The Affection of the Vegetation Change on the Eroded Soil in South-East Asia
Takeharu Kawai, Shunji Murai
Institute of Industrial Science The University of Tokyo
7-22-1, Roppongi, Minato-ku, Tokyo 106 Japan
Shiro Ochi
United Nations Centre for Regional Development (UNCRD)
1-47-1, Nagono, Nakamura-ku, Nagoya 450 Japan
Abstract
Worldwide decrease of forest causes of disaster such as flooding due to the reduction of water retention, desertification, and soil erosion etc. Especially in South-East Asia, soil erosion due to deforestation has become a serious problem in this decade.
In this paper, the relationship between water runoff, vegetation, and soil erosion is studies. The quantity of eroded soil at the test area is estimated from the Digital Elevation Model (DEM) and present vegetation which is calculated from the SPOT imagery. The final objective is to examine the affection of the vegetation change on the water runoff and the change of the quantity or eroded soil.
Introduction
Soil erosion, which influences not only the agriculture but also on the environment, has become a very serious problem at many places in the world. It is important to assess the impact on soil due to the vegetation change, which means the large-scale deforestation in many cases so far as South-East Asia in concerned, because it is one of the reasons why erosion happens.
The objective of this study are :
- to propose a new soil erosion model with use of Remote Sensing data, and
- to examine the relationship between the vegetation change and the extent of soil erosion using the proposed model.
Review of the Existing Soil Erosion Models
The existing models on soil erosion are divided into two categories from the point of the scale.
- the “Micro-scale” type
- the “Macro-scale” type
Type a) is used for relatively small areas(1). It is based on the mechanism and the theory which cause erosion in order to precisely reflect its procedure, therefore, a lot of experiments are done in order to estimate parameters of the equations. However, even if the parameters are defined, it is still difficult to apply to larger area.
Concerning type b) which is used for relatively large areas, the elements which causes erosion is not attributed to the individual processes but to the integrated processes, for example the inclination and length of the slope, the condition of the vegetation on the ground, and so on. Among them, the Universal Soil Loss Equation (USLE) is the most well-known model and has been modified to adopt to many actual cases.
Proposal of the Soil and Water Flow Model
3.1 Overview of the model
The objectives of the proposed model are as follows :
- to utilize Remote Sensing data effectively, and
- to present the relationship between vegetation change and soil erosion
Raster format is adopted in order to satisfy the first objective. Test area is also divided into grids to match them. It is assumed that soil and water move to the grid which is the lowest of the eight neighboring ones per a unit time(3). This method enables to simulate the movement easily, but it is difficult to perfectly realize the increase of the velocity corresponding to the increase of the flow quantity, as shown by Manning’s Formula and so on. The proposed model will be divided into two submodels which are for soil and for water.
Vegetation influences soil erosion mainly in two ways. One is water retention effect at the roots of plants, which influences rill and gully erosion it is taken into consideration in the model to meet the second objective. The other is natural canopy effect that leaves and branches interfere with rain’s strike on the ground, which influences splash and sheet erosion. However, the amount of eroded soil because of splash and sheet erosion is usually relatively small. Therefore a substitutive method may be applicable, for example to incorporate a certain portion of rainfall into inflow volume previously and regard the rest as “appropriate” rainfall.
3.2 The Submodel for Water Flow
The outline of the submodel for water is shown in Figure 1. This figure describes the movement of water and soil at one grid and one step. “Standard” total outflow means the outflow volume presumed for a certain “standard” inclination when the flowing destination after this step has not been determined yet. The bold solid arrows mean the flow of water.
As described above, the flowing direction will be selected from eight neighboring grids. If the selected grid which is of course the lowest of eight ones, is still higher than the central grid, it is assumed that water is accumulated on the surface of the central grid until its volume excesses the product of the grid area and height difference.
3.3 The Submodel for Soil Flow
The outline of the submodel for soil flow is also shown in Figure 1. The flow of soil is described as the bold broken arrows. A remarkable thing is that the quantity of soil outflow is mainly determined by the water outflow. Therefore, the submodel is dependent upon the water submodel.
Figure 1 The Outline of the Proposed Model
3.4 The Whole Model
The whole model is constructed by means of the two submodels adapted to each grid and for computer simulated steps successively.
The mark of ? on figure 1 means the need of empiric equations. It will be useful for their assessment to combine experimental data, existing statistic data, digital elevation model (DEM), and Remote Sensing data.
Application
The proposed model is examined at the test area where a large slope failure happened in the Southern Thailand in 1988. Three alternatives related to vegetation are prepared for comparison.
- when existing forest remains all over the area.
- when existing forest remains at the regions which are higher than a certain height (900m in this case) but the others are deforested.
- When all of the area are deforested.
