and Chittagong during 1953-2008, 1964-2008, 1948-2008, 1949-2008, 1956-2008 and 1949-2008, respectively. The missing data were filled in by the median of the corresponding years. The seasonal data for the three crop seasons namely Kharif, Prekharif and Rabi were derived by averaging the monthly data taken from June - October, March - May and November - February, respectively. In this study, the linear trend for annual and seasonal ADBT were fitted with the least square method taking the following form of equation - Y= a + bX where, Y - ADBT, X - time, a and b - parameters Stationarity of residuals for ADBT trend was tested using autocorrelation function (ACF) and partial autocorrelation function (PACF) display and the normality was checked by normal probability plot. The value of classical ‘t’ test was used for the identification of significant ADBT trend, when residuals follow the normality and stationarity pattern. Univariate Box-Jenkin’s ARIMA model was fitted to forecast the monthly ADBT data for January 2009 - December 2012. After confirming that the series was stationary, an effort was made for an ARIMA model to express each observation as a linear function of the previous value of the series (autoregressive parameter), and of the past error effect (moving average parameter). The available data were divided into training, validation and test sets. The training set was used to build up the model, the validation set was used for parameter optimization and the test set was used to evaluate the model. The adequacy of the above model was checked by comparing the observed data with the forecasted results. The data for the last ten years were used to compare with the fitted model forecasts for the years and the models were selected for the minimum root mean square forecasting error of the data set of mentioned time range. The diagnostic techniques namely histogram of residuals, normal probability plot of residuals, ACF and PACF display of residuals, times series (TS) plots for residual versus fitted values and TS plots for residual versus order of the data were used for checking residuals of ARIMA models. Box-Cox transformation was used for variance stabilization and the transformation of the data to get stationary series from nonstationary series. The software package ‘Minitab 13’ was used to fit the ARIMA univariate models. A detailed description of the nonseasonal and seasonal ARIMA models and the standardized notation used.

Source of Information: J. Environ. Sci. & Natural Resources, 5(1): 243-252, 2012; ISSN 1999-7361

Article Link (Online Journal):

Funding Source:

Total Budget (Tk.) :

Achievement/Findings:

The rates for Seasonal ADBT were positive for all the three seasons for Dhaka and Chittagong, but negative for all the three seasons for Barisal while positive for Kharif and Rabi seasons and negative for Prekharif seasons for Khulna and Sylhet. The rates were positive for Kharif season but negative for Prekharif and Rabi seasons for Rajshahi.

Knowledge Management: Journal