KR20180000619A - Soil loss evaluation method based GIS - Google Patents

Abstract

Disclosed is a method to evaluate a soil loss based on a geographic information system (GIS). According to the present invention, the method comprises: a step of using an image classification method (forest-normalized difference vegetation index (NDVI), water area-near infrared band, rice paddy and field-wetting index, downtown (city)-heat infrared band) by each land cover item, such as a downtown, a farmland (rice paddy, field), a forest, a road, a water area, etc., from the latest Landsat image to construct a land cover map used for soil loss analysis; a step of constructing GIS data, such as a land map, a digital elevation model (DEM), etc., to calculate a revised universal soil loss equation (RUSLE) factor required for an RUSLE model; a step of calculating the amount of soil loss (A); and a step of simulating the amount of soil loss of each small area to check the small area weak at the soil loss. Accordingly, the method is used for establishing a soil loss reduction means by using analysis data of the small area weak at the soil loss in addition to the evaluated amount of soil loss of each small area.

Description

{Technical Field} The present invention relates to a method for evaluating a soil loss based on GIS,

The present invention relates to a GIS-based soil loss assessment method, and more particularly, to a GIS-based soil loss evaluation method for estimating a GIS-based soil loss from a landat image by a land cover item such as a downtown area, agricultural land (rice field, field), forest area, road, Image classification [Land-NDVI, water-near-infrared band, agricultural area (rice fields) - wet index, urban area (urban area) - thermal infrared band] is used to construct a land cover map used for soil loss analysis. The GIS data such as DEM (Digital Elevation Model) and the land cover map were constructed to calculate the RUSLE factor needed for the RISLE (Revised Universal Soil Loss Equation) model. The soil loss was calculated, We developed an evaluation method of soil loss assessment to examine subwatershed vulnerable to loss, and using the analysis of subwatershed vulnerable to soil loss, And to a GIS-based soil loss assessment method for establishing the GIS.

1.1 Overview of earth loss

Soil loss is a series of natural processes in which soil constituting the crust by wind or rainfall is relaxed, separated and transported from the surface. This process of soil loss can be summarized in three major themes: energy, resistance and protection (Hudson, 1977). Here, energy implies the potential for rainfall runoff, which affects soil loss. The soil is resistant to erosion in various forms based on the physical and chemical properties of the soil itself. The resistance of soil to separation and transport from soil particles is an important parameter of soil loss. Variables that promote water penetration reduce runoff and, as a result, reduce soil loss. Various anthropogenic activities causing soil degradation increase the penetration of water, but on the other hand there is a tendency to crush the surface of the earth and to suppress the penetration and enlarge the loss. Soil forms respond to a variety of anthropogenic influences. The collapse of clay often plays a role in reducing loss, while in the case of sandy soil, it plays a role in increasing loss. The protection of the soil depends on the presence of vegetation cover. Vegetation is an effective function to directly block rainfall and reduce the rate of runoff associated with runoff (Dissmeyer and Foster, 1981). The effect of rainfall blocking on vegetation varies depending on vegetation density, and vegetation structure is also an important factor affecting soil loss.

To be more specific, the sediment loss consists of two distinct aspects: the separation of soil particles from soil particles and the transfer of particles through the medium of flowing water (Hudson, 1977).

The separation of the soil particles begins with the source of the physical erosion process when raindrops fall on the surface as shown in Figure 1 (Lal, 1977; Ritter, 1986; Marsh, 1987).

Figure 1 shows that as the pores of the soil increase under the occurrence of rainfall, the water flows downstream as a result of gravity and kinetic energy of the rainwater. Raindrops are the most important mediator of soil separation (Morgan, 1981). In general, millions of raindrops fall within 1 ㎢ during a single rainfall event, and long-term exposure of the soil to continuous rainfall, along with large amounts of rainfall, causes soil softening. In particular, tropical soils are affected not only by biochemical processes but also by meteorological factors such as wet season and dry season (Lal, 1977).

Rainfall has the power to destroy soil surfaces, but it also promotes soil consolidation (Morgan, 1995). In other words, rainfall acts simultaneously with both cementation and soil decay mediators. Soil consolidation by rainfall is more pronounced in the upper soil layer, which leads to the resolidification of soil masses and subsequent consolidation of soil after drying (Morgan, 1995). The degradation of surface soils is indicated as a major cause of inhibiting soil penetration which can promote many surface currents (Morgan, 1995).

Raw loam and sandy loam is a form of soil that is severely damaged by soil degradation, whereas binding properties of soil organic matter and clay have resistance from soil degradation (Hudson, 1977). In general, coarse-grained soils have much resistance to soil separation as a result of their weight. Clay and organic matter are composed of particles with charge, which is the source of chemical forces of adhesion, and these forces form a protective barrier against the forces of raindrops (Morgan, 1981, 1995).

The slope is an important parameter for assessing the effect of rainfall on the collapse (Wischmeier and Smith, 1978; Morgan, 1981, 1995; Desmet and Govers, 1996). With increasing tilt angle, more soil particles will migrate downstream as a result of gravity and this will result in soil particles moving in the downstream direction as shown in Fig. The erosion by raindrops occurs irregularly on the surface, and under these conditions, the areas protected by the leaves, roots, and rocks are much different from the unprotected areas in terms of soil erosion. Erosion by raindrop frying is the source of decisive forces that move the surrounding soil in the form of convex terrain, such as the watershed boundary (Morgan, 1995).

Fig. 2 is a view showing the "erosion process: movement of water and soil particles ".

The surface water flow moves the rock or soil loosely, as shown in Figure 2, in the downstream direction (Wischmeier and Smith, 1978; Marsh, 1987). As a result of fine topography, peak flow rates of surface water often result in higher dynamic energies that can transfer large quantities of material at high flow rates. The surface flow is divided into two distinct patterns, interrill erosion and rill erosion, as shown in Figure 2 (Ritter, 1986; Morgan, 1995). The form of the advanced stage of rill erosion is gully, which is a noticeable form in erosion-prone terrains as shown in Figure 3.2 (Ritter, 1986; Marsh, 1987). The surface sewage can be expressed as a result of sapping as shown in Fig. Where sapping occurs, surface soils become loose enough to cause even greater damage to surface flow (Marsh, 1987).

Whether or not vegetation has stabilized in nature is an important factor affecting the rate at which erosion can change landforms (Chonghuan and Lixian, 1992). Land cover plays an important role as an intermediary medium for restraining the movement of the soil which exists under agricultural or forest water culverts. As a result, the absence of vegetation under certain conditions will result in the rapid transfer of soil particles, and thus the role of vegetation can be discussed in terms of physical and biological background. From a physical point of view, vegetation can act as a cushioning buffer for kinetic energy due to rainfall through interception (Wischmeier and Smith, 1978; Dissmeyer and Foster, 1981). Therefore, rainfall interruption is expressed as the intermediate mediator role of rainfall caused by the vegetation before the contact between the soil and the raindrops, thereby reducing rainfall energy significantly, thereby reducing soil loss (Ondro et al., 1995).

Rainfall interception appears in a variety of forms depending on the results of animal species, water density and degree of anthropogenic activity. Leaf Area Index (LAI) is an important indicator for comparing the barrier capacity of different vegetation, where LAI is the area occupied by leaves on the total surface area covered by vegetation (Marsh, 1987) . In addition to land cover, the density of water pipes is also known to have a very large impact on rainfall interception (Wischmeier and Smith, 1978).

Vegetation coverings make a lot of difference in stabilizing the soil depending on the species. Studies have shown that grass-like roots and grass-like plants play a major role in soil stabilization compared to leafy plants (Wischmeier and Smith, 1978; Dissmeyer and Foster, 1981; Morgan, 1981). In addition, fine roots function to bind soil particles to act as inhibitors against erosion such as raindrops and surface water flow. From a morphological point of view, these roots are densely packed, curled like a line, and often serve as a continuous connection of the middle soils (Ondro et al., 1995). Conversely, vegetation, such as tropical soybeans and sugar cane, has very low roots and results in soil vulnerable to soil erosion (Ondro et al., 1995).

