Mapping Tropical Forest Cover Types using Optical Satellite Data And GIS In Gentral in Central Sumatra, Indonesia
Yousif Ali Hussin John S. Mwantdha Hamid Marwat
International Institute for Aerospace Survey
and Earth Science (ITC)
7500 AA Enschede, The Netherlands
Fax : (31) (53) 4874-399
E-mail: HUSSIN@ITC.NL
Abstract
Never before in the history of mankind, tree and forests have assumed that global importance which is evident today. For many years the problems of deforestation in the tropical forest had raised considerable international interest. From 1850-80 about 15% of the earth's forest cover disappeared as a result of human activities (World Resources Institute 87), according to this report the tropical rain forest of Asia, Africa, and South America declined by 27 percent. This depletion is very significant, where about 2.5 million people depend n natural forest resources for many economic and environmental goods and services. According to the Food and Agriculture Organization of the United Nation report of 88, between 80 and 85 the estimated annual rate of tropical forest degradation was 0.6 percent or 11.4 million hectares. the major causes of this degradation in the tropics are causes is the population growth. In general, high population pressure has its impacts on the changes, especially in tropical countries in heavily populated areas, it is essential to monitor information is essential. The demand for timely and accurate geographic information has forest conditions. Geographic information can be derived easily from the remotely sensed Landsat Thematic Mapper ™, SPOT Panchromatic and MSS, AVHRR NOAA, radar etc. The main objective of this research was to map tropical forest cover types using optical satellite data i.e., Landsat TM and Spot XS satellite images in an Indonesian test site of Jambi Provice, Central Sumatta.
1. Introduction
FAO estimates that the forest area in the world was about 3.4 billion hectares in 1992. This area is known to be decreasing at a rate of 2.1% per year calculating from the 1962 figures. The reduction in area is however not the same everywhere. Developing countries in Latin America, Asia and Africa have higher deforestation rate than 2.1% while in the European Union, an increase in area up to 20% has been recorded between 1961 and 1992.
The fact that the forest area is decreasing worldwide is an indicator of its importance. Unfortunately most of the benefits of a forest are of destructive nature to its. Some of these benefits are the provision of wood (timber and poles) and wood products, food, fodder, medicine and fuel. Others benefits include employment, recreation, carbon dioxide sinks, wildlife habitats, genetic reservoirs and maintaining biodiversity (Buschbacher, 1986; de Gier, 1995; Hamilton, 1984).
A lot of effort, is being exerted to try and conserve the remaining forests and even create others. This has been a result of concern about the disastrous effects of deforestation. Buschbacher, (1986), summarizes these as permanent degradation of forest, loss of cultural and biotic diversity and especially in the developing nation, should of course not be forgotten.
1.1 The Indonesian Situation
In Indonesia about 75% (143 million hectares) of land surface has the status of forest land. The National Land Use Law of 1992 has designated 34% of this land as Protected forest, 45% as Production forest and 21% as Conversion forest (Hoekman, 1995). The planning, management and surveillance of such vast areas requires up-to-date information on land cover and land cover change. An accurate assessment of forest and land cover change is a prerequisite, not only for studies on climate and global environmental change, but more importantly for formulation and identification of ways to slow down, or even, half the rate of deforestation.
Several methods of acquisition, processing and analysis of various of data to monitor cover change are currently in use in Indonesia. Most methods involve use of aerial photographs and Landsat and SPOT satellite data. These have a great limitation with respect to cloud cover and haze in as far as provision of information on a regular basis is concerned. Central Sumatra, where Jambi is located, is one of those areas most affected by severe cloud cover. A study is presently under way to evaluate the usefulness of radar data (both aircraft and satellite) for forest monitoring and management especially to take advantage of its cloud and haze penetration properties. Whatever land cover classification system using these data is necessary. All data acquisition and handling techniques have some limitations. These limitations are not the same for all data. In the case of Landsat and SPOT XS (multispectral) data integration may enable the exploitation of the different information content and advantages forest cover changes. It may also improve the certainly with which predictions of the effects of different management options can be made to slow down forest degradation and deforestation.
To accurately detect and monitor forest cover types and changes, proper data collection and processing is a prerequisite. With the current advances in satellite data collection and image processing, there is a growing concern as to which data to use for what purpose and after what kind of processing. Furthermore, if a given data set is used, with what confidence can one say, that conclusions drawn from such an analysis are correct? This is particularly important for forest management in the tropics, such as in Indonesia, where the majority of the world's species are found.
