Examination of Land Degradation based on Erosion Potential using Revised Universal Soil Loss Equation (A Study Case of Bandung Regency, West Java, Indonesia)

Research Title

:

Examination of Land Degradation based on

Erosion Potential using Revised Universal Soil

Loss Equation

(A Study Case of Bandung Regency, West Java,

Indonesia)

Name

:

Yee Yee Maw

Student ID

:

G 051048021

Study Program

:

Master of Science in Information Technology for

Natural Resources Management

Approved by,

Advisory Board

Dr. Ir. Tania June ,M.Sc

Iwan Setiawan ,S.Si,PM

Supervisor

Co-supervisor

Endorsed by,

Program Coordinator

Dean of the Graduate School

Dr. Ir. Tania June ,M.Sc

Dr. Ir. Khairil A. Notodiputro, MS

STATEMENT

I, Yee Yee Maw, here by stated that this thesis entitled:

Examination of Land Degradation Based on Erosion Potential Using Revised

Universal Soil Loss Equation (Case study Bandung Regency, West Java, Indonesia)

Are result of my own work during the period of January to May 2006 and that it has not

been published before. The content of the thesis has been examined by the advising

committee and the external examiner.

Bogor, June 2006

ACKNOWLEDGEMENT

First, I would like to express my deepest gratitude to Buddha the All Mighty for

His Mercy, Favor, Blazing me to carry out this task with sound health.

I wish to express my Earnest thanks and Sincere appreciation to the Government

of the Union of Myanmar,

Minister

and

Deputy Minister for Ministry of Agriculture

and Irrigation, my special Kind and Respected to

Mr. Tin Htut Oo, the

Director

General for Dept. of Agricultural Planning, Myanmar for their official endorsement,

kindly permission, allowed me to attend the Master for Information Technology for

Natural Resources Management Study Program in Bogor Agricultural University (IPB),

Indonesia.

I am very grateful to those who have helped me throughout this research period.

My especially gratitude and appreciation due to

Dr. Ir. Tania June

(Program

Coordinator of MIT) and

Ir. Iwan Setiawan

who acted respectively as my respected

Chairman and Member of Advisory committee. I am deeply appreciate and respected

them for their constant guidance, continuous encouragement, interest, and support during

the work and for providing the study area data throughout the duration of this project. I

really wish there were some way of showing my gratitude for all they have done for me.

I am especially grateful to my sponsor "

ASEAN FOUNDATION

" for financial

support by awarding me the Scholarship during of my two years Master Degree study

which without this support, this Degree would not be possible. I am deeply appreciated.

My special thanks goes to the

Director

,

Deputy Directors

, and

staffs

at the

SEAMEO Regional Center for Tropical Biology (SEAMEO BIOTROP). Also the MIT

staffs for their help and support to me during my study period in Indonesia. I thank very

much to

Drs Eddy Nugroho

and Dr. Budi Kartiwa

(Hydrology Research Lab) who

took their personal time to track down knowledge and provide insight and support

Citarum watershed metrology data. Without their assistance during some trying times, the

results of my work would have been less than desirable.

My special dedication and gratit ude goes to my sister,

Ms. Tin Tin Win

, the

cheerful lady who stream spirit, and provided as accompanied me during my

hard-pressed time. My profound gratitude goes to my friends in MIT 2004 odd and even

semester; 2005 even semester students, especially Mr. Adang Setiawan, Wen Wen and

Tran Xung Sang. I really grateful to appreciate for the advice, assistance and camaraderie

and best wishes to you all.

Finally, I would like to thank my Grandpa, my Parents and my lovely youngest

sister for making all of this possible through their continued support and love. They have

always been there to set me on the right path and for that I am eternally grateful. They are

the reason that I have been able to succeed.

Curriculum Vitae

Yee Yee Maw was born in Yangon, Myanmar at June 6, 1974 as a

second daughter of U Kyaw Hla and Daw Khin Khin Myint. She

receives her undergraduate Degree from Yezin Agricultural

University in 1998 in the field of Soil and Water Management.

From the year of 1998 to present, she works for Department of

Agricultural Planning under the Ministry of Agricultural and Irrigation as a deputy

program office in Trade and Marketing Section.

In the year 2004, Yee Yee Maw received a scholarship from ASEAN

FOUNDATION to peruse her graduate study in Bogor Agricultural University. She

received her Master of Science in Information Technology for Natural Resources

Management from Bogor Agricultural University in 2006. Her thesis was on

“Examination of Land Degradation Based on Erosion Potential by using Revised

Universal Soil Loss Equation, Cased Study of Bandung Regency, West Java, Indonesia”.

ABSTRACT

YEE YEE MAW(2006).

Examination of Land Degradation Based on Erosion

Potential Using Revised Universal Soil Loss Equation (Case study Bandung

Regency, West Java, Indonesia).

Under the supervision of Dr. Ir. TANIA JUNE and

IWAN SETIAWAN.

Soil erosion is a serious environmental problem in the world. With the generally

high erosion rate in many parts of the world, efforts should be directed towards curtailing

its hazard. This requires quantitative data to identify critical areas where urgent

conservation is needed. Traditional approaches based on runoff plots are expensive, time

consuming and generate point-based data. A comparative study was conducted to

estimate soil loss under landscapes, major land uses and slope gradients in Bandung

Regency, West Java Province, Indonesia. The soil erosion model,

i.e

_{. the Revised}

Universal Soil Loss Equation (RUSLE) was applied in GIS and remote sensing (RS)

environment.

GIS based spatial analysis model is used to answer questions about what exists

now or exists at some points in the past. Perhaps most importantly, it can be used to

predict what will happen or has happened in another location or at another point in a

certain time. A GIS provides these capabilities by means of its spatial analysis function.

Erosion that is caused by rain and human’s activities can be assessed with the Revised

Universal Soil Loss Equation in ton/hectare/year units based on soil loss rate of a certain

area. RUSLE involves erosivity from rain, erodibility from soil properties, slope

conditions, land conservation, and cultivation system factors on a land.

The classification of land cover in study area was conductive by using Landsat

satellite image of different years, i.e.1989, 1993, 1998 and 2003. Changes in land use/

cover was led to increase the rate of erosion by water in the Bandung regency which

could be concluded from the statistical analysis of the annual result from RUSLE. The

potential eroded area by districts in the study area can be examined. Human factor, land

use and cover change, is the most dominant factor in RUSLE equation for prediction of

land degradation was shown in the sensitivity analysis of land use/cover change by years.

According to the outcome of this research, deforestation and the agricultural

practice for vegetable farming in the study area were causing problems of land

degradation. Intensification of agriculture area and residential area are major factors of

land degradation in the study area. Finally, this research can be concluded that 1825

hectare of land in the Bandung regency is continuously suffered high and very high

erosion in four different period of land cover type. That area need to be recognized as

“rehabilitant area” and must be maintained with the way of high soil conservation

practice. The forested area which located more than 40 percent slope gradient in the

study area is needed to be identified as highly value conservative forest area. Erosion

must be noticed in a land degradation management and RUSLE can help farmers not only

to assess erosions on their lands but also to policy maker for implementing the land use

planning.