Roots change the soil texture and have a direct impact on penetration rate (Marsh, 1987). The rate of soil penetration is governed by a series of dynamic parameters such as the shape or density of vegetation cover that varies over time based on land cover, climatic conditions and soil texture (Wischmeier and Smith, 1978). Vegetation cover is usually affected by anthropogenic activities such as agriculture, livestock, and forests. Root tissues also play a role in generating a lot of soil pores so that rainfall and surface runoff can naturally penetrate the soil (Morgan, 1981). In addition, vegetation provides a source of life for the animals that live in oysters, and such animals can easily penetrate the soil through the oysters and cause massive soil loss and landslides (Hudson, 1977).

The presence of forest water pipes not only increases runoff during the dry season, but also reduces peak flow during many storm events. Rainforests have a significantly larger leaf area index (LAI) that blocks large amounts of rain. These vegetation absorb water in the form of fog and evapotranspiration, and the evapotranspiration falls again to the surface as erosion rainfall. Healthy rainforests not only protect the soil from erosive rainfall during the wet season, but also minimize damage during the dry season (Lal, 1977).

However, these natural protection walls were destroyed as the forests were damaged by agricultural activities. In many tropical regions where agriculture is a livelihood, water-based agriculture is dominant. During the rainy months, corn and soybean cultivated lands are covered with forests and weeds, and farmers start farming after picking up weed-like surface plants. These activities result in the substantial removal of all soil protection coatings that protect the soil during the rainy season. In addition, there have been many studies that artificial behaviors such as plowing and planting have also changed the surface of the soil, and that various forms of agriculture can play a role in increasing or decreasing the erosion tendency to soil (Wischmeier and Smith, 1978; Dissmeyer And Foster, 1981).

1.2 Type and selection of soil loss model

1.2.1 Overview

In GIS models, GIS is important because it can store and manage vast amounts of data needed to describe and predict erosion outcomes. Although there are many tools to store and manage data, GIS is distinguished from others in terms of storing and managing spatio-temporal data (Burrough et al., 1988). In the soil loss model, the use of GIS enabled the study of spatio-temporal patterns related to the potential for soil loss in the watershed (Hickey et al., 1994). Appropriate soil conservation measures can be established by looking at the response patterns of land use changes (Laflen et al., 1991).

Modeling using GIS has enabled the overlapping of complex spatial inferred data in soil loss studies (Burrough et al., 1992). These overlapping functions can assume various forms based on the requirements of a particular model. From simple operations related to soil erosion to complex functions can be performed through the GIS environment. In addition, GIS is useful for interpreting 3D terrain information, and in particular, DEM generation is used to calculate the inclination angle and slope, which are considered to be the hydrological factors of the erosion model (Desmet and Govers, 1996). Kuhkman et al. (1994) used GIS to successfully carry out massive quantities of data with characteristics of new soil erosion prediction models such as ANSWERS (Areal Nonpoint Source Watershed Environmental Response Simulator) and WEPP (Water Erosion Prediction Project).

It is also recognized that GIS is a tool that helps user decision making and reflects the quantity and quality of the data used. . Data collection and operation for such an appropriate soil loss model will affect the final accuracy of the modeling (Barfield et al., 1989). GIS is gaining popularity as a tool for modeling the spatio-temporal characteristics of soil loss around the world (Burrough et al., 1992).

When using GIS as a data management tool, it can be used in conjunction with existing database and data management system. However, if the reliability and accuracy of the GIS data are secured by using the GIS data in the soil loss assessment study, it is possible to perform a quick and accurate analysis in connection with the soil loss model.

1.2.2 Types of soil loss models

1) Overview

When choosing the appropriate model to understand the process of soil erosion in a watershed, the usefulness and scale of the data entered into the model is a decisive factor (Lo et al., 1992). If input data is given clearly, the model should be derived so that it can perform the optimal function. This model can be divided into empirical model and deterministic model.

Empirical models are derived through empirical data without conceptual or theoretical backgrounds (Barfield et al., 1989). Empirical models are based on important statistical relationships between variables when a rational database is constructed (Morgan, 1995). Therefore, the empirical model produces the most plausible optimal results and yields meaningful results within the range of the given data, but is somewhat difficult to apply to areas outside the data range (Morgan, 1995).

Another model, deterministic model, is based on theoretical background and formulas derived from physical relationships (Barfield et al., 1989). The deterministic model yields numerical results based on mathematical relations of erosion processes (Barfield et al., 1989). The criteria for selecting the model are the availability of the data, the purpose of use, and the size of the data.

It is important to recognize the purpose of each soil loss model, and a heuristic model is generally used when finding practical solutions for field conservation activities, whereas a deterministic model is more useful for an overall understanding of soil loss processes (Hudson , 1981).

2) Types of soil loss models

The soil loss model has been shifted from studying soil loss in a single slope, such as the Universal Soil Loss Formula (USLE), to studying soil loss and retirement in the watershed. The implication of this transition is that it is necessary to explain the erosion process to be more generalized so that it can be applied to decision making. Furthermore, this transition also implies a transition from an empirical soil erosion model to a deterministic model (de Roo et al., 1989). Recently, deterministic soil loss models are divided into two types, concentrated and decentralized factors. When applied to watersheds, the deterministic concentration model assesses the overall response of the watershed to the forces contributing to erosion. On the other hand, the distributed model maintains and manages information related to the area distribution in the spatially varying erosive processes (de Roo et al., 1989).

When selecting the appropriate models for the study among the centralized and distributed models, the research objectives and objectives should be carefully selected.

end. RUSLE

The whole body of the RUSLE (Revised Universal Soil Loss Equation) model was first devised by Wischmeier and Smith (1965) as a USLE model for a large area, and then re-published in 1978 (Wischmeier & Smith, 1978). The USLE model is designed to be used for agriculture and is designed to allow farmers to estimate soil loss for individual compartments and to predict changes in soil loss from the use of conservation plans (Wischmeier and Smith, 1978; Wall et al. , 1978). The USLE model is an empirical model and is designed to perform functions according to the size of the compartment. Both USLE and RUSLE models are based on the same factor and combine two or more factors to calculate the total soil loss.

When the USLE model came out in the 1970s, the USLE model underwent change and development for a better understanding of the erosion process (Laflen et al., 1991). Initially, the USLE model has been changed to have more predictive functions as well as the need to integrate retirement as part of the soil erosion process. The RUSLE model was developed by Renard et al. (1991) because there are many criticisms that the USLE model is 'extraordinary'. Based on the empirical data used to develop and apply the USLE model, the RUSLE model has made much progress and improved its function (Renard et al., 1991). In the RUSLE model, the length factor L and the slope factor S of the erosion slope are regarded separately, making it possible to use the LS factor more appropriately in complicated terrain conditions (McCool et al., 1987; Renard et al., 1997).

(1)

(One)

Where A is the annual average soil loss due to erosion and R is the rainfall erosion factor indicating rainfall intensity and rainfall energy. K is a soil erosion factor and indicates sensitivity for soil erosion under certain conditions. LS is a combination of length factor (L) and slope factor (S) of erosion slope. C is a vegetation cover factor and represents the annual seasonal soil loss ratio according to the previous land use, water pipe, surface cover, degree of surface roughness and wetting condition, and P is a tillage factor such as tillage technique, cultivation type, And so on.

While the USLE model was developed for agriculture, the RUSLE model was developed under various scenarios to be applied to 'watershed' (Chakroun et al., 1993). One of the reasons for using the RUSLE model is that they can effectively combine the use of telemetry data (Chakroun et al., 1993). In addition, the RUSLE model has many advantages when considering the compatibility of the data required by the GIS data processing system, especially when it is applied to the watershed, it is easy to integrate with the GIS, and it is easy to understand and use functional aspects (Hickey et al. 1994).

When used in combination with grid - based GIS, the RUSLE model can predict the loss potential of the soil in grid units, which is useful for understanding the spatial type of soil loss in the watershed.

I. ANSWERS

The ANSWERS model, which was developed in the early 1980s, can evaluate soil separation, transport, and sedimentation processes, and performs modeling based on distributed factors in soil loss assessment. The ANSWERS model is a hydrological model that includes erosion models such as soil separation and transport elements, as well as several descent trajectory equations that are essential for describing surface, subsurface, and stream flows proposed by Huggins and Monke (1966) (Beasley et al., 1980). The ANSWERS model has high accuracy as a tool for predicting soil loss, but it has many data needed for modeling and some data are difficult to obtain. The ANSWERS model has characteristics that allow for consideration of temporal and spatial variation data, and there is a limit to the accuracy of accuracy in the wider watershed. Using the ANSWERS model to simulate erosion in a watershed requires a large amount of memory and numerical computation capabilities and is therefore difficult to apply to large watersheds (Charkroun et al., 1993).

