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

Variability and Trend of Decadal and Annual Climatic Variables in Dinajpur District

Jahanara Afrose Syeda
Department of Statistics, Hajee Mohammad Danesh Science and Technology University, Dinajpur, Bangladesh

Abstract/ Executive Summary:

An attempt is made to investigate the decadal and annual trend and variability for 20 climatic variables of Dinajpur district for 1948-2004. The variety of exploratory data analyses (EDA) tools and different robust and nonrobust measures are used for the analyses. The significant positive rates are found for annual average relative humidity in the evening (0.175%/yr) and soil temperature (0.0730C/yr) at the depth of 5cm but significant negative for average range temperature (0.0290C/yr), maximum temperature (0.0200C/yr), difference of relative humidity at morning and evening (-0.044%/yr) and total frequency of insignificant rainfall (0.261days/yr). The fairly high positive rates are observed for annual total rainfall, maximum rainfall, average relative humidity, relative humidity in the morning, cloud, minimum temperature, wet bulb temperature and maximum wind-speed but negative for total frequency of zero rainfall, average evaporation, difference of dry and wet bulb temperature and sunshine-hour with the nonnormal and/or nonstationary residual. The very low positive rate is found for ASLP (+0.0007mb/yr) with normal and stationary residual but negative for ADBT (-0.00040C/yr) with nonstationary residual. The historical climatic data needs exploratory analysis and warrants tougher justification in classical analyses for outlier and residual’s nonnormality and nonstationarity. The findings support that the climate of Dinajpur is changing in different terms which may affect its agricultural production

Keywords: Variability, Decadal, Annual, Climatic variables, Dinajpur

Location: Dinajpur District

Start Date: 00-01-1948

End Date: 00-12-2004

Program Area: Risk Management in Agriculture

Non-commodity: Adoption and mitigation technology

Objectives:

1. To investigate the decadal and annual trend and variability for 20 climatic variables of Dinajpur district for 1948-2004.

Methodology:

The daily and monthly data for 1948-1972 and 1981-2004 on climatic factors of Dinajpur are collected from Bangladesh Meteorological Deptt., Agargaon, Dhaka, Bangladesh. The collected data are the total rainfall in mm (TR) and maximum rainfall (MXR) in mm, total frequency of insignificant rainfall in days (TFIR), average dry bulb temperature in 0C (ADBT), average maximum temperature in 0C (AMXT), average minimum temperature in 0C (AMNT), average range temperature in 0C (ARNT), average wet bulb temperature in 0C (AWBT), average difference of dry bulb and wet bulb temperature in 0C {AT(D-W)}, average relative humidity in percentage (ARH), average difference of relative humidity between morning and evening in percentage (ARH(0-12)), average wind speed in knots (AWS), average maximum wind speed in knots (AMWS), average sea level pressure in mb (ASLP), average cloud in octas (AC). Besides, the data of daily and monthly evaporation in percentage (AE) for 1987-2000, average soil temperature (AST) in 0C at the depths of 5, 10, 20 and 50cm, respectively for 1987–2000 and the average sunshine-hour (ASH) for 1989-2004 are collected. In this study, the daily and the monthly missing data for 1973-1976 are filled in by the medians of the observed data for 1948- 1972 and the missing data for 1977-1980 are filled in by the medians of the observed data for 1981-2004. Daily missing values are placed by the median of the corresponding daily data of the months/years. Analytical methods The within-year and between-year decadal and annual patterns are investigated for the data for 1948-2004 with the exploratory data analyses techniques like boxplot, stem-and-leaf plot, and median polish table. The statistical package Minitab 11.12 is used for the analysis. The nonrobust measures like mean, coefficient of variation and robust measures like median, 5% trimmed mean and the percentage ratio of quartile deviation to median are used to investigate the within-year variability pattern of data. The trend of the number of weeks containing 0–5 millimeter (ml) rainfall are investigated as it is considered insignificant from agricultural view and this analysis can focus about the drought. To test whether the trend is deterministic or stochastic, the stationarity of residuals are checked after trend fitting using the sample autocorrelation function (ACF) and from the partial autocorrelation function (PACF) and Box-Peirce test statistic. The normality of the residuals are examined from these fits by normal probability plot and the rescaled moments (RM) test for normality.

Source of Information: J. Agrofor. Environ. 5 (2): 27-32, 2011; ISSN 1995-6983

Article Link (Online Journal):

Funding Source:

Total Budget (Tk.) :

Achievement/Findings:

Some residuals demonstrate stationarity pattern but several follow non stationarity and a few error support normal distribution but others show abnormality. Sometimes robust and non-robust measures come up with conflicting conclusions. So it is imperative for us to avoid normality assumption based statistical procedures for further analysis of historical climatic data and emphasize the view in favor of EDA and robust techniques in analyzing climatic variables. Significant changes in some variables support that the climate of Dinajpur district is changing, which ultimately may affect its agricultural production.

Knowledge Management: Journal