TABLE OF CONTENTS

STATEMENT... I

ACKNOWLEDGEMENT ... II

CURRICULUM VITAE...III

ABSTRACT...IV

TABLE OF CONTENTS... V

LIST OF TABLE...VI

LIST OF FIGURE...VII

REFERENCES

……….88 – 94

APPENDIX ………..

LISTOFFIGURE

FIG(1.1):THE EROSION ASSESSMENT RATE OCCURRENCE IN WEST JAVA AREA

(JAWA BARAT)...6

FIG (2.1):DISTRIBUTION OF RAIN EROSION THROUGHOUT THE WORLD (HUDSON 1995_)...10

FIG (3.1):THE GENERAL INFORMATION OF RUNOFF EROSION MODEL BASED ON SPATIAL ANALYSIS...16

FIG (3.1):THE LOCATION OF STUDY AREA...28

FIG (3.2):THE FLOW CHART OF THE WHOLE PROCEDURE OF STUDY...31

FIG (3.3):THE PROCESS OF DRIVIN G DIGITAL TERRAIN MODEL (DEM) FROM CONTOUR LINE MAP...32

FIG (3.4):THE FLOW CHART OF PREDICTION EROSION PROCESS BY RUSLE MODEL...36

FIG (3.5):THE PROCESS OF MANIPULATE LS FACTOR FROM DEM ...40

Fig (4.1): Flow Chart of Image Preprocessing ...44

Fig (4.2): The geometric correction image of 1989 and 1993 ...45

Fig (4.3): Images after geometric rectification of the years 1989 and 1993 ...46

Fig (4.4): The scatter gram layer of each class of classified image ...48

Fig (4.5): The histogram shown the yearly percentage of land use changed...64

Fig (4.6): The correlation between land cover and erosion in 2003 ...66

Fig (4.7): The total soil lo ss area in (km square) in each district in lower potion ...67

Fig (4.8): The soil loss rate in each district in lower potion...68

Fig (4.10): The soil loss rate by erosion in middle portion...69

Fig (4.11): The total soil loss area in (km square) due to erosion ...69

Fig (4.12): The total soil loss area in (km square) by erosion of each district ...70

Fig (4.13): The effectiveness of different type of land cover upon soil loss rate ...71

Fig (4.14): The soil loss in area amount in year 1998...71

Fig (4.15): The soil loss rate of Lower portion in year 1998...72

Fig (4.16): The soil loss in area amount based on year 1998 land cover type ...73

Fig (4.17): The soil loss in rate based on year 1998 land cover type...73

Fig (4.18): The soil loss in area amount in the year 1998 land cover type ...74

Fig (4.19): The soil loss rate in the year 1998 land cover type ...74

Fig (4.20): The land cover (C) factor effectiveness on erosion rate in 1993 ...75

Fig (4.21): The soil loss area by erosion in lower portion of study area ...76

Fig (4.22): Soil loss rate in lower portion of study area...76

Fig (4.23): Soil loss rate in each district in year 1993...77

Fig (4.24): Soil loss rate in each district in year 1993...78

Fig (4.25): The soil loss rate in area by erosion in lower portion of study area...78

Fig (4.26): Soil loss rate amount in study area at 1993...79

Fig (4.27): The correlation of soil loss rate with Land used type (LUT) in 1989...79

TABLE (2.1): Data Layer And GIS Description For Rusle Factor ...20

Table (4.1): Name and History of Satellite Imagery for doing analysis ...43

Table (4.2): The Accuracy assessment of each class of land cover classification (2003) ...49

Table (4.3): The Physical Characteristics of the Soil Types in the Study Area...52

Table (4.4): The Classification of Soil Group in Hydrology based on Texture...54

Table (4.5): The Yearly Erosion Rate of Study Area by using RUSLE...65

LIST OF APPENDIX

APPENDIX-1: THE LOCATION OF WATER LEVEL GAUGE STATION IN

UPPER CITARUM WATERS HED ...1

APPENDIX-2: RAINFALL EROSIVITY (R) FACTOR FOR BANDUNG

REGENCY...2

APPENDIX-3: SOIL MONOGRAPH...3

APPENDIX-4 :THE SOIL ERODIBILITY(K FACTOR) OF BANDUNG

REGENCY...4

APPENDIX-5: SLOPE LENGTH (LS) FACTOR MAP...5

APPENDIX-6: LAND COVER CLASSIFICATION OF BANDUNG

REGENCY IN 2003 ...6

APPENDIX-7: LAND COVER CLASSIFICATION OF BANDUNG

REGENCY IN 1998 ...7

APPENDIX-8: LAND COVER CLASSIFICATION MAP OF 1993...8

APPENDIX-9: LAND COVER CLASSIFICATION MAP OF 1989...9

APPENDIX-10: MAP LAYOUT OF EROSION POTENTIAL AREA (2003) ..10

APPENDIX-11: MAP LAYOUT OF ER OSION POTENTIAL AREA (1998) ..11

APPENDIX-12: MAP LAYOUT OF ER OSION POTENTIAL AREA (1993) ..12

APPENDIX-14: THE COMPARISON OF EROSION IN BANDUNG

REGENCY BASED ON FOUR DIFFERENT LAND COVER SCENARIO ...14

APPENDIX-15: LAND DEGRADATION MANAGEMENT MAP ...15

APPENDIX-16: FACTOR C CLASSIFICATION VALUE...16

APPENDIX-17: P - FACTOR FOR SEVERAL CONSERVATION

PRACTICES ...17

APPENDIX-18: LIST OF THE NAME OF DISTRICT AND ITS

I. INTRODUCTION

1.1. Background

Erosion is the removal of surface material from the Earth's crust, primarily soil and rock debris, and the transportation of the eroded materials by natural agencies. It’s most important agent is moving water (Sposito, 1999). The concept of multiple soil functions and competition is crucial in understanding current soil protection problems and their multiple impacts on the environment. Soil erosion, in particular, is regarded as one of the major and most widespread forms of land degradation, and, as such, poses severe limitations to sustainable land use.

Soil erosion is a detrimental process both on-site and off-site. Soil erosion not only reduces soil depth, but also reduces the capacity of soils to hold water due to sealing, and depletes plant nutrients in the soil. This reduces soil productivity and causes long term reduction in crop yields (Nanna, 1996), since the necessary plant nutrients are washed away. It is estimated that annual crop production becomes uneconomical on 20 million hectares of land in the world (Elirehema, 2001). This raises concern about the ability of land to feed the ever-increasing population. Moreover, soil erosion also creates off site environmental problems, such as water pollution, silitation of reservoirs and degradation of coastal ecosystems. It is thus necessary to understand where erosion is taking place in order to design sound conservation measures (Kadupitiya, 2002a).