All. WEPP

The WEPP model was developed by Purdue University in 1989 and focused on techniques for predicting future erosion based on understanding of the current erosion process (Laflen et al., 1991). The current WEPP model consists of three modules: profile, watershed, and grid. As a function to directly replace the USLE model, the WEPP profile module includes predicting the loss factors from the currently accepted soil as an important part of the erosion process. The WEPP watershed module has been introduced to manage areas larger than the WEPP profile module covers, and this module is able to control the spatial extent within the watershed boundary. In addition, the WEPP watershed module can also explain the erosion process of mountains, fields and even rivers (Laflen et al., 1991). The last module of the WEPP model, the grid module, was developed to explain the erosion process of watersheds with different spatial boundaries (Laflen et al., 1991).

Like the ANSWERS model, the WEPP model is a deterministic model that places importance on mathematical relationships. However, it differs from the ANSWERS model in that the WEPP model has taken into account the erosion process alongside rivers. The WEPP model is calculated with four input files, such as climate, soil characteristics, slope and vegetation characteristics, and has many similarities to the data required by the ANSWERS model.

The WEPP model is a recent model for explaining the erosion process, which can be used for a variety of topographical representations that could not be performed by the ANSWERS and RUSLE models (Laflen et al., 1991).

1.2.3 Selection of soil loss model

In order to select the model for predicting the soil loss, the spatial extent of the target area, the construction of the GIS data, and the form of the final analysis result should be considered together. The Nakdong River and Nankang Dam basins, which are the study sites, have a spatial range of the watershed size, and most of the GIS data are well established. Therefore, in this study, the RUSLE model was selected as a model for the evaluation of soil loss after a comprehensive review of the scale of the site and the possibility of using GIS data.

Although the WEPP model has the advantage that it can predict the erosion inside the stream along with the river and the grazing effect can be included in the erosion process, the WEPP model does not provide a convenient environment to use the GIS. The RUSLE model is used, which allows various combinations of possible factors. In addition, the WEPP model requires specific information in the watershed, which is another reason for choosing the RUSLE model. In addition, the ANSWERS model does not meet the objectives of this study to assess general soil loss in that the necessary data are rigorous, too vast and perform event-oriented soil loss prediction functions.

The argument-based RUSLE model has a structure that can be easily linked to the GIS data of the Nam River dam basin. The main reason for selecting the RUSLE model is that it can predict the soil loss under various scenarios. In addition, by operating the RUSLE model in conjunction with the grid-based GIS, it is possible to query and extract regions of many soil loss potentials in a grid-by-grid manner, and to identify spatially separated regions (Wischmeier and Smith 1978 ; Desmet and Govers, 1996).

Recently, various information constructed through GIS and remote sensing technology are very useful for calculating the factor of RUSLE model. Especially, length factor (L) of erosion slope, slope factor (S) of erosion slope and vegetation cover factor (C) (Desmet and Govers, 1996).

Recently, there has been a serious problem of water pollution due to landslides and landslides in the watershed and muddy water due to concentrated rainfall due to climate change. In addition, the sediments that have flowed into the river have been degraded in the river, causing flooding damage in the flood.

Therefore, it is necessary to prepare earthquake disaster countermeasures according to landslide, soil collapse, loss of soil. In order to reduce such soil loss, various measures such as dam construction, field foundation maintenance, and coating improvement are needed in the country.

An object of the present invention to solve the problems of the prior art is to reduce the land cover of urban areas, agricultural lands (rice fields, fields), forest areas, roads, The land cover map used for soil loss analysis is constructed using the image classification by item [forest-NDVI, water-near-infrared band, agricultural land (rice field and field) -wet index, urban area (urban area) -thermal infrared band] (DEM, Digital Elevation Model), and land cover map. The RUSLE factor is calculated for the RISLE (Revised Universal Soil Loss Equation) model, and the soil loss is calculated. The results of this study are summarized as follows. First, we investigate the subsoil loss assessment method which is vulnerable to the earthquake loss, To provide a loss assessment methods, GIS-based landslide to establish a sense of measure.

In order to achieve the object of the present invention, a GIS-based soil loss evaluation method is classified into (a) a land cover class including a city (urban area), a cropland (rice field, field), forest, road, Constructing a land cover map used for soil loss analysis by applying a star image classification technique; And (b) construct GIS data of soil, digital elevation model (DEM), and land cover map using ArcGIS software, and apply the GIS spatial analysis technique required for GIS-based RUSLE (Revised Universal Soil Loss Equation) (A), and by analyzing the RUSLE factor for each geographical area, it is possible to extract subwatershed vulnerable to soil erosion in terms of soil, topography and land cover. .

The GIS-based soil loss evaluation method according to the present invention is a method for estimating the amount of soil loss according to the land cover items such as urban land, agricultural land (rice field, field), forest area, road, Image classification [Land-NDVI, water-near-infrared band, agricultural area (rice fields) - wet index, urban area (urban area) - thermal infrared band] is used to construct a land cover map used for soil loss analysis. (GIS) data such as elevation model (DEM) and toe cladding were constructed to calculate the RUSLE factor required for the RISLE (Revised Universal Soil Loss Equation) model, and then the soil loss amount (A) was calculated. And to establish countermeasures to reduce soil erosion by using subwatershed analysis data that is vulnerable to soil erosion including the amount of soil erosion by each site. There is an effect that is used in determining means.

In GIS - based RUSLE model, the largest amount of soil loss was simulated at 163,025 ton / yr in the Imcheon subwatershed in one embodiment. In the analysis of unit soil loss, which reflects the potential of soil loss, yr ², which is the most vulnerable to the loss of soil in the event of rainfall.

In this study, a land cover map was constructed from the latest satellite image (landsat image) by using image classification technique for each land cover item such as urban area, agricultural area, forest area, road area, and water area. Also, GIS data such as digital elevation model (DEM) were constructed for the soil, and the factors required for the RUSLE model were calculated and used for soil loss calculation. And we could examine the subwatershed vulnerable to the earth loss by simulating the amount of soil loss by each site.

Therefore, the subwatershed analysis data, which is vulnerable to soil erosion, including the soil erosion amount by each site, can be very useful for establishing a business scale for establishing measures to reduce earth loss and to select major business areas.

1 is a view showing "raindrops: a source of a physical erosion process ".
Fig. 2 is a view showing the "erosion process: movement of water and soil particles ".
Figure 3 shows Erickson's K value estimation triangle plot.
4 is a diagram showing a "contour line length model along the direction of a cell ".
Fig. 5 is the location map of the Namgang dam basin for analyzing the soil loss by selecting the Namgang dam basin as a study area.
Fig. 6 shows the model of the Namcheon Dam constructed for the analysis of factor and soil loss in RUSLE model.
7 is a DEM distribution chart of the Namgang dam basin.
8 is a slope distribution map of the Namgang dam basin constructed through GIS lattice analysis.
Fig. 9 is a detailed soil map of the Nanjang dam basin.
Figure 10 is a land cover map of the Nanjang dam basin.
11 is a view showing a soil loss evaluation process using a RUSLE (Revised Universal Soil Loss Equation) model.
12 is a view showing the soil erosion factor (K) in the Nam River dam basin.
13 is a view showing the erosion slope length factor (L) of the Namgang dam watershed.
14 is a view showing an erosion slope gradient factor (S) of the Namgang dam watershed.
15 is a view showing the vegetation cover factor (C) of the Nanjang dam basin.
Fig. 16 is a view showing the vegetation cover factor (P) of the Nanjang dam basin.
Fig. 17 is a distribution chart of the amount of soil loss in the Namgang dam watershed.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings.

2. Analysis of soil loss based on image classification data

2.1 Research subject

For soil loss analysis, many GIS data such as precision soil map, digital elevation model (DEM), slope, and land cover are required. In addition, there is a lot of limitation in evaluating the soil loss for the entire Nakdong River basin because the calculation processing by spatial analysis is required for each data. Therefore, in this study, as shown in FIG. 5, in this study, the Namchang dam basin, which is a part of the Nakdong River, was selected as a study site and the soil loss was analyzed.

The Namgang Dam basin is located in the lower part of the existing Namkang Multipurpose Dam, about 80 km upstream from the Nakdong River confluence point. The basin area is 2,285㎢, accounting for 9.6% of the entire Nakdong River basin area.

