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

Map of Nearly Assured Monthly Maximum and Minimum Temperature in Bangladesh 2003-2004

Manotosh Howlader, MSc. (Stat.)
Senior Scientific Officer
Agricultural Statistics Division, BRRI-Gazipur

Md. Shahjahan Kabir
SSO
Agricultural Statistics Division, BRRI-Gazipur

Md. Rokonuzzaman
Scientific Officer
Agricultural Statistics Division, BRRI-Gazipur

Abstract/ Executive Summary:

The objective of this study is to determine best fitted probability distribution for monthly maximum and minimum temperature. Normal distribution and Pearson Type I distribution were examined as the probable candidates to identify the best fitted model for graduating the probability distribution of monthly average maximum and minimum temperature for the months from May to October. Data from 15 weather stations of Bangladesh Meteorological Department (BMD) where minimum of 30 years’ daily maximum and minimum temperature were available. Goodness of fit was judged by chi square test. Probability distribution of monthly maximum and monthly minimum temperatures can be graduated by Pearson Type I distribution in Bangladesh. The maps created for the nearly assured (at 90% probability) monthly maximum and minimum temperature constructed for the months from May to October based on Pearson Type I probability distribution can be used to estimate the monthly maximum and minimum temperature at any location in Bangladesh.

Keywords: Map, Max Temperature, Min Temperature

Location: BRRI HQ, Gazipur.

Start Date:

End Date:

Program Area: Risk Management in Agriculture

Non-commodity: Temperature

Objectives:

To determine best fitted probability distribution for monthly maximum and minimum temperature and
To construct maps of nearly assured monthly maximum and minimum temperatures in Bangladesh.

Methodology:

Normal distribution and Pearson Type I distribution were examined as the probable candidates to identify the best fitted model for graduating the probability distribution of monthly average maximum and minimum temperature for the months from May to October. Data from 15 weather stations of Bangladesh Meteorological Department (BMD) where minimum of 30 years’ daily maximum and minimum temperature were available. Goodness of fit was judged by chi square test. After choosing the appropriate distribution, monthly average temperatures with 90% probability were computed by solving the equation for x, where are the estimates of the parameters. These estimate of temperature thus obtained was treated as nearly assured monthly average temperature. These estimates at 30 weather stations of BMD were used to construct the surfaces of nearly assured monthly minimum and maximum temperature for the months from May to October in Bangladesh. ARCView 2.1 spatial analyst was used to create the surface following Inverse Distance Weight (IDW) interpolation module.

Source of Information: BRRI Annual Report 2007-08

Article Link (Online Journal):

Funding Source:

1. Government

Budget:

Total Budget (Tk.) :

Achievement/Findings:

This is summarize the results for goodness of fit of candidate distributions for maximum temperature and Tables 8-9 for minimum temperature. It was found that, in most cases the value of l²was smaller for Pearson Type I distribution than it was for Normal distribution. Out of 90 (15 stations x 6 months) cases, there were 25 cases where neither of these two distributions fitted well. However, Pearson Type I distribution showed lower value of l² in 14 cases out of these 25 cases. Of the remaining 65 (90-25) cases, Pearson Type I distribution fitted well in 60 cases and Normal distribution in 30 cases. In each of the months from May to October Pearson Type I distribution showed better fit than Normal distribution as the number of stations having insignificant value of l² is higher for Pearson Type I distribution than for Normal distribution. Both in May and October there was only one station where Pearson Type I distribution did not fit well. In each of the months from July to September, the number of stations showing insignificant value of l² was much higher for Pearson Type I distribution than for Normal distribution.

Knowledge Management: Report/Proceedings