Soil losses due to erosion can be considered as irreversible in relation to the time needed for soil to form or regenerate itself. Natural erosion rate ranges

from less than 0.02 to more than 10 metric tons per hectare of soil lost annually, rates of soil loss exceeding 10 metric tons per hectare annually occurrences of accelerated erosion. It is important to note that this accelerated soil loss is equivalent to less than 1 mm of soil depth, making erosion damage very difficult to observe over short time spans (Sposito, 1999). Erosion is extremely costly for developing countries. Besides the damage to infrastructure, fisheries, and property, erosion of precious top soils costs tens of billions of dollars worldwide each year.

Vegetation cover is a very crucial factor in reducing soil loss (Petter, 1992). In general, as the protective canopy of land cover increases, the erosion hazard decreases (Mkhonta, 2000). It protects the soil against the action of falling rain-drops, increases the degree of infiltration of water into the soil, maintains the roughness of the soil surface, reduces the speed of the surface runoff, binds the soil mechanically, diminishes micro-climatic fluctuations in the uppermost layers of the soil, and improves the physical, chemical and biological properties of the soil (Petter, 1992). As long as vegetation cover is unbroken, erosion and runoff are small despite erosivity of the rainfall, slope steepness and soil instability. The effects of vegetation cover on erosion processes especially on surface erosion are varied depending on the type of vegetation cover, density, undergrowth cover and litter. These determine the interception loss, absorption of kinetic energy and increasing water infiltration. Land with good cover allows soil redundancy to overland flow. Vegetation acts as a protective layer or buffer between the atmosphere and the soil. The above ground cover absorbs energy of falling raindrops, running water and wind, so that less is directed to the soil. The below

ground components comprising the root system contribute to the mechanical strength of the soil (Hagos, 1998).

Controlling erosion requires data on relative erosion rates, spatial extents, vulnerable areas, current sources, relative contributions from different sources and likely effects on land use (Meijerink and Lieshout, 1996). In many areas, quantitative data on erosion rates is severely lacking (Nadeem, 1999). This data is necessary for land management decisions in assigning priorities for erosion control (Jack, 2002; Moore and Burch, 1986). It is financially impractical to have conservation in all areas, rendering the need to identify and prioritize critical areas (Wessels et al., 2001). Such information enables prevention of various forms of degradation before they caused irreparable damage (Wessels et al., 2001).

Soil erosion modeling has proved to be a sound approach in generating quantitative data (Shigeo et al., 1998). Models are effective predictive tools of soil loss (Nearing et al., 1994). Models are particularly useful for evaluating the impacts of intensified land use on soil loss, water quality and for evaluating the potential effectiveness of mitigation or remedial measures before large sums of money are invested in such measures (Moore and Burch, 1986).

Several studies (Shrestha, 1997; Shrestha, 2000; Wessels et al., 2001) have shown that GIS is an excellent tool in erosion modeling. GIS modeling does not only predict consequences of human actions on erosion, but it is also useful in the conceptualization and interpretation of complex systems as it allows decision-makers to easily view different scenarios. Most of the data used in models i.e. vegetation type, soil, relief, climatic, etc can be processed in a GIS and used as

first stage input to identify and map of degraded lands (Jaroslav et al., 1996; Shigeo et al., 1998).

1.2. Identification

The island of Java is one of the most densely populated areas in the world. A chain of volcanoes, some still active, have enriched the soil so that it is generally very fertile. Since ancient times, Java has been a center of educational, economic, cultural and political activity in Indonesia. Java has been until recently the main producer of rice and sugar in the country. The fertile plain in the north of the island is intensively cultivated throughout the year. To the south, fertile agricultural lands have been developed in several river basins. If these are farmed without conservation measures, the result is erosion and increased flooding.

However, because of the growing population has already occupied all available land in the plains, people have invaded their agricultural purpose to hilly and mountainous regions. For this reason, many reservoirs have been constructed to manage water resources for irrigation and sanitation, as well as hydroelectric power in west Java. About 60% of the population of Indonesia lives in Java and Madura, and almost all of the land are being utilized for agriculture (75%). With a population growth of 1.7% annually, the number of families that depend upon agriculture will be increased by about 150,000 each year. This will result in the conversion of forest land at a rate of 18,000 ha annually, with about 40,000 ha of agricultural land being converted each year into residential and industrial purposes in west Java (World Bank 1990).

Because of the high population density, much of the land on west Java (Java Barat), Indonesia that should have been preserved for forest and

conservation purposes has already been used for urbanization and agriculture. Now only just over 6 million (hectare) of Java Island are covered with forest. Because of population growth, these areas will undoubtedly continue to diminish. The land currently within the forestry department boundary in west Java is about 22% or about 800,000 hectares less than the recommended area. Furthermore, many of the designated forest lands are not in fact forested in the late 1980's, the island of Java especially west Java ( Java Barat) was losing 770 million metric tons of topsoil every year at an estimated cost of 1.5 million tons of rice, enough to fulfill the needs of 11.5-15 million people. Only the West Java area (Jawa Barat), the erosion rate per year is around 32,931.061 ton/th (1 juta truk tronton=

30 ton). (Source: BAPEDA 2002, DATA 2001/ RTRW Province Jawa Barat 2010,

Agro-ecological Analysis for Agricultural Development in Indonesia).

There are 13 watersheds in Java have which have critical erosion problems. The Citarum watershed which is most important for electricity and drinking water for Java Island is situated in west Java. The upper Citarum watershed is now facing many environmental problems caused by the relatively rapid population growth demanding change and development of new settlement areas. The density of the population is relatively high: 1,640 people per km2.

There has been a 7-fold increase in silt load in the Citarum River over a recent 3-year period. This is rapidly filling Indonesia's largest reservoir which located downstream at Jatiluhur. For the whole Java, "Critical" lands outside of covering forests include 5690 km2. Soil erosion rates are 990 to 4040 t/ km2/ year and are increasing. Indonesia's main island of Java must be one of the most eroded places in Asia, if it is not the most in the world. The government of Indonesia

classifies more than 10,000 km2 (8% of croplands) as critically eroded. The land is said to be so badly degraded that it is already, or soon will be, unable to sustain even subsistence agriculture. Some small fields are losing 5 cm of soil/ year (150,000 t/ km2/ year).(Source: The Earth's Carrying Capacity --Some Literature Reviews, http://home.alltel.net/ bsundquist1/index.html available: 10 pm, 14 January, 2006).

Fig(1.1): The erosion assessment rate occurrence in west Java area (Source: http://home.alltel.net/ bsundquist1/index.html)

The Bandung basin which is the main portion of upper Citarum watershed seriously faces routine annual floods caused by an increase of erosion, sedimentation and pollution problem. So that area needs to maintain environmental condition for future uses.

This study will attempt to evaluate the impacts of different land use cover types on soil loss by employing GIS based modeling techniques in order to identify the effects of land degradation on its generation.

Very light (0-15 Ton/ Ha/ yr) Light (15-60 Ton/ Ha/ yr) Intermediate (60-180 Ton/ Ha/ yr) Critical (180-480 Ton/ Ha/ yr) Very Critical (>480 Ton/ Ha/ yr)

1.3. Objective

The objectives of this study are as follows:

• Estimation of soil erosion potential based on soil loss condition in the study area.