The average annual temperature is 13 ℃, and it is the Dow area accompanied by heavy rainfall and typhoon due to the monsoon climate and the southern coast turbulence in summer. The average annual rainfall is 1,416 ㎜. The administrative districts of the Nanjang dam basin are located at 3 o'clock 1 o'clock and 11 o'clock 67 municipalities. The main industry is agriculture and forestry, and some industries are developed in Jinju city.

Fig. 5 is the location map of the Namgang dam basin for analyzing the soil loss by selecting the Namgang dam basin as a study area.

Fig. 6 shows the model of the Namcheon Dam constructed for the analysis of factor and soil loss in RUSLE model.

2.2 Building GIS Data

As shown in Table 1, DEM, precision soil map, and land cover map were used for GIS data constructed in this invention for soil loss analysis.

   Name of material         Data details        source     DEM - 1 / 5,000 digital topographic map (340 degree) - National Geographic Information Service  Precision soil map - 1 / 25,000 Precision Soil (20 sheets) - Rural Development Administration  Land cover - Image Classification Method by Object Class - Build your own

To construct a precision DEM, a 1 / 5,000 digital topographic map constructed by the National Geographic Information Institute was used. We first used the dxfarc command in ArcGIS software to check the contour and elevation layers on a 1 / 5,000 digital topographic map and then convert the contours and elevations in DXF format to ArcGIS coverage. When converting the contour and elevation values of DXF format into coverage, ACODE and elevation values are generated as a separate file called XCODE. Therefore, in case of contour line, ACODE file is connected to AAT file. In case of elevation, XCODE You have linked the file to a PAT file. Programmed with AML (Arc Macro Language) to automatically extract ArcGIS coverage from the contour and elevation layers of a 1 / 5,000 digital topographic map in DXF format. The TIN (Triangle Irregular Network) was constructed from the elevation and elevation, and the digital elevation model (DEM) with 10m resolution was constructed from the TIN.

Fig. 7 shows the DEM distribution map of the Namgang dam basin. The elevation distribution was 25 ~ 1,911m. 8 is a slope distribution map of the Namgang dam basin constructed through GIS lattice analysis.

Fig. 9 is a detailed soil map of the Nanjang dam basin.

The precision soil map of the Nam River dam basin was constructed by superimposing 1 / 25,000 scale precision soil maps constructed by computerization at the Agricultural Science and Technology Center of RDA. The precision soil layer also contains information on soil name, soil code, soil, slope, drainage, land use, and geological characteristics. Therefore, the soil erosion factor, which is the input factor of the soil erosion model, can be calculated by classifying the soil group secondarily based on such soil property information.

Fig. 9 is a detailed soil map of the Namgang dam basin. Detailed information on each soil code is shown in Appendix 1.

The covering condition constituting the indicator plays a decisive role in the loss of soil and the generation of turbid water due to the rainfall. Therefore, in order to overcome the limitation of the Landsat image with the resolution of 30m, the accuracy and accuracy of the land cover is very important. In this study, the land cover map constructed by the image classification technique by the land cover object class, The coatings were combined and constructed as shown in Fig. Figure 10 is a land cover map of the Nanjang dam basin.

Table 2 shows the results of the land cover analysis for the Nam River dam basin.

As a result of land cover analysis including urban area (agricultural area), agricultural area (agricultural area), forest area, grassland, marsh area and paddy field, more than 73% of the Namchang dam basin consisted of forest area, 19.817%, and the ratio of urbanization area was 2.382%.

Land cover (major classification) Land cover (subdivision) Categories code area
(㎢) Occupancy rate
(%) Categories code area
(㎢) Occupancy rate
(%) Poetry
Dry area
10 54.427 2.382 Residential area 11 36.271 1.588 Industrial area 12 1.52 0.067 Commercial area 13 3.586 0.157 Amusement facility area 14 0.117 0.005 Traffic area 15 10.292 0.45 Public utility area 16 2.641 0.116 Agricultural area 20 452.753 19.817 Rice field 21 303.166 13.269 field 22 85.96 3.762 A highland field - 30.137 1.319 House plantation 23 6.787 0.297 orchard 24 19.819 0.867 Other plantations 25 6.884 0.301 Forest area 30 1677.227 73.411 Broad-leaved forest 31 599.348 26.233 Coniferous forest 32 760.479 33.286 Mixed forest 33 317.4 13.892 Grassland 40 23.556 1.031 Natural grassland 41 11.032 0.483 golf course 42 Other grasslands 43 12.524 0.548 marsh 50 20.042 0.877 Inland wetland 51 20.042 0.877 Me 60 22.953 1.005 Mining area 61 1.957 0.086 Others 62 20.996 0.919 water system 70 33.742 1.477 water system 71 33.742 1.477 system 2284.7㎢ (100%)

The GIS based soil loss assessment method consists of (a) applying the image classification technique by land cover class including the urban (urban), agricultural (rice field, Establishing a land cover used for analysis; And (b) construct GIS data of soil, digital elevation model (DEM), and land cover map using ArcGIS software, and apply the GIS spatial analysis technique required for GIS-based RUSLE (Revised Universal Soil Loss Equation) (A), and by analyzing the RUSLE factor for each geographical area, it is possible to extract subwatershed vulnerable to soil erosion in terms of soil, topography and land cover. .

In the step (a), the image classification by the land cover class is classified into the forest-NDVI, the water-near-infrared band, the agricultural land (paddy field) -the wet index, and the urban (urban) -thermal infrared band from the satellite image, Forests are extracted by applying a Normalized Difference Vegetation Index (NDVI) in combination with a near-infrared band and a red band, and the water bodies are classified by applying the most sensitive NIR , The water layer is extracted by applying a cell expansion technique to take into account the influence of the water level change due to the rainfall, the agricultural land (paddy field) is extracted using the wetting index, and the urban area .

The process for evaluating the soil loss using the RUSLE (Revised Universal Soil Loss Equation) model is based on the GIS DB used for evaluating the respective RUSLE factors including the precision soil map, the digital elevation model (DEM) Using a coating,

The RUSLE factors include rainfall erosion factor (R), soil erosion factor (K), erosion slope length factor (L), erosion slope factor (S), vegetation cover factor (C) .

2.3 RUSLE model

The equation for calculating the soil loss (A) using the RUSLE model is shown in (2).

Soil loss A has units of ton / ha / yr and is composed of rainfall erosion factor (R), soil erosion factor (K), length factor of erosion slope (L), slope factor of erosion slope (S) C) and cultivation factors (P).

(2)

(2)

2.3.1 Rainfall erosion factor (R)

Rainfall erosion factor (R) is influenced by rainfall and rainfall intensity. It is known that rainfall intensity has more influence on rainfall erosion factor than rainfall amount and relatively non - erosion result below constant rainfall intensity. From the results of the study, the rainfall intensity expected to erode is about 25 ㎜ / hr. Also, in temperate climates, approximately 95% of rainfall has non-erosive rainfall intensity, while in tropical regions approximately 60% of rainfall is known to have non-erodible rainfall intensity (Hudson, 1977).

The rainfall erosion factor (R) is calculated by using rainfall or rainfall intensity. Rainfall was used in this study.

Rainfall erosion factor (R) is the product of the sum of the rainfall energy during the simulation period and the maximum rainfall intensity of 30 minutes. Rainfall energy, E, is expressed by Eq. (3).

(3)

(3)

Where E is kinetic energy (mt / ha / cm) at a given rainfall intensity X and X is rainfall intensity (cm / hr). The rainfall erosion factor for a specific storm event can be evaluated by using Equation (4), which is calculated for each storm after the rainfall multiplied by Equation (3) in energy units.

(4)

(4)

은 30분 지속 최대강우강도(cm/hr)이다. Where R is the rainfall erosion factor (J / m < 2 >),

Is the maximum rainfall intensity (cm / hr) sustained for 30 minutes.

The study on rainfall erosion factor (R) was carried out by Jeong Pil Kyun et al. (1983). Rainfall erosion factors were determined over three years from 1979 to 1981 by selecting 51 out of the stations under the Korea Meteorological Administration.

2.3.2 Soil erosion factor (K)

The relationship between soil erosion factor (K) and soil particle size distribution is presented in triangulated form by Wischmeier and Smith, pioneer of general soil loss formula studies. Erickson (1997) complements this and completes the triangle diagram shown in Fig. 3, and the soil erosion factor obtained from this is in the range of 0.02 to 0.69.

