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

Monitoring Gamar (gmelina Arborea Roxb.) Plantation in the Chittagong Hill Tracts of Bangladesh Using Remote Sensing Technique-synthesis of Limitations

A. Z. Md. Zahedul Islam
Bangladesh Space Research and Remote Sensing Organization (SPARRSO), Agargaon, Sher-e-Bangla Nagar, Dhaka 1207, Bangladesh.

M. Mahmudur Rahman
Bangladesh Space Research and Remote Sensing Organization (SPARRSO), Agargaon, Sher-e-Bangla Nagar, Dhaka 1207, Bangladesh.

Mozammel Haque Sarker
Bangladesh Space Research and Remote Sensing Organization (SPARRSO), Agargaon, Sher-e-Bangla Nagar, Dhaka 1207, Bangladesh.

Abstract/ Executive Summary:

This paper has demonstrated the limitation of monitoring Garnar plantation, based on the development of a canopy growth model and estimations of plantation areas, in a hilly area using Landsat TM images. The primary difficulty arose in extracting Gamar plantation areas from the images due to the shadowing effect of the hills. As a result, estimations of the whole plantation areas for majority of the plantation years were not possible. However, estimations of plantation areas from two TM images of different years were seen mutually consistent and reasonably accurate for areas having no shadow zone. The other difficulty arose from the absence of the developing phase in the NDVI/Age curves of the Gamar canopy, which made it impossible to develop the canopy growth model correlating the spectral response of the Gamar canopy with its measured growth parameters. The exercise presented in this parer has contributed towards understanding the complex variability of the spectral response of vegetation in a hilly area.

Keywords: Gamar canopy, Plantation, Chittagong Hill Tracts, Remote Sensing Technique

Location: Chittagong

Start Date: 09-10-1988

End Date: 06-04-1990

Program Area: Knowledge Management

Non-commodity: Remote sensing

Objectives:

To develop a growth model for planted Garnar canopy using TM data and measured growth parameters.

Methodology:

The study area is located between 22°10' 26" and 22°28' 26" N latitudes and 92°06' 50" E and 92°20' 34" E longitudes in the Chittagong Hill Tracts region of Bangladesh. The area receives annual average rainfall of 2750 mm with five months less than 200 mm. The topography of the area is characterized by the small and medium high hills with elevation ranging from 100m to 400m. The soil of the area is shallow to moderately deep, pH varies from 5.4 to 6.3. Soil is well to excessively drained. The ecosystem of the area naturally thrives on tropical wet evergreen and semi-evergreen forests. It should be mentioned here that once Dipterocarps were the dominant species in this study area, but, presently they have been removed. Gamar is a major pulpwood species. Pulpwood plantation has been initiated in this region since 1974 to meet the demand of raw materials of Karnaphully Paper Mills. The plantation has been raised by clearing the degraded tropical wet evergreen and semi-evergreen forests. In the plantation, Gamar is a major species which is fast growing and suitably used for the production of pulp. Field campaign was conducted in order to collect information on the physical environment and to measure the relevant canopy parameters in the study area. The hills in the area were seen to have considerable variations in elevation, slope and orientation. The plantation areas were found to have Gamar canopy as the main vegetation and perennial bushes, evergreen bamboo, etc. as the co-existing vegetation. The perennial bushes were the main co-existing vegetation in the study areas. The measured canopy parameters were stand density, mean basal and mean height of the Gamar canopies of different age-classes. Gamar trees were planted each year between 1975 to 1998 in different areas, and the different age-classes of Gamar trees were obtained from different areas. Thus, the growth parameter/age relationship obtained in this measurement for each age-class of the canopy was specific of the individual site quality to which the canopy of the particular age belonged. The measured height/age scatter plot is not always progressive indicating comparatively poor quality of sites of some age-classes (years of 10, 11, 15, etc.). It is, therefore, very important to consider the individual site quality in the canopy growth model. In the scatter plot of three growth phases has been identified: Initial developing phase (from 1 to 5 years, average growth rate is about 1.05 m/yr), Middle developing phase (6th and 7th years, average growth rate is about 2.185 m/yr) and developed phase (from 8th year on- wards, average growth rate is about 0.43 rn/yr). Average height/age curve of the Gamar canopy is constructed using the average growth rate of each growth phase. The growth rate in the middle developing phase was very high, and having this growth the Gamar canopy became more or less compact at the end of this phase. In order to examine the relationship between the growth parameter and the spectral response at different non-deciduous stages of the Gamar canopy, TM images of 9 October 1988 and 6 April 1990 were used in this study. The later one belonged to the young growing period, and the former one belonged to the matured non- growing period of the Gamar leaves. The plantation map of the Kaptai Pulpwood Division was used to delineate the plantation boundaries of different years. The plantation map was checked in respect of scale deformation using a topographic base map of the area in the scale of 1 :50000; scale deformation was not found. The images was geometrically corrected using the base map. The plantation areas of each year were digitized as separate ARC/INFO coverage, which was transferred as digitized polygons to ERDAS images processor and was used to extract the plantation areas of the year from the Images. The Gamar trees were planted each year starting from 1974. The measured data on growth parameters of the Gamar canopy and the areas of the Gamar plantation in the plantation map were available for each plantation year between 1975 to 1989. Therefore, in the images of 1988 and 1990, it was possible to obtain Garnar Canopy of different age-classes. However, during field campaign it was observed that the plantation areas delineated in the plantation map contained Gamar canopy as well as other co-existing vegetation (mainly perennial bushes). Therefore, in order to obtain the said age-classes, Gamar canopies must be separated from its co-existing vegetation in the plantation area (of each year) extracted from the images. The nex t task in this study was to construct an DVII Age curve of the Gamar canopy in order to examine the possibility of development of the said growth model.

Source of Information: Journal of Remote Sensing and Environment, 1998, Vol. 2 ISSN 1560-201X

Article Link (Online Journal):

Funding Source:

Total Budget (Tk.) :

Achievement/Findings:

This paper has synthesized the difficulty at monitoring vegetation in a hilly area having co-existing vegetation environment using remote sensing data. One of the primary difficulty arose from the shadowing effect of the hills. In shadow zonex, the Gamar canopy was found not separable from the co-existing vegetation. As a result, estimation of plantation areas for majority of the plantation years was not possible. However, area estimations from the two images were mutually consistent and was found reasonably accurate for those areas where shadow zone of hill did not exist or existed with lower percentage. The next difficult arose from the high variability of the spectral response of the vegetation facing direct sunlight. This variability is due to the combined effect of the topographic characteristics of the terrain (slope and orientation diversity) and the co- existing vegetation environment in the study area. Due to this variability, it became difficult to select training samples for masking-in the Gamar canopy in the majority of the plantation areas facing direct sunlight. However, unsupervised clustering method was found applicable for masking-in the said Gamar canopy in the plantation areas. The final difficulty arose from the anomalous NDVI/Age curve of the Gamar canopy. The expected developing phase in the curve was not present due to the modulation of the background vegetation and the topographic characteristic of the terrain. On the other hand, spectral saturation of the Gamar canopy was attained at the beginning of the developed phase (after the age of 7 years), and there existed no meaningful NDVI/Age relationship in this phase. Development of Gamar canopy growth model from TM data was, therefore, not possible.

Knowledge Management: Journal