• Examination of land cover change effect on soil erosion.

Output

• Soil erosion potential area using RUSLE.

• Correlation between land cover changed with potential soil erosion.

Benefit

• Provide the snapshot view of erosion potential assessment for decision maker to implement the land use planning for declining land degradation

1.4. Thesis Structure

The thesis is divided into 5 chapters. The first chapter identifies the research background, scope and objectives emphasizing on the need for this study. Chapter two discusses the background on the theoretical aspects of erosion with highlights on factors that influence it focusing on soils degradation problems. Various methods of erosion modeling are briefly specified in this chapter and finally a short description of the approaches employed in this study is given. In chapter 3, a description of the study area is given mentioning among others its location, climate, soil characteristics and geology. Moreover, Chapter 3 introduces the modeling approaches implemented in this study. It gives detailed descriptions on the components used and specifies the data requirements of each approach. Following this the activities involved in attaining the required data for the modeling procedures are discussed. Chapter 4 discusses the procedures

used in preparing the input maps and the parameterization of the parameters required is mentioned. The results of the study are presented in chapter 4. This includes the results on the analysis of soil loss by erosion based on RUSLE and the model predictions relating them to different land use/ cover types. Finally in chapter 5, conclusions on the results of the study are made. Limitations of the study are highlighted and some recommendations based on the study are given in chapter 5.

II.

LITERATURE REVIEW

2.1. Erosion

Soil erosion is one form of land degradation besides soil compaction, low organic matter content, loss of soil structure, poor internal drainage, stalinization and soil acidity problems. In particular, soil erosion is defined as: “physical removal of topsoil by various agents, including falling raindrops, water flowing over and through the soil profile, wind velocity and gravitational pull” (Lal 1990). Historically, soil erosion began with the beginning of intensive agriculture activities, where people are removing protective vegetation cover and growing various food crops on disturbed soil surface.

In addition, some other large-scale opening of vegetation through commercial logging, preparation of timber and crop estates, and expansion of human settlement accelerated it. Nowadays, soil erosion is almost universally recognized as serious threat to human’s well being. This is confirmed by facts of active supports given by most governments to soil conservation programmes (Hudson 1995). A general figure of rain erosion susceptibility is presented in Figure (2.1).

Fig (2.1): Distribution of rain erosion throughout the world (Hudson 1995) Soil erosion caused by water is a serious problem in sub humid, semiarid, and arid regions. Inadequate moisture and periodic droughts reduce the periods when growing plants provide good soil cover and limit the quantities of plant residue produced. Erosive rainstorms are not uncommon and they are usually concentrated within the season- when cropland is least protected (Wischmeier and Smith, 1978).

This energy in the form of rainfall causes splash erosion. The potential energy for erosion is converted into kinetic energy, the energy of motion of the running water. This kind of energy formed by runoff causes inters rill, rill, gully, and riverbank erosion.

2.2. Factors Affecting Soil Erosion

Agents of erosion are the carriers or the transport system in the movement of soil (e.g. water, wind). Factors of erosion are those natural or artificial parameters that determine the magnitude of perturbation, e.g. climate, topography, soil, vegetation and management. Erosion may not occur even when the agents

as farming practices, deforestation and cropping systems that facilitate the effects of agents and factors of erosion and accelerate the various erosion processes (Bergsma 1996; Lal 1990). The factors affecting soil erosion by water are:

2.2.1. Climatic Erosivity

Erosivity refers to the aggressively of the climate, or more precisely the energy of such climatic elements to cause erosion. Climatic factors that affect erosivity are precipitation, wind velocity, water balance, mean annual and seasonal temperatures, etc.

2.2.2. Soil Erodibility

Erodibility is the susceptibility of soil to erosion. This is an inherent property of the soil and is influenced by soil characteristics (e.g.) texture, structure, permeability, organic matter content, clay minerals and contents of iron and aluminum oxides.

2.2.3. Landforms

Erosion also affected the terrain relief through degree and length of slope, shape of slope and slope aspect. In general, the higher the slope gradient, the more soil erosion by water occurs.

2.2.4. Human

Human activities affect soil erosion through their measures to natural resources. Human activities related to erosion are deforestation, grazing, faulty farming system and cropping intensity. However, some activities in terms of soil conservation measures are reducing the amount of soil erosion (e.g. contouring planting)

2.3. Land degradation and land use/ land cover change

Land use and land cover change have become a central component in current strategies for managing natural resources and monitoring environmental change. Since the late 1960’s, the rapid development of the concept of vegetation mapping has lead to increased studies of land use and land cover change worldwide. Providing an accurate assessment of the extent and health of the world’s forest, grassland, and agricultural resources has become an important priority.

2.3.1. Land use and land cover

Every parcel of land on the Earth’s surface is unique in the cover it possesses. Land use and land cover are distinct yet closely linked characteristics of the Earth’s surface. Land use is the manner in which human beings employ the land and its resources. Examples of land use include agriculture, urban development, grazing, logging, and mining. In contrast, land cover describes the physical state of the land surface. Land cover categories include cropland, forests, wetlands, pasture, roads, and urban areas. The term land cover originally referred to the kind and state of vegetation, such as forest or grass cover, but it has broadened in subsequent usage to include human structures such as buildings or pavement and other aspects of the natural environment, such as soil type, biodiversity, and surface and groundwater. (Myers, 1988)

Land use change is generally conscious, volitional responses by humans or human societies to changes in biophysical or societal conditions. It is a response indicator, therefore, reflecting how and to what extent society is responding to meet its changing needs and goals or to adapt to changing environmental

conditions. This does not exclude the possibility that some land use changes may, in turn, constitute a driving force for changes in the state of the environment. That is in the very nature of the complex causal network (not a simple causal chain), including a number of feedback loops, that is society's relationship with its environment.

As is the case for land use change, it is doubtful whether a single or aggregate measure of land condition change would be feasible. What is feasible in principle is an estimation of the change in the different land qualities that influence the suitability of the land for one use or another, or for conservation purpose, for example, of biodiversity and erosion for land degradation. (Land qualities are discussed in FAO, 1976).

2.4. GIS and RS in Soil Erosion Modeling and Land Use/Cover Change

Soil erosion is spatial phenomena, thus geo-information techniques play an important role in erosion modeling (Yazidhi, 2003). While this is agreeable, the quality of the results matches the quality of the input data used (Svorin, 2003). Land use data required to run erosion model can be derived from remotely sensed data. In a GIS environment it is possible to link data generated from remote sensing with their spatial location (Mkhonta, 2000). In general, the use of geo-information techniques offers the following advantages in erosion modeling:

(i) fast and cost effective estimates, (ii) possibilities to investigate larger areas,

(iii) greater possibilities of continuous monitoring of these areas and (iv) possibilities to refine the soil erosion model depending on the required

According to Yazidhi (2003), the use of digital elevation models and GIS offers possibilities to estimate more relevant topographical parameters that are useful in soil erosion modeling.