3 is Erickson's K-estimated triangular plot.

In the previous studies, the value of soil erosion factor (K) based on the rough soil map of 1 / 250,000 scale was mainly used, but it is true that there is a limit to express the distribution of precise soil erosion factor in the rough soil map. Especially, considering the similarity of topographical relief in the domestic watershed, the proportion of the soil characteristics has a great influence on the soil loss. Therefore, in this study, soil erosion factor K was calculated from the soil group based on 1 / 25,000 precision soil map by using Erickson 's K - value triangle chart.

2.3.3 Topographical factors

The topographic factor is divided into the length factor (L) of the erosion slope and the slope factor (S) of the erosion slope. First, the length factor (L) of the erosion slope represents the ratio of the horizontal slope to the slope of 22.13m, which is the slope of the unit slope. There are many empirical equations presented to calculate this, but the representative empirical equation There are Renard and Desmet & Govers expressions. First, Renard et al. (1997) developed equation (5).

,

,

는 침식사면의 경사각이다.m is the exponent of the erosion slope length,

Is the inclination angle of the erosion slope.

(5)

(5)

는 셀에 대한 침식사면의 길이인자이며,

는 셀크기다.

Is the length factor of the erosion slope for the cell,

Is the cell size.

This formula uses the number of cells entering the observation cell and uses the length of the cell as a multiple flow algorithm to determine the total slope of the segment. This cell-based technique for calculating slope lengths is derived from the segment formulas of Foster and Wischmeier (1974). The use of cell-based algorithms is more objective than the existing evaluation techniques that can only define subjectively where the slope starts and ends. This method can take into account the bifurcation and confluence of the flow and the complex natural topography. Especially, it can extract effectively using GIS cell based model.

Desmet and Govers (1996) developed Equation (6) to calculate the length factor (L) of the erosion slope.

(6)

(6)

는 셀에 대한 침식사면의 길이인자,

는 셀에 유입되는 상류기여면적,

는 셀크기,

는 흐름방향에 직교하는 등고선 길이로 (

)),

는 셀의 방향이다. 이 공식에서, 흐름방향에 직교하는 등고선 길이

는 도 4와 같다.

Is the length factor of the erosion slope for the cell,

Is the upstream contribution area into the cell,

Cell size,

Is a contour line length orthogonal to the flow direction (

)),

Is the direction of the cell. In this formula, contour lengths perpendicular to the flow direction

Is shown in FIG.

4 is a diagram showing a contour line length model along the direction of a cell.

Here, the upstream contribution area A flowing into the observation cell is calculated as shown in equation (7).

(7)

(7)

는 관측셀 i에 대한 기여면적, A는 이용 가능한 상류기여면적,

는 이웃한 셀의 경사,

는 가중인자(직교방향은 0.5, 대각선 방향은 0.354)이며

는 하류방향의 셀수이다.here,

Is the contribution area for observation cell i, A is the available upstream contribution area,

Is the slope of the neighboring cell,

Is a weighting factor (0.5 in the orthogonal direction and 0.354 in the diagonal direction)

Is the number of cells in the downstream direction.

The Desmet and Govers equation uses the number of cells entering the observation cell, which is similar to the Renard equation, but it is a multi-flow direction model considering the direction of flow into the cell. The Desmet and Govers equations also have the advantage of being able to efficiently extract the results of a given cell unit under GIS, rather than just the entire watershed.

The slope factor (S) of the erosion slope indicates the effect of the slope slope on the slope failure. The slope factor (S) of the erosion slope is sensitive to the soil loss factor compared to the length factor (L) of the erosion slope. The relationship between the slope and the soil loss depends on the density of the vegetation and the size of the soil particles.

In the USLE, which predicted erosion on agricultural land, the average value of the S factor was not unreasonable because the agricultural land generally showed an even slope distribution. However, in case of RUSLE which predicts erosion on the watershed, Considering the point, it is desirable to use a cell-type value that can evaluate the S factor locally. The provision of cell-based S values can be effectively handled more quickly by the introduction of GIS (Andrew, 1998).

McCool et al. (1987) developed a RUSLE model using Eq. (8) to solve the problem of excess soil loss caused by the very high evaluation of slope factor (S) of erosion slope in USLE.

(8)

(8)

Here,? Is the inclination angle of the erosion slope.

Equation (8) was developed by considering 20 study areas and taking into consideration the slope effect in the field and river. The slope of the target area has a slope range of 0.1% to 32%, but this formula is actually developed for applications where slope is less than 18%. McCool's formula does not apply to areas with a slope greater than 50%, ie, a slope greater than 26.6 °. According to Liu 's study, it is suggested that the soil slope increases linearly with the inclination angle in the slope region of 9 ~ 50%. Liu et al. (1994) proposed equation (9) for the slope region of 9 to 50% in order to express the result in the form of an equation.

(9)

(9)

Nearing (1997) proposed equation (10) that can be applied to the RUSLE model by generalizing the existing linear functions applied to slopes of less than 25% by applying Eqs. (8) and (9) This formula also confirmed that McCool et al. (1987) and Liu et al. (1994) show better results than those of equations even when the slope exceeds 25%.

(10)

(10)

2.3.4 Vegetation Covering Factor (C)

The vegetation cover factor (C) is a factor that reflects vegetation cover and crop condition effects on soil loss. The effect of the covering condition of the surface of the earth on the soil loss is due to the reduction of the kinetic energy of the raindrops caused by vegetation, the reduction of the transport of soil particles due to the reduction of the flow rate of the surface currents and the enhancement of sedimentation, the inhibition of movement of soil mass by the vegetation roots, And the reduction of soil moisture due to the enhancement of biological activity in the soil. Since the effects of vegetation vary by season, crops, and crop growth, this should be considered seasonally for estimating short-term soil loss or sediment loss, but it may be sufficient to consider only the annual average effect in long-term models.

Land cover can be classified into urban area, agricultural area, forest area, road, sleep, etc., and it is closely related to vegetation except for urban areas and roads where there are many artificial changes.

It is impossible to artificially control natural phenomena and natural conditions related to rainfall erosion factor (R), soil erosion factor (K), and topographic factor among the soil loss factors. However, the vegetation cover factor (C) has the function of suppressing the soil loss in that anthropogenic regulation is possible to some extent, and this function can change into a structure resistant to the soil loss through artificial forest formation.

The effect of vegetation cover changes over time. The values of vegetation cover factors in a particular area depend on the type of vegetation, the state of vegetation growth, the type of cultivation and management factors. The ratio of the vegetation cover factor C value is about 1.0 in the same area as in the ground before vegetation growth, and is low in dense forested areas and high grain density areas.

(1999) suggested the values of vegetation cover factor for each land cover as shown in Table 3. However, the vegetation cover factor suggested by the New Territories is set too high for the values of forest and grassland compared to the values of paddy field and field. In particular, the land cover is used to compute the tillering factor P, and the remaining areas except the paddy field, the field and the water level area are classified by the contour tillering method. In this case, considering the vegetation covering factor (C) There is a contradiction in the area where the soil loss is larger than the rice field and the field. Therefore, in this study, the value of vegetation cover factor (C) of Table 4 published by the US Department of Agriculture was used (Park, Kyung-Hoon, 2003). In particular, since Table 4 can consider the subdivision of land cover form as compared with Table 3, it is possible to apply the land cover of the subdivision system constructed in this study more precisely.

Table 3 shows the value of vegetation cover factor for each type of land cover classified as urban area, paddy field, field, forest, grassland,

Land cover type Vegetation Covering Factor (C) Shiga Area 0.01 Rice field 0.30 field 0.40 Forest 0.10 Grassland 0.20 Nagai 0.50 water system 0.00

Table 4 shows the vegetation cover factor (US Department of Agriculture) for each land cover classified as urban, agricultural, forest, grassland, or inland.

Land cover type Vegetation Covering Factor (C) Main Category Middle class City Low density 0.002 High density 0.001 Industrial area 0.000 Traffic area 0.000 Farmland field 0.400 Rice field 0.300 orchard 0.200 Forest softwood 0.009 Hardwood 0.004 Mixed forest 0.007 Grassland Park, cemetery, pasture 0.050 Nagai Nagai 1,000 Water Rivers and reservoirs 0.000

2.3.5 Cultivation factors (P)

The cultivation factor (P) considers the effects of cultivated land types such as paddy fields and fields and slopes of cultivated land. Contained cultivars include contouring, contouring, and terracing. This type of cultivated land plays an important role in regulating soil loss, and the variety of cultivated land varies depending on local characteristics. Generally, the field shows the contour plot type and the rice field shows the terrace plot type. However, in the case of field, it is necessary to understand the type of cultivation that is mainly used in the area through field survey as it shows various forms according to the region.

