|
|
|
Forestry / Vegetation
|
Mapping Tropical Forest Cover Types using Optical Satellite Data And GIS In Gentral in Central Sumatra, Indonesia
Several data problem have been identified in Indonesia. Apart from problems related to satellite data acquisition such as cloud cover, the reliability of such data for classification and therefore monitoring of forest land has not been investigated. At the same time no image processing guidelines are in place to help in dealing with the huge amount of digital data available. Apart from the foregoing, the following issues call for further investigation in the use of optical satellite data in the classification of tropical rain forests. A lot of satellite data are being produced for almost all parts of the world for various uses. Some of these data are currently being used in Indonesia in the field of forestry. There is still a need to know which data is appropriate for forest cover type classification and what limitations, if any, there are, in the use such data.
As in other parts of Indonesia, there is a general trend towards an increase in human population in the Jambi area due to both to birth (in the area) and transmigration from more densely populated areas. The increase in population leads to pressures on the land for settlement, agriculture and recreation, among other. Most of this pressure is being exerted on forest land, resulting in the illegal encroachment for establishment of cropland, provision of timber and other wood products decisions to be made about how they should be managed, relevant and accurate data about the forest and its surrounding must be available, only then can the needs of the population be met and conservation for posterity be possible.
Several logging concession have been signed between the Ministry of Forest and various logging companies (Pantimena, 1996). Knowledge of how much is cut, the rate of regeneration, what forest classes are available and other related information are important for management planning. Whereas most of this information can be obtained from the normal state-of-the-art forest survey methods and that have greater temporal resolution and expensive and often unreliable after some time. Data collection methods more desirable. Though satellite data promise to meet this requirement, a problem arises as to how forest do not only depend on the activities themselves but are also known to change with time and with changes in environmental conditions. Constant monitoring of both the activities and forest cover changes is therefore a necessity. This could then help in determining the cause and effect relationships between human activities and forest degradation and designing of appropriate measure to handle them.
Tropical rain forests vary considerably (in species composition, size of stems, basal area, crown cover) form place to place, even within the same forest type. Furthermore, transition from one type to another does not often have a clear-cut boundary. These variations make classification complicated. In this regard, an assessment of classification accuracy should take into consideration the effect of variability. This would be inline with Lowell (1994) who suggested the creation of forest maps in which not all boundaries are definite and fixed, but where some are just transition zones. An investigation of these alternatives for use in classification of forest types, especially while using remotely sensed data, needs to be done.
The relationship of man and natural resources is complex, dynamic and intricate. Consequently, not all problems of a resource can be addressed in single research. The objectives of this study were: to compare and contrast the effect of spectral resolution difference of TM and SPOT XS data on the detection and classification of forest cover types; to assess the suitability of selected image transformations (vegetation index, normalized difference vegetation index, tasseled cap transformation and principal component analysis), of Landsat TM and SPOT data to improve forest cover type classification; to investigation the potential of using the fuzzy set approach to determine the accuracy of forest type classification.
2. Study area
2.1 Location and Climate
The study area lies between lines of latitude 1° 15' and 1°45' south and lines of longitude 102° 45' east. It is located in Bungotebo country, in the northwestern part of Jambi, Sumatra island, Indonesia. The Batang Hari river is a dominant feature in this area (Pantimena, 196). Other rivers in order of size are Batang Tabir, Batang Tebo, Sungai Sumai and Sungai Ajai.Muratebo found at the point where river Batang Tebo joins river Batang Hari is the main town in the study area. The area is relatively flat ranging between 5 and 30 m above sea level in most parts.
The climate is hot and et most of the year, classified as group 1a. The mean annual rainfall received in the area is about 2500 mm, most of which is between is between November and May. Temperature are highest in the period of July to September (33° to 37° C). Minimum temperature are recorded at 20° to 22° C, though they sometimes drop to as low as 18° C during the months of December and January.
2.2 Forest Resource
Several methods of classification of forest land exist in Indonesia. One of these, the Consensus on Forest Land Use classification, locally known as Tata Guna Hutan Kesepakatan (TGHK), is mainly based on susceptibility to soil erosion. Using the system, forest classification in Indonesia falls in 5 broad categories according to their main function. These are : nature reserve, protection forest, other uses such as agricultural settlement (transmigration) (Weir and Djajono, 1994).
The land cover in the study area, like in most of Indonesia, is dominated by forest. The major forest types found in the study area as classified by the TGHK system are production forest, conversion forest, protection forest and other forest. Other forest land includes unproductive dry land, agriculture fields and estates.
3. Material and Methods
3.1 Data Used
In this study, the major data sources were maps and satellite image data. The maps were used in this study : Land use map of Kabupaten Bungotebo (scale 1:200 000), Forest vegetation and land use map (land cover map) of Muara (scale 1:250 000), TGHK map of Kabupaten Bungo Tebo (scale 1:250 000), Land system (land suitability map) of Bungo Tebo (scale 1:200 000), Transmigration and agriculture development area map of Kabupaten Bungo Tebo (scale 1:200 000), Agroclimatic zone map of Kabupaten Bungo Tebo (scale 1:200 000), Soil type map of Kabupaten Bungo Tebo (scale 1:200 000) Topographic maps of Lubuk Punggai, Muara Ketato, Muara Killis, Muara Tebo of scales 1:50 000 and of Muara Bungo of scale 1:250 000. The satellite image were used: Landsat TM of 15th September 1993 and Spot XS of 5th September 1994.
|
|
|
|
|
|
|