2.4.1. Land Cover Mapping

Land cover mapping is one of the most important and typical applications of remote sensing data. Land cover corresponding to the physical condition of ground surface, for example, forest, glass land etc., while land use reflects human activities such as the use of the land, for example, industrial zone, residential zone, agricultural fields etc.

To prepare, the land cover mapping from digital images “land cover classification” should be done. There are two kinds of classification, i.e. supervised and unsupervised classification.

2.4.2. Supervised Classification

Supervised classification is the method used to transform multi spectral image data into thematic information classes. This procedure typically assumes that imagery of a specific geographic is gathered in multiple regions of the electromagnetic spectrum.

In supervised classification, the identifying and location of feature classes or cover types (urban, forest, water, etc) are known beforehand through fieldwork, analysis of aerial photographs, or other means. Typically, identify specific areas on the multispectral imagery that represent the desire known feature types, and use the spectral characteristics of theses known areas to train the classification program to assign each pixel in the image to one of these classes. Multivariate statistical parameters such as means, standard deviation, and correlation matrices

are calculated for each training region, and each pixel is evaluated and assigned to the class to which it has the most likelihood of being a member (according to rules of the classification method chosen).

One of the sample classification strategies that may be used is Maximum Livelihood Classifier. The maximum livelihood was adopted by using the training samples of the landsat image and ground truth. Actually, this is one of the most popular methods of classification in remote sensing, in which a pixel with the maximum likelihood is classified into corresponding class.

2.5. Runoff Erosion Potential with GIS Based Spatial Analysis Model

Spatial Analysis extends the basic set of discrete map features of points, lines and polygons to surfaces that represent continuous geographic space as a set of contiguous grid cells. The consistency of this grid-based structuring provides a wealth of new analytical tools for characterizing “contextual spatial relationships”, such as effective distance, optimal paths, visual connectivity and micro-terrain analysis. In addition, it provides a mathematical/statistical framework by numerically representing geographic space.

Spatial Statistics, on the other hand, extends traditional statistics on two fronts. First, it seeks to map the variation in a data set to show where unusual responses occur, instead of focusing on a single typical response. Secondly, it can uncover “numerical spatial relationships” within and among mapped data layers. The model assumes that erosion potential is primarily a function of terrain steepness and water flow. Then the result is combined with the human factor. Admittedly the model is simplistic but serves as a good starting point for a spatial analysis example in natural resources.

Fig (3.1): The general information of runoff erosion model based on spatial analysis

2.6. Hybrid Erosion Modeling Approach

Empirical models which is one of statistical model, describe erosion using statistically significant relationships between assumed important variables where a reasonable database exists (Kadupitiya, 2002a). Empirical models are based on

defining important factors through field observation, measurement,

experimentation and statistical techniques relating erosion factors to soil loss (Petter, 1992). In empirical models, the inherent processes involved are not used and the models can only be operated in the designed direction where inputs go into one side of the equation and the out put on the other side. Empirical models are quick in predicting erosion, but are site specific and require long-term data (Elirehema, 2001). Most models used in soil erosion studies are empirical models. The most widely used empirical model is the Universal Soil Lose Equation (USLE). Others include Derived Universal Soil Loss Equation (DUSLE), Revised

Universal Soil Loss Equation (RUSLE) and Modified Universal Soil Loss Equation (MUSLE) etc, which are based on modifications made on USLE. The GIS can also be used as a controlling tool for application ranges of model parameters, especially the digital elevation model (DEM) related topographic variables used in erosion modeling.

2.6.1. Revised Universal Soil Loss Equation: RUSLE

The Revised Universal Soil Loss Equation is being developed by the

USDA's Agricultural Research Service. The model will be refined and improve the accuracy of the original Universal Soil Loss Equation (USLE) to estimate the effects of various conservation systems on soil erosion.

(i) Model and Components

The RUSLE is an empirical equation that predicts annual erosion (tons/ha/yr) resulting from sheet and rill erosion in croplands. The USLE is factor-based, which means that a series of factors, each quantifying one or more processes and their interactions, are combined to yield an overall estimate of soil loss. It is the official tool used for conservation planning in the US and many other countries have also adapted the equation. The equation is:

A = R * K * L * S* C* P

where,

A = Annual soil loss (tons/hectare/yr) resulting from sheet and rill erosion. This is the predicted value resulting from the execution of the equation above.

R= Rainfall - runoff erosivity factor.

This factor measures the effect of rainfall on erosion. The R factor is a summation of the various properties of rainfall including intensity, duration, size etc. It is computed using the rainfall energy and the maximum 30 minutes intensity (EI30) K= Soil erodibility factor.

The soil erodibility factor measures the resistance of the soil to detachment and transportation by raindrop impact and surface runoff. Soil erodibility is a function of the inherent soil properties, including organic matter content, particle size, permeability, etc. Because these properties vary within a given soil, erodibility (K values) also varies.

L= Slope length factor.

This factor accounts for the effects of slope length on the rate of erosion.

S = Slope steepness factor.

This factor accounts for the effects of slope angle on erosion rates. All things being equal, higher slope values have greater erosion rates.

C = Cover management factor.

Accounts for the influence of soil and cover management, such as tillage practices, cropping types, crop rotation, fallow, etc..., on soil erosion rates.

Accounts for the influence of support practices such as contouring, strip cropping, terracing, etc...

Once these factors have been determined for a field of interest A can be computed. Also, the equation can be used to determine the desired cover management factor (C) or erosion control (P) if the allowable soil erosion rates are known. Thus, in this research use the RUSLE to simulate the impact of changes in land use and land cover on soil erosion, anthropogenic impacts on the environment.

Table (2.1): Data Layer and GIS Description for RUSLE Factor

Erosion databases,

factors

Data layers Description of GIS procedures (include cross-references)

Erosivity (R) Rainfall data

Spatial interpolation of station EI values; stored as R- factor map.

Erodibility

(K) Soil data

Assignment of numerical K-values to soil units by reclassification of the soil unit polygon map with the K- value column from the soil attribute table; stored as K-factor map. Combining slope length and gradient (LS) Geomorphic

If regional geomorphologic relief classes exist, combined LS-gradient or terrain factor values can be obtained using a 2-Dim table with row wise, relief steepness classes and column wise the slope length classes, resulting directly in LS-value distribution for the area; stored as an LS-map file.

Land Cover Land cover, Farm dbs

If necessary pre-processing or spectral classification of remote sensing data; assignment of C-factor values to land cover RS data classes using cover attribute table; stored as C- factor map;

Conservation Practice

Farm

Land cover

For land use types with soil conservation practices, reclassify C-factor map with P-factor values of land cover attribute table, using P=1 for non-conservation areas.