The contour line cultivation method is a cultivated form with constant direction and slope, and it can be estimated that soil loss is the highest in that it causes the soil transfer immediately in case of rainfall. This type of tillage is a plate form with a constant slope from the top to the bottom of the cultivated land, and it is the most likely form of erosion when considering stability. Contour cultivation is similar to contouring but it is relatively non-erosive rather than contour tiller because it is along contour lines. The terraced tillage method is the most effective way to alleviate soil loss by using a short distance and a steep slope angle within the cultivated land in the form of stairs. The erosive soil is moved downstream along the slope, where the erosion in the terrace is drastically reduced when the terrace is encountered and the soil is accumulated on the terrace. If terraces do not exist, then the multitude of tofu transported to a single outlet will cause disasters such as landslides, but if the terraces are present, the multitude of tofu transported to one outlet will provide intensive kinetic energy Disruption and induction of multiple outflows minimizes catastrophic disasters such as sudden landslides.

Thus, the tillering factor P depends on the cultivation type and the slope slope, and the values are shown in Table 5 (Shin et al., 1999). Table 5 shows the cultivation factors (P) according to the cultivation type and slope.

slope(%) Contour Contour target Terrace 0.0 to 7.0 0.55 0.27 0.10 7.0 to 11.3 0.60 0.30 0.12 11.3 to 17.6 0.80 0.40 0.16 17.6 to 26.8 0.90 0.45 0.18 26.8 or higher 1.00 0.50 0.20

The tillering factor (P) used was 0.55 for contour line cultivation, 0.27 for contour line cultivation, and 0.10 for terrace cultivation in the case of 0.0 to 7.0 slope (%) using the cultivating factor (P) , Contour line cultivation 0.60, contour line cultivation 0.30, and terraced cultivation 0.12 were used for the slope of 7.0 ~ 11.3 slope (%), 0.80 for contour line contouring 0.30 for contour line 11.3 ~ 17.6, , Contour line cultivation (0.90), contour line cultivation (0.45) and terrace cultivation (0.18) were used for slope (17.6 ~ 26.8) ) Value.

2.4 Earth Loss Analysis

Figure 11 shows a process for evaluating soil loss using a RUSLE (Revised Universal Soil Loss Equation) model. The GIS DB used to evaluate each RUSLE factor includes precision soil, digital elevation model (DEM), and land cover map. In particular, since the slope is used to evaluate the tillage factor (P), the distribution characteristics of the slope are also evaluated from the DEM data by a certain resolution.

Rainfall erosion factor (R) has two methods to apply rainfall and rainfall intensity. In this study, we used the method to utilize the rainfall data. For this purpose, we searched WAMIS (Water Management Information System) to find the monthly precipitation data for the last 10 years for the Pearl River Observation Station Respectively. The rainfall erosion factor (R) can be calculated by applying the equation (2.1) to the annual mean rainfall using this method.

(2.1)

(2.1)

은 연강우량(㎜/yr)이다. here,

Is the annual rainfall (mm / yr).

Year of observation Monthly rainfall (mm / month) Sum January February In March April In May June In July August September October November December 2013 20 77 58 75 251 91 202 136 56 84 53 One 1,104 2012 7 12 130 172 47 52 277 376 462 44 60 74 1,713 2011 0 0 34 258 168 244 737 450 58 93 147 3 2,192 2010 29 141 131 189 198 53 375 364 304 70 5 20 1,879 2009 12 59 79 114 144 181 695 104 35 41 31 28 1,523 2008 44 10 38 71 105 307 83 114 28 30 9 2 841 2007 One 57 86 21 100 40 208 519 432 92 0 32 1,588 2006 0 0 74 131 274 163 1,012 139 90 35 0 0 1,918 2005 10 46 87 99 91 114 240 299 56 8 34 0 1,084 2004 0 54 36 97 129 164 212 355 189 7 73 34 1,350

The soil erosion factor (K) was calculated by applying the Erickson diagram after analyzing the contents of sand, silt, clay (Clay), and sand in the soil surface by using the precision soil map. In addition, the topographic factor (LS) was evaluated using Desmet & Govers equation and Nearing equation based on DEM data. In particular, the length factor (L) of the erosion slope was implemented to simulate soil erosion more realistically . The vegetation cover factor (C) was evaluated by the combination of the results of the analysis by the image classification technique of the land cover object class and the middle class land cover of the environment part in this study. The tillage factor (P) , Field, and other areas.

Table 7 shows the results of the RUSLE factor analysis for the Nam River dam basin, and Figures 12 to 16 show the RUSLE factor distribution.

factor at least maximum Average Standard Deviation K 0.000 0.505 0.322 0.092 L 0.248 3.394 1.444 0.878 S 0.049 15.128 6.509 4.116 C 0.000 1,000 0.070 0.130 P 0.100 1,000 0.800 0.315

12 is a view showing the soil erosion factor (K) in the Nam River dam basin.

13 is a view showing the erosion slope length factor (L) of the Namgang dam watershed.

14 is a view showing an erosion slope gradient factor (S) of the Namgang dam watershed.

15 is a view showing the vegetation cover factor (C) of the Nanjang dam basin.

Fig. 16 is a view showing the vegetation cover factor (P) of the Nanjang dam basin.

In order to analyze the distribution characteristics of the RUSLE factor for each Namgang dam owned area as shown in Tables 8 and 9, the Zone grid for each subwatershed was constructed and evaluated as follows.

The soil erosion factor (K) was the highest at 0.377 in the subwatershed downstream of the Deokcheon River, and the lowest at 0.297 at the confluence of the Yiwon and Wichon rivers. Therefore, the soil characteristics of the subwatersheds at Yiwang Yiechen confluence area were found to be resistant to rainfall erosion, and the downstream subwatershed of Deokcheon river showed soil characteristics vulnerable to rainfall.

The length factor (L) of the erosion slope was the highest at 1.583 in the Sichon subwatershed and the lowest in the Namchang dam subwatershed at 1.326. In addition, sloping factor (S) of erosion slope was highest at 8.900 in the upstream area of the Deokcheon River and lowest at 4.862 in the Namgang Dam subwatershed.

Own Station Name factor at least maximum Average Standard Deviation Namgang Dam
(165.623㎢) K 0.000 0.505 0.302 0.138 L 0.249 3.367 1.326 0.795 S 0.049 15.085 4.862 4.019 C 0.000 1,000 0.067 0.126 P 0.100 1,000 0.743 0.318 Upstream of Namgang Dam
(70.840㎢) K 0.000 0.500 0.336 0.111 L 0.251 3.372 1.405 0.838 S 0.049 15.032 5.030 4.063 C 0.000 0.400 0.084 0.140 P 0.100 1,000 0.725 0.340 Upper Nam River
(160.053㎢) K 0.000 0.445 0.302 0.056 L 0.249 3.365 1.439 0.894 S 0.049 15.116 7.575 4.010 C 0.000 1,000 0.081 0.209 P 0.100 1,000 0.853 0.274 Upstream of the Deokcheon River
(105.322㎢) K 0.000 0.450 0.367 0.086 L 0.253 3.367 1.577 0.933 S 0.049 14.927 8.900 3.679 C 0.000 0.890 0.027 0.088 P 0.100 1,000 0.930 0.120 Down the Deokcheon River
(142.971㎢) K 0.000 0.490 0.377 0.085 L 0.251 3.351 1.457 0.874 S 0.049 14.904 5.974 4.091 C 0.000 1,000 0.076 0.138 P 0.100 1,000 0.775 0.330 Shimcheon
(264.108㎢) K 0.000 0.490 0.335 0.078 L 0.248 3.394 1.461 0.877 S 0.049 15.128 6.165 4.220 C 0.000 0.890 0.107 0.217 P 0.100 1,000 0.774 0.338 Sancheong water mark
(195.674㎢) K 0.000 0.490 0.327 0.097 L 0.248 3.359 1.423 0.874 S 0.049 15.111 6.369 4.161 C 0.000 1,000 0.083 0.142 P 0.100 1,000 0.769 0.329

