|
|
|
1 KM Land Cover Data Base in Asia
|
GIS of management of forest in eastern Sri Lanka through a semi-knowledge based approach
L. K. Perera and Ryutaro Tateishi
Weathernews Inc., D21, 1-3, Nakase, Mihama-ku, Chiba-Shi 261-01, Japan Fax: +81-43-274-5506
Center for Environmental Remote Sensing (CEReS), Chiba University, 1-33 yayoi-cho, Inage-ku,
Chiba-Shi 263, Japan. Fax: +81-43-290-3857
Abstract
Geographical information System (GIS) gives a strong approach to solve environmental problems by linking remotely sensed data with geographical data. The conventional approach is ranking these data files and merging respective classes to select most suitable to least suitable areas for some specific need. But, when the objects of the environment management become more intensive, it is important to consider the practical capability to conduct some new activity on land. By this paper, we introduce a "general value unit" approach which is based on probability calculation to merge different GIS files in order to produce a more practical land suitability map for reforestation in Sri Lanka.
Sri Lanka has a very limited forest coverage (37%) of th total land area), and this area is also decreasing at on alarming rate. Eastern Sri Lanka has many poddy lands feed with mahaveli river water. These agriculture based human activities have rendered some adverse effects to the available limited forest cover (including Wasgomuwa strict natural forest reservation) and the wild life of the are. In this study, buffer zones with practical high capability to conduct reforestation were selected and geographical data of those buffer areas were ranked through a semi-knowledge based method. The resulted land suitability map for reforestation will bring an environmental guide line for the ongoing and further environmental plans of Sri Lanka.
1.Introduction:
Decisions to change the use of land will cause great benefits or losses to many parties, not only to the mankind, but also to all living things. Land may be evaluated directly, by trial, that is by growing a crop or building a length of pipeline, to see what happens. But these kind of evaluations are based on crop or building a length of pipeline, to see what happens. But these kin of evaluations are based on inadequate information and may not be closely related to the inherent productivity of the land. Therefore most land evaluation systems are based on indirect methods which assume certain soil and site properties influence to the success of the particular land use (McRae & Burnham, 1981). Avery (1962) and Buter (1964) pointed out that there can be no general relationship between any one soil or site attribute and crop performance since very many properties influence yield and they do not act independently. Today, remote sensing and GIS (Geographical Information Systems) are playing a very significant roll in land evaluation systems such as production of land suitability maps (Perera & Tateishi, 1995).
The first sections of this paper contains a brief explanation about the methodology which was used to produce a land suitability map based on satellite data and GIS. Then some results extracted from a bigger study also included. The study area was selected from eastern Sri Lanka and covers over 9300 sq. km. of land (see figure 01). The methodology re-classifies the land with digital data processing methods, while using human knowledge and experience to make sub-divisions of each GIS file, which we could name as a 'semi-knowledge based approach' or as a 'semi-automatic approach' in other words.
2.The new approach
According to the new approach we considered that; 'all physical factors on or effect to the land are equally important'. It is important to do least changes in economically or environmentally valuable land suitability assessment. Hence, present land cover becomes the most important data file of the entire analysis. All other facts effecting to the land or related to the land can be considered with equal impotence when those elements are treated individually. This idea gives a clue to construct a new foundation to integrate GIS data file using a 'general value unit' . Also under the practical conditions, there are some land cover types those are not practical to consider in a land suitability assessment . With this background, the entire research approach can be summarized into four main divisions;
Figure. 01. Location of the study area
|
|
|
|
|
|
|