2.6.2. Difference between RUSLE and USLE

The USLE (Wishmeier and Smith, 1978) is the most widely used model in predicting soil erosion. It is used in education and research as a starting point in developing understanding of erosion hazard prediction because of its simplicity and clarity (Hagos, 1998). Many scientists have proposed changes, but all are woven around the same concept of rainfall erosivity, soil erodibility, slope length, slope class, land cover and land management factors are taken as directly proportional to the rate of annual soil erosion (Sohan and Lal, 2001).

RUSLE is a revised version of USLE, intended to provide more accurate estimates of erosion (Renard et al., 1994). It contains the same factors as USLE, but all equations used to obtain factor values have been revised. It updates the content and incorporates new material that has been available informally or from scattered research reports and professional journals. The major revisions occur in the C, P, and LS factors. The cover factor (C) and management factor (P) in RUSLE consider not only agricultural land but also multifunctional land use type and management. The slope length and aspect gradient factor combine to become slope length and steepness factor (LS) in RUSLE.

2.7. High Conservation Value Forest (HCVF)

One of the Forest Stewardship Council (FSC) principles is the management of High Conservation Value Forest (HCVF). This is relatively a new principle, which has been developed to replace the previously used concept of old growth or virgin forest. Through this principle, FSC requires unique approach in managing forest ecosystem and conserving the biodiversity value (FSC 2001). In this case, forest management unit should maintain or enhance such value, not

preserve it. The key of HCVF principle is the concept of conservation values. HCVF have nine principles.

The use of remote sensing and spatial information to support identification of HCVF is certainly potential. Some of the HCVF elements could be assessed through the remote sensing and GIS analysis resulting the location of forest area containing some High Conservation Values (HCV). One HCV element, which is potentially assessed by the support of remote sensing and GIS is HCVF principle four. The principle four (HCVF 4) mention that “Forest areas which provide basic services of nature in critical situations (e.g. watershed protection, erosion control)”. That principle including (3) sub – factors, i.e.

• Functions as unique source of drinking water for local communities (HCV 4.1) • Part of critical major catchments (HCV 4.2)

• Has critical erosion risk (HCV 4.3)

Ancillary data that can be used are topographic information and its derived products (DEM, slope map and other terrain features to support prediction of potential soil erosion risk (after Rainforest Alliance and ProForest 2003).

2.8. Accuracy of Modeling in GIS Environment

Accuracy is the degree of likelihood that the information provided is correct. This definition focuses on two components of accuracy. The first and more familiar aspect of accuracy is that it predicts the proportion of information that is expected to be correct or the magnitude of error to be expected. The second and often ignored aspect of accuracy is that it involves a probability. When a map or other data set is asserted to be 80% accurate it means that when the data set is used, it can be expected that on average 80% of the information will be correct.

The measure of this probability of having a higher or lower accuracy than expected is termed the level of confidence. So, when a map is rated 80% accurate with a 90% level of confidence it means that if a large number of accuracy tests were done on the map, then 80% or more of the test points would be correct in 9 out of every 10 tests. The level of accuracy depends on the information to be provided and the level of detail required. An acceptable level of accuracy is that level where the costs of making the wrong decision are equal to the costs of acquiring more accurate information.

In the GIS environment, map accuracy depends on many factors. At the micro level, there are components such as positional accuracy, attribute accuracy, logical consistency, and resolution. At the macro level, there are components such as completeness, time, and lineage. Finally usage components are accessibility and direct or indirect costs. There are also different sources of errors associated with all geographic information. Some of the more common errors are related to data collection, data input, data storage, data manipulation, data output, and the way of using and understanding results.

Paper data such as different maps and associated geographic attributes and data are used as one of the sources of input data to the GIS environment. In this process the paper data are converted to digital data. The level of accuracy of the digital data will be the same as paper data if they are correctly converted to the digital form with a suitable package in an acceptable resolution. Once the data are converted, the accuracy of the output data resulting from different manipulations depends on the resolution power of the software done with the skill of the operator.

2.9. Validation of soil erosion models

Model validation involves a procedure to determine how best the model predicts soil loss rates in the real world. Traditionally validation of soil erosion models has been implemented through the comparison of model output and measured values from real world data (Nearing, 1998).

III. RESEARCH METHODOLOGY

3.1. Physical Condition of the Study Area

3.1.1. Location

The Bandung regency which include major portion in the catchments area of the Saguling Reservoir is so called Bandung basin (Indonesia), with an area approximately 2,283 square km2, geographically located between 6° 4' S to 7° 10' S and 107° 15' E to 107° 45' E. The Bandung regency include (40) districts. For doing sensitivity analysis, the study area need to make zone into three portions, i.e. upper, middle and lower portion. Upper portion of study area consist of (12) districts and middle and lower portion of study area consist of (14) sub-districts. The detail list of sub-district which include in each portion was mention in Appendix (18).

The topography of the Bandung basin in Bandung regency various from flat to mountainous with a height from 650 meter to 2,000 meter above sea level. And also the slope varies from gentle to very steep.

3.1.2. Climate

The study area has tropical climate under the influence of monsoon wind as the same as other places in Indonesia. There are two seasons in this area, rainy and dry season. The rainy season occurs from October until April, whereas the dry season is taken place from June till October. The period from May till June can be considered as transition period. Average rainfall in the surrounding area is between 1,782 millimeter to 3,426 millimeter, for 30 years under consideration, and the average temperature of 220° C to 240° C and average humidity 83%.

3.1.3. Geology and Geomorphology

The Land Mapping Units comprised: (1) flood plains, (2) alluvio-lacustrine plains, (3) colluvial plains, (4) volcanic plains, (5) alluvio-volcanic fan, (6) volcanic fans, (7) volcanic foot-slopes, (8) lower volcanic ridges, (9) middle volcanic ridges, (10) upper volcanic ridges, (11) hills, and (12) mountains.

In general the Lithology of the upper portion of the study area which is the location of Cikapundung catchment, consists of volcanic rock which has been resulted from eruption of Tangkuban Perahu mountain.

According to Silitonga (1973), the study area consists of four units as fallows:

1. The differentiated rock unit coming from the old sunda volcano comprises volcano breccia, lahar and lava alternately.

2. The differentiated rock coming from the Tangkuban Perahu eruptions, comprises sandy tuff, lapilly and breccia, lava and conglomerate.

3. Tuff pumice which came from Tangkuban Perahu eruption phase A, comprises sandy tuff, lapilly, bombs, halloured lava, solid andest and fractions of pumice

4. Lava which consist of basalt and gastubes, coming from the lava flows of the Tangkuban Perahu.

The catchments can be divided into two main geomorphologic units, which are volcanic origin and structural origin. Volcanic origin can be divided into four sub units, which are volcanic cone, volcanic middle slope, volcanic lower slope and fluvial volcanic plain. While the structural origin unit can be divided into two

parts; northern and western part, separated by an east-west running normal fault (Lembang fault).

Indonesia Map

Fig (3.1): The location of study area

3.1.4. Time

The study will be conducted from February 2006 up to May 2006.