Own Station Name factor at least maximum Average Standard Deviation Sichuan cloth
(106.255㎢) K 0.000 0.460 0.338 0.100 L 0.248 3.357 1.583 0.935 S 0.049 15.127 8.400 3.522 C 0.000 1,000 0.030 0.094 P 0.100 1,000 0.921 0.206 New spring
(172.173㎢) K 0.000 0.490 0.311 0.089 L 0.248 3.365 1.403 0.863 S 0.049 15.116 5.739 3.903 C 0.000 1,000 0.087 0.150 P 0.100 1,000 0.760 0.339 Yangcheon
(252.530㎢) K 0.000 0.505 0.301 0.089 L 0.249 3.376 1.409 0.859 S 0.049 15.111 6.496 4.088 C 0.000 1,000 0.069 0.128 P 0.100 1,000 0.803 0.317 Yangcheon Junction
(99.951㎢) K 0.000 0.490 0.306 0.094 L 0.248 3.378 1.455 0.884 S 0.049 15.116 6.029 3.961 C 0.000 0.890 0.069 0.131 P 0.100 1,000 0.794 0.311 Imcheon
(218.217㎢) K 0.000 0.490 0.334 0.097 L 0.248 3.393 1.522 0.920 S 0.049 15.123 7.527 3.696 C 0.000 1,000 0.064 0.148 P 0.100 1,000 0.857 0.279 Yi Yang
(177.939㎢) K 0.000 0.490 0.301 0.068 L 0.248 3.392 1.421 0.874 S 0.049 15.116 6.487 4.079 C 0.000 0.890 0.087 0.162 P 0.100 1,000 0.784 0.320 Hamyang Weichon Junction
(161.765㎢) K 0.000 0.490 0.296 0.066 L 0.252 3.340 1.402 0.859 S 0.049 14.986 5.983 4.026 C 0.000 0.890 0.093 0.155 P 0.100 1,000 0.755 0.336

Namkang Dam subwatershed is located in the most downstream area, and relatively topographical relief and slope characteristics are gentle, indicating that the geomorphic characteristics are strong against rainfall. The vegetation cover factor (C) was highest at 0.107 in the Lamcheon subwatershed and 0.027 in the upstream area of the. From this result, it can be seen that the farm area of Ramsun subwatershed is more distributed than other subwatersheds.

The cultivating factor (P) was the highest in the subcatchment area of the Deokcheon River at 0.930 and the lowest at 0.725 in the upstream area of the Namgang Dam. Tillage factor (P) is a factor which depends on the cultivation type and terrain slope. It is found that the cultivation condition and topography of the upstream area of the Duckcheon River are vulnerable to soil loss.

Soil loss was calculated from the RUSLE factor calculated by GIS spatial analysis. FIG. 17 is a distribution chart of the soil loss amount in the Nam River dam basin, and Table 10 shows the results of the soil loss analysis.

division at least maximum Average Standard Deviation value
(ton / ha) 0 10,443 25 92

The Zone grid was constructed for each subwatershed in order to analyze the soil erosion characteristics according to the possessing area in Table 11.

As a result, the largest amount of soil loss was simulated at 163,025 ton / yr in the subcatchment area of Imcheon subwatershed and 29,341 ton / yr in the upstream area of Namchang Dam. The assessment of soil loss per unit area is the basis for evaluating the potential for soil loss. In the estimation of unit soil loss, the area of Imcheon subwatershed was the highest at 747 ton / yr2 km, and the upstream area of Namchang Dam was the lowest at 414 ton / yr2 km.

Own Station Name Soil loss (ton / ha) Soil loss
(ton / yr) unit
Soil loss
(ton / yrkm2) at least maximum Average Standard Deviation Namgang Dam
(165.623㎢) 0 7,820 19 75 69,695 421 Upstream of Namgang Dam
(70.840㎢) 0 1,767 18 50 29,341 414 Upper Nam River
(160.053㎢) 0 8,012 30 130 108,923 681 Upstream of the Deokcheon River
(105.322㎢) 0 4,095 20 59 47,186 448 Down the Deokcheon River
(142.971㎢) 0 8,955 29 117 94,329 660 Shimcheon
(264.108㎢) 0 4,492 23 76 139,140 527 Sancheong water mark
(195.674㎢) 0 7,721 29 102 129,838 664 Sichuan cloth
(106.255㎢) 0 8,241 23 88 55,731 525 New spring
(172.173㎢) 0 8,775 24 82 95,783 556 Yangcheon
(252.530㎢) 0 7,980 25 82 141,062 559 Yangcheon Junction
(99.951㎢) 0 2,677 21 57 48,804 488 Imcheon
(218.217㎢) 0 10,443 33 130 163,025 747 Yi Yang
(177.939㎢) 0 3,607 29 86 115,671 650 Hamyang Weichon Junction
(161.765㎢) 0 3,338 25 71 90,808 561 Sum 1,329,337

Recently, there has been a serious problem of water pollution due to landslides and landslides in the watershed and muddy water due to concentrated rainfall due to climate change. In addition, the sediment that has flowed into the river has decreased the flowability of the river, causing flood damage in the flood. In order to reduce such loss of soil, the government is establishing various measures such as construction of the dam, construction of the field foundation, improvement of the cloth.

For example, land cover items include urban areas (urban areas), agricultural lands (rice fields), forest areas, roads, and water bodies.

In the present invention, the image classification (forest-NDVI, water-near-infrared band, paddy field-wetting index, urban area-urban-rural area) of the land covering items such as urban area, agricultural area, forest area, road, Thermal infrared band] technique to construct a land cover map. In addition, the GIS data such as the soil elevation, the DEM, and the land cover map were constructed to calculate the factors required for the RUSLE model and then used for calculating the soil loss. And, we could review the subwatershed vulnerable to the loss of soil by simulating the amount of soil loss by each station.

Therefore, the subwatershed analysis data, which is vulnerable to soil erosion, including the soil erosion amount by each site, can be very useful for establishing a business scale for establishing measures to reduce earth loss and to select major business areas.

[conclusion]

In this study, we propose an image classification (forest - NDVI, water - near - infrared band, paddy field and wet - wetland) of land cover class (urban area, (City - area) - thermal infrared band), and the land cover was constructed by using the ArcGIS software and analyzed with the GIS - based RUSLE model. The main results and conclusions of this study are as follows.

1. By applying the image classification technique according to the land cover class from the recently taken satellite image (landsat image), it was possible to secure the up-to-date property of the land cover used for the analysis of soil loss.

Especially, after classifying the water by applying NIR which is most sensitive to water (water), we could effectively extract the water layer by applying the cell expansion technique to consider the influence of the water level change by rainfall .

In addition, forests could be extracted by applying Normalized Difference Vegetation Index (NDVI), which is a combination of NIR band and red band.

Cropland (paddy fields) could be extracted using the wetting index, and urban areas (urban area) could be extracted using thermal infrared bands.

2. In order to solve the problem of misclassification of urban images by RGB band, we extract the downtown area using thermal infrared band and compare the classification result with that of the Ministry of Environment to prove the effectiveness of the urban image classification technique by thermal infrared band I could. In addition, farmland (paddy field) could be extracted from the images before and after planting by applying the wetting index by band combining technique.

3. Comparison of Land Cover Classes According to Image Classification [Land-NDVI, Watershed-Near Infrared Band, Rice Field-Wet Index, Urban Area (Urban Area) - Thermal Infrared Band) We analyzed the characteristics of land cover change in Nakdong River basin for 6 ~ 8 years. As a result of analysis, water area, forest, and paddy fields decreased by 2.53%, 0.37% and 4.52%, respectively. Urban areas and fields increased by 3.71% and 6.00%, respectively. Therefore, as the urbanization progressed, the area of the forest and the area of the discussion decreased, and the area of the city area increased greatly.

4. GIS-based RUSLE input factors could be calculated using DEM, precision soil map, and land cover map constructed by image classification. In addition, by analyzing the RUSLE factor for each zone by constructing the Zone Grid, it was possible to effectively analyze subwatershed vulnerable to soil loss in terms of soil, topography, and land cover.

5. Analysis of soil loss by GIS-based RUSLE model In the estimation of soil loss by each station, the largest amount of soil loss was simulated at 163,025 ton / yr in the subcatchment area of Imcheon subwatershed. And it was found that the watershed is the most vulnerable to the soil loss in the event of rainfall.

