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

Estimation of Area Under Agroforestry Practices in Modhupur Sal Forest Using Remote Sensing

M. Golam Mahboob
T&C Wing, Bangladesh Agricultural Research Institute (BARI), Joydevpur, Gazipur

G.M. Mujibar Rahman
Department of Agroforestry, Bangladesh Agricultural University (BAU), Mymensingh

M. Jinnahtul Islam
Bangladesh Space Research and Remote Sensing Organization (SPARRSO), Agargaon, Dhaka

M. Abdus Salam
Department of Aquaculture, BAU, Mymensingh

Abstract/ Executive Summary:

The objective of this study was to detect and discriminate the area under agroforestry practices through satellite remote sensing in Modhupur sal forest. This study was conducted in the year of 2003 at Bangladesh Space Research and Remote Sensing Organization (SPARRSO) laboratory analyzing SPOT image of March 3, 1999 with bands 4, 3, 2 and 1 using ERDAS Imagine professional software. All individual bands and their different combinations of the imagery were tried for depicting the agroforestry areas. From the best combination of bands False Color Composite (FCC) was prepared. Extensive ground verification was conducted with the help of GPS and plantation journals. Unsupervised classification of SPOT image was conducted through the ERDAS Imagine and 15 landuse classes were derived. Similar landuse classes were merged together incorporating the ground truth information. Best FCC was obtained from the combination of SPOT spectral bands 3, 4 and 2 (RGB) for discrimination of the desired features in Modhupur sal forest. After merging similar landuse features incorporating ground data, a final map with four classes: sal forest, agroforest, rice field and depressed land were generated. The extents of agroforestry area were mainly found distributing around the peripheral zones of remnant sal forest. Some agroforestry practices were found scattered in all over the study area. The results of this study exhibit the way of depiction and estimation of agroforestry and other landuse features through remote sensing technique which could be refined in future with updated satellite images to study different aspects of forestry and agriculture in Bangladesh.

Keywords: Agroforestry, Remote sensing, Landuse classification, Sal forest, FCC, ERDAS Imagine, SPOT

Location: Modhupur sal forest

Start Date: 00-01-2003

End Date: 00-12-2003

Program Area: Conservation and Biodiversity

Non-commodity: Adoption of technology

Objectives:

1. To detect and discriminate the extent of existing agroforestry practices in Modhupur sal forest; 2. To measure the area under agroforestry practices within the study region of Modhupur sal forest.

Methodology:

All laboratory-based operations were undertaken at GIS and Remote Sensing laboratory of Bangladesh Space Research and Remote Sensing Organization (SPARRSO) situated at Sher-e-Bangla Nagar, Agargaon, Dhaka. The database contained several raster images extracted from SPOT image of March 3, 1999, and digitised from a wide range of paper maps and tabular data from different sources. SPOT image of the seen J 301 (path), K 235 (row) band 432 and 1 was loaded to the computer equipped with the ERDAS Imagine professional software version 8.4. The software was run in the networked GIS Workstations powered by Windows 2000 version 5.00.2195 operating system (OS). SPOT multispectral images used in this study have a maximum resolution of 20m x 20m. The same is also capable to show a higher resolution of 10m x 10m in panchromatic mode. The image covers a vast area of Tangail, Mymensingh and Sirajganj districts.

Source of Information: J. Agrofor. Environ. 1(1): 2007; ISSN 1995-6983

Article Link (Online Journal):

Funding Source:

Total Budget (Tk.) :

Achievement/Findings:

The results of this study are indicative of discriminative power of Remote Sensing which could be refined in future with most updated satellite images and can also be used for other type of land uses such as forestry as well as agriculture.

Knowledge Management: Journal