3.2. Materials and Tools Requirement

The following materials are needed:

(1) Digital Topography maps at scale of 1:25,000 (2) Soil classification map at scale 1:250,000 (3) Soil survey data

(4) Satellite image: Landsat TM+ for four different years (1989, 1993, 1998 and 2003)

(5) Time series rainfall data (6) Rainfall isohyetal map

3.2.1. Data Source

The materials that were used in this project divided into two types, which are: primary data or field data and secondary data. The secondary data include soil survey data, (10) years rainfall data and discharge data of at least two rainfall stations in the study area. The spatial data raster format are soil type classification map (1:25,000 scale), soil depth classification, slope length and isohyetal map of the study area. The primary data consist of landset TM image and topographic map. The rainfall intensity data and discharge data are provided by Agro-climate and Hydrology research center, Ciomas, Bogor.

3.2.2. Tools

Hardware required minimum computer device: Intel Based PC or compatible

machine with Pentium IV processor.

Software required: ER Mapper 6.4 for Landsat TM image processing, ArcView

GIS 3.3 and Spatial Analyst Extension, Digital Elevation Model Extension, Soil and water assessment tool extension and Hydrology modeling extension is recommended for analysis the data, Microsoft excel 2003 for meteorological data processing and Microsoft words 2003 for preparing the report. Moreover, MapInfo (version 7) is highly recommended for digitizing the analog map.

3.3. Procedure of application

In this study, an estimation of land degradation in the study area was carried out based on the flowchart of the procedure to obtain the purpose of the study. Literature Review Problem identification Research objective Data collection Primary Data Secondary data Land use and vegetative cover data Soil profile description Soil erodibility factor (K) Land cover management factor (C) Soil conservation factor (P) Climate data Topographic map Rainfall runoff erosivity (R) Rainfall data Contour line map Elevation (DEM)

Slope length & Slope steepness

factor (LS)

Preparation of map layers

Calculate erosion potential parameter

RUSLE erosion model

Land Degradation based on erosion

potential

Soil Survey

Data

1989,1993 Landsat TM+, 1998 and 2003 Landsat ETM+, Using Supervised and Unsupervised Classification for testing land use and cover change

Statistical analysis of land use and cover

change Given weight based on land used / cover

characteristic

3.4. Deriving Digital Elevation Model (DEM)

Digital Elevation Model (DEM) is an essential intermediate product that was derived in this study. DEM is used as main input in calculating slope length and slope steepness for prediction of potential erosion risk. In this study, a Digital Elevation Model was derived from contour line map with original scale of 1:25.000 and contour interval 12.5 meter. The general process of deriving the DEM is presented in figure (3.3). The contour line map was digitized and stored in format of ESRI PC Arc/Info® line coverage. Since the elevation interpolation algorithm requires point map as the input, the contour line map was converted into grid (which have same grid cell size with desired DEM grid cell) and then converted into point coverage. After the conversion was done, spline point interpolation of ArcView® was carried out to produce elevation grid with resolution of 30 metres. Nevertheless, the grid size of 30 meter was chosen to have more detailed result and to have the same grid cell size with other spatial data (i.e. land cover type map).

3.5. Classified the Digital Image

Classification of remotely sensed data used to assign corresponding level with respect to group with homogenous characteristics with the aim of discriminating multiple objects from each other within the image.

Classification will be executed on the based of spectral or spectrally defined features, such as density, textured in the future space. Its can be said that classification divides the features space into several classes based on the decision rule. Classification will be done according to fallowing procedure:

Step1: Definition of classification classes

Depending on the objective and the characteristics of image data, the classification classes should be clearly defined.

Step2: Selection of feature

Features to discriminate selected classes should be established using multi-spectral characteristics or features.

Step3: Sampling of training data

Training data should be sampled in order to determined appropriate decision rules. Classification technique such as supervised or unsupervised learning will then be selected on the basis of the training datasets.

Step4: Estimation of universal data sets

Various classification techniques will be compared with the training data, so that appropriate decision rule is selected for subsequent classification.

Step5: Classification

Depending up on the decision rule, all pixels are classified in the single class. There are two methods of pixel by pixel classification and pre-field classification, with respect to segmented areas.

Popular techniques are as fallowed:

• Multi-level slice classifier

• Minimum distance classifier

• Maximum likelihood classifier

• Other classifier such as fuzzy set theory and expert system

Step6: Verification result

The classified result should be checked and verified for their accuracy and reliability.

3.5.1. Supervised Classification

In order to classify a decision rule for classification, it is necessary to know the spectral characteristics of features with respect to the population of each class. The spectral features can be measured using ground based spectrometers. However due to atmospheric effects, direct used of spectral effect is not always available. For this reason, sampling of training data from clearly identified training areas, corresponding to define classes is usually made for estimating the population statistics. This is called supervised classification. Statistically unbiased sampling of training data should be made in order to represent the population correctly.

In this classification the user plays a primary role; based on ground data, training sets have to be selected. These training sets will be used to define the

decision space of the different classes. For each category, has a number of field locations throughout the study area has been determined on the ETM imagery. The maximum likelihood classifier was used. Since there is a considerable spectral difference between soil and vegetation on the image, classification for erosion hazard was possible.

3.6. Preparation for RUSLE Model

To spatially estimate the potential erosion risk, distribution of rainfall intensity, slope length and slope steepness factor derived from Digital Elevation Model (DEM), soil map and land cover map were used to establish a map of potential erosion risk. A revised universal model developed by USDA-ARS, Revised Universal Soil Loss Equation (RUSLE) (Wischmeier & Smith in 1978) is used to estimate the erosion risk of the study area, with the following equation:

A = R x K x LS x C x P

• R = Rainfall erosivity

• K = Soil erodibility

• LS = Topographic Factors

• C = Cover management

• P = Land Conservation

• A=Rate of erosion at certain area

Landset TM+ Geometric correction / Georeferencing Radiometric correction (Haze removal) Supervised

Classification Land Cover Map

Information erosion management practice Land Conservation Value Land conservation value map Cover Management Factor (C) Rainfall Intensity distribution

Calculate rainfall erosivity (R) Rainfall Intensity Map

Soil Classification

Map Calculate Soil erodibility (k)

Soil Erodibility Map (K factor) Rainfall Intensity distribution DEM (Raster)

Contour Line Map (12.5 interva)

TIN (Grid Coverage)

Interpolation Slope in % Slope Aspect

Topographic Map (LS factor)

Calculate with Map Calculator (A= R K LS C P)

Erosion Potential

Fig (3.4): The flow chart of prediction erosion process by RUSLE model

3.6.1. Rainfall-runoff Erosivity Factor (R)

The R factor represents the erosivity of the rainfall at a particular location. Rainfall Erosivity (R) is originally calculated as a product of storm kinetic energy (E) and the maximum 30-minute storm depth (I 30) summed for all storms in a year. An average annual value of R is determined from historical weather records and is the average annual sum of the erosivity of individual storms (Weischmeier and Smith 1978). Regarding the study area, available dataset are records of monthly rainfall amount (mm/month) and raindays (days/month), as well as an isohyets map representing the general distribution of annual rainfall. As common situation found in developing countries, there was no rainfall intensity data collected in the rainfall station. Therefore, the following equation, which is

derived for tropical areas (El-Swaify et al. 1985), was used to calculate R-Factor based on annual rainfall amount.