As described above, the method of the present invention can be implemented as a program and recorded on a recording medium (CD-ROM, RAM, ROM, memory card, hard disk, magneto-optical disk, storage device, etc.) Lt; / RTI >

While the present invention has been described in connection with what is presently considered to be practical exemplary embodiments, it is to be understood that the invention is not limited to the disclosed embodiments, but, on the contrary, It will be understood that various modifications and changes may be made without departing from the scope of the present invention.

A: Dust loss (ton / ha / yr)
R: rainfall erosion factor,
K: soil erosion factor,
L: Length factor of erosion slope,
S: slope factor of erosion slope,
C: vegetation covering factor
P: Cultivation factor

Claims (12)

(a) Applying the image classification technique of each land cover class including the urban area (urban area), agricultural land (rice field, field), forest, road and water area from the latest satellite image (Lansat image) Building; And
(b) By using ArcGIS software, we construct GIS data of soil, digital elevation model (DEM), and land cover map, and by using GIS spatial analysis technique required for RISLE (Revised Universal Soil Loss Equation) Extracting the subwatershed vulnerable to soil erosion in terms of soil, topography, and land cover by analyzing the RUSLE factor by calculating the calculated RUSLE factor and calculating the soil loss amount (A)
Based on the GIS-based method. The method according to claim 1,
In the step (a), the image classification according to the land cover class
From the satellite image, it is classified into forest-NDVI, water-near-infrared band, agricultural land (paddy field) - wet index, urban (urban)
The forest is extracted by applying Normalized Difference Vegetation Index (NDVI) in combination with a near-infrared band and a red band,
In order to consider the influence of the water level change due to rainfall, a water layer is extracted by applying a cell expansion technique to the water (water) by classifying the water using the most sensitive NIR,
The agricultural land (paddy field) is extracted using a wetting index,
Wherein the urban area (city area) is extracted using a thermal infrared band. The method according to claim 1,
The process for evaluating the soil loss using the RUSLE (Revised Universal Soil Loss Equation) model is based on the GIS DB used for evaluating the respective RUSLE factors including the precision soil map, the digital elevation model (DEM) Using a coating,
The RUSLE factors include rainfall erosion factor (R), soil erosion factor (K), erosion slope length factor (L), erosion slope factor (S), vegetation cover factor (C) And the GIS-based soil loss evaluation method. The method of claim 3,
Using the RUSLE model, the soil loss volume (A) has units of ton / ha / yr,

By equation (2)
The soil loss factor was calculated as the product of rainfall erosion factor (R), soil erosion factor (K), length factor of erosion slope (L), slope factor of erosion slope (S), vegetation cover factor (C) And calculating a GIS-based soil loss. The method of claim 3,
The rainfall erosion factor (R) used rainfall data. For this purpose, the WAMIS (Water Management Information System) was searched for monthly rainfall data for the last 10 years for rainfall observatories. To calculate the average annual rainfall

[here,

Wherein the rainfall erosion factor (R) is calculated by applying the equation (2.1) to the annual rainfall (mm / yr). The method of claim 3,
The soil erosion factor (K) was determined by analyzing the content of sand, silt and clay in the topsoil layer from the soil group based on 1 / 25,000 precision soil map and then calculating the Erickson K value Wherein the soil erosion factor K is calculated for each soil level using a triangle chart. The method of claim 3,
The topographical factor LS is divided into a length factor L of the erosion slope and a slope factor S of the erosion slope, and the length factor L of the erosion slope is the horizontal slope of 22.13 m, R, and Desmet & Govers equations are used to calculate the ratio,

,

m is the exponent of the erosion slope length,

Is the inclination angle of the erosion slope,

Renard equation (5)

Is the length factor of the erosion slope for the cell,

Cell size,
This formula uses the number of cells entering the GIS cell-based observation cell and uses the cell length as a multiple-flow algorithm to determine the total slope of the segment, which measures the bifurcation and confluence of the flow and the complex natural topography Considering,
Desmet and Govers (1996) use Equation (6) to calculate the length factor (L) of the erosion slope,

(6)

Is the length factor of the erosion slope for the cell,

Is the upstream contribution area into the cell,

Cell size,

Is a contour line length orthogonal to the flow direction (

)),

Is the direction of the cell, and in this formula, the contour line length perpendicular to the flow direction

(A contour line length model according to the direction of the cell of Fig. 4)
Here, the upstream contribution area A flowing into the observation cell is calculated as shown in equation (7)

(7)
here,

Is the contribution area for observation cell i, A is the available upstream contribution area,

Is the slope of the neighboring cell,

Is a weighting factor (0.5 in the orthogonal direction and 0.354 in the diagonal direction)

Is the number of cells in the downstream direction,
The Desmet and Govers equation (6) uses the number of cells entering the observation cell, which is similar to the Renard equation (5), but is a multiple flow direction model considering the flow direction into the cell. And the resultant value is extracted in units of predetermined cells of the GIS. The method of claim 3,
The slope factor S of the erosion slope represents the influence of the slope inclination against the soil erosion slope and the slope factor S of the erosion slope is more sensitive to the soil erosion than the length factor L of the erosion slope The relationship between slope and soil loss is influenced by vegetation density and soil particle size. In the USLE, which predicted erosion on agricultural land, the agricultural land has a generally uniform slope distribution. In the case of the RUSLE, which predicts erosion of the watershed, considering the fact that the watershed topography change is severe, use the cell-based S value as the introduction of the GIS that can evaluate the S factor locally, or
The slope factor (S) of the erosion slope in the USLE was calculated by using Eq. (8), which was developed by considering the slope effect in the field and river by selecting the study area to solve the problem of excess soil erosion. , &thetas; is the inclination angle of the erosion slope

(8), or
According to Liu et al. (1994), the slope angle increases linearly with the slope angle in the region of 9 ~ 50% slope. Therefore,

(9), or
Applying Equations (8) and (9) to a wide area, we can approximate the existing linear function applied to the slope of less than 25% to approximate the Nearing (1997) equation (10)

(10)
Wherein the GIS-based soil loss evaluation method comprises the steps of: The method of claim 3,
The vegetation cover factor (C) is a GIS-based soil-soil loss (CIS) value using the value of the vegetation cover factor (C) for each land cover classified into city area, agricultural area, forest, grassland, Assessment Methods. The method of claim 3,
The tillering factor (P) used was 0.55 for contour line cultivation, 0.27 for contour line cultivation, and 0.10 for terrace cultivation in the case of 0.0 to 7.0 slope (%) using the cultivating factor (P) , Contour line cultivation 0.60, contour line cultivation 0.30, and terraced cultivation 0.12 were used for the slope of 7.0 ~ 11.3 slope (%), 0.80 for contour line contouring 0.30 for contour line 11.3 ~ 17.6, , Contour line cultivation (0.90), contour line cultivation (0.45) and terrace cultivation (0.18) were used for slope (17.6 ~ 26.8) ) Is used as the GIS-based soil loss evaluation method. 11. The method of claim 10,
The cultivating factor (P) depends on the type of cultivated land in farmland, the type of cultivated land in the field and the inclination of the slope of the cultivated land. The cultivated land types include contouring, contouring and terracing, The same type of cultivated land plays an important role in regulating the soil loss, and the cultivated land forms vary according to the regional characteristics. The field shows the cultivation type of the contour line and the rice field shows the terraced cultivation type. However, As a result of the field survey, it is necessary to identify the type of cultivation that is mainly used in the region,
The above-mentioned contour line cultivation is a cultivation type having a certain direction and inclination, and it can be estimated that the soil loss is the highest in that it causes the soil transfer immediately in the event of rainfall,
The above-mentioned contour line cultivation is similar to the contour line cultivation but is relatively non-erosive rather than the contour line cultivation method in that it is formed along contours,
The terraced tillage method can reduce the soil loss by using a short distance and a sharp inclination angle in the cultivated land in the form of a stair, and the eroded soil is moved to the downstream along the slope. At this time, The erosion is rapidly reduced and the soil is accumulated on the terrace. If the terrace does not exist, the soil that has been transported for several times is collected into a single outflow, causing a disaster such as a landslide. The method according to claim 1, characterized in that the transferred soil is dispersed in intensive kinetic energy gathered in a single outflow port and is guided to several outlets to minimize the sudden disaster of landslides. The method of claim 3,
Wherein the slope is used for evaluating the tillage factor (P), and the distribution characteristic of the gradient is also evaluated from the DEM data at a predetermined resolution.

Images