R = 38.5 + 0.35P

Where R: annual rainfall-runoff erosivity factor (MJ.mm.ha-1.h-1.yr-1) P: annual rainfall amount (mm) summed from monthly-recorded

rainfall amount (mm)

P value was derived from averaged annual rainfall record collected from two rain station. Considering the spatial variation of the rainfall amount, the isohyets map, with additional data from the rain stations, was interpolated to grid map with 300 meter cell size to come up with spatial distribution of annual rainfall. The grid map of annual rainfall was resample to 30 meters to allow spatial overlay with grid map of other factors of RUSLE. All climatic data were collected from BAHL (Agro-climate and Hydrology Laboratory).

3.6.2. Soil Erodibility Factor (K)

The soil erodibility factor is the average long-term soil and soil profile response to the erosive power of rainfall and runoff. Soil erodibility factor represents the effect of soil properties and soil profile characteristics on soil loss. To determine the K-value for each soil type, the following equation was used (Weischmeier and Smith 1978), given the silt fraction does not exceed 70%.

Where:

K : Soil Erodibility Factor (t.ha-1.MJ-1.ha.mm-1.h) OM : Organic matter content (%)

M : Product of primary particle size fractions

(3.2)

S : Code of Soil Structure P : Code of Soil Permeability

3.6.3. Slope Length and Slope Steepness Factor (LS)

Slope length (L) is defined as horizontal distance from the origin of overland flow to the point where the slope gradient decreases considerably, so that the deposition begins, or where runoff becomes concentrated in a defined channel (Weischmeier and Smith 1978). Surface runoff will usually concentrate in less than 400 ft, which is the recommended practical slope-length limit in RUSLE.

The LS factor is the most difficult one to derive in GIS, because the length aspect is not direct. The digital elevation model (DEM) of the study area was used in generating the LS factor. Because of the difficulties commonly experienced in generating the LS factor, two methods were used. In the first method, the slope steepness (S) and length (L) factor layers were generated separately and later overlaid to get the RUSLE slope length factor layer (Mongkolsawat et al., 1994).

In the second method, an ArcView based technique was used. The LS factor can be estimated from the DEM. The technique described for computing LS requires a flow accumulation theme. Flow accumulation can be computed from a DEM using the hydrologic extension or other watershed delineation techniques. Flow accumulation is used to estimate slope length. First we will compute slope steepness using the DEM. Before beginning the analysis, check the Analysis Properties to make sure the extent and gird cell size are acceptable. The grid cell size should be set to the grid cell size of the DEM for this analysis.

• Make the DEM active, select the Surface pull down menu as shown below and select the Derive Slope option.

• Use the map calculator to create new themes of the flow accumulation classification as 0 and 1.

• Use map calculator to compute the LS factor as shown below. Note that the slope must be converted to radians from degrees by multiplying the slope by 3.14 (pi) and dividing by 180.

The technique for estimating the RUSLE’s LS factor that will be used here was proposed by Moore and Burch (1986a and 1986b). They derived an equation for estimating LS based on flow accumulation and slope steepness. The equation is:

LS = (Flow Accumulation * Cell Size/22.13)^0.4 * (sin slope/0.0896)^1.3

Fig (3.5): The process of manipulate LS factor from DEM

3.6.4. Cover Management Factor (C)

C factor reflects the effects of vegetation cover on soil erosion, hence, this factor is often used to compare relative impacts of management options on conservation plan. The land cover-management factor (C) is the most important in RUSLE because it represents conditions that can easily be managed to reduce erosion (Renard et al., 1994). The C factor reflects the effect of cropping and management practices on erosion rates. It indicates how conservation affects the average annual soil loss and how soil loss potential will be distributed during cropping and other management schemes (Renard et al., 1997). It is based on the

concept of deviation from a standard plot under clean continuous fallow cultivations. In RUSLE, C is computed from soil loss ratios (SLR) i.e. prior land use (PLU), canopy cover (CC), surface cover (SC) and surface roughness (SR).

In this study, a land cover type map was produced as the result of image classification of Landsat ETM+ Image. The C factor value can be ranked based on the FAO standard tables (see Appendix-15) given information about land use and management for land degradation assessment.

3.6.5. Land Conservation Factor (P)

The assign value of factor (P) as a attribute field according to FAO standard value and rating. Put the value of P as 1 for non-conservative area and zero for conservative forest area. For Agricultural area, the P factor classification based on the agricultural practice, which is normally divided into: terraces cropping, contour cropping, mulch cropping and permanent ground cover type. The table showed the P- factor classification for several conservation practices in Indonesia (Hammer, 1980) in Appendix-16.

3.6.6. Calculate the Erosion Rate

Using map calculator can estimate RUSLE soil erosion for every grid cell in the area of interest.

3.7. Analysis the Model

In this research, propose to use analysis method with changed the parameter “land use and land cover” differencing between four landsat TM images of different dates to compare the rate of erosion by soil loss in each district based on four different land used and cover characteristic from of scenario.

IV. RESULTS AND DISCUSSIONS

4.1. Land Cover Change Classification

The large areas of land cover are mapped at 1:50,000 scale using Landsat TM and/or ETM data. The land cover mapping was carried out using the FAO Land Cover Classification System (LCCS), a new methodology especially created for land cover mapping and now used worldwide. The following list describes the main phases applied in the present study, for the creation of the land cover maps:

1. Satellite data selection 2. Satellite data preprocessing 3. Satellite data classification

4. Satellite data interpretation and vectorization of the resulting units 5. LCCS classification

6. Field checking

7. Composition of final land cover maps

4.1.1. Satellite Data Selection

As previously indicated, all areas under study were used in July 1989 for a rapid assessment of local physiographic and of the main land cover classes occurring there. Satellite data were selected on the seasonal based of the crop calendar for the main crops. Although the purpose of the study was to prepare land cover maps and not crop inventories, it was considered of some importance to be able to separate the crops in the field at the time of satellite data acquisition. Consequently, an image acquired in December is included in the wet season,

based calendar of the satellite image, we can consider not only permanent land cover changed by human factor but also temporal changed by season. The information of images use in this research is shown in Table(4.1).

Table (4.1): Name and History of Satellite Imagery for doing analysis

Types of image Path row Date of acquisition Landsat ETM + T 1220650126034 122-065 December 6, 2003 Landsat ETM + T 1220650816982 122-065 August 16, 1998 Landsat ETM + T 1220650919932 122-065 September 19,1993

Landsat TM + T 1220650706892 122-065 July 6, 1989

4.1.2. Satellite data Processing

This is the first step prior to the remotely sensed data to be analyzed for removing some errors of satellite imagery. There are two common types of errors, radiometric and geometric, in remotely sensed satellite image to perform for enhancing quality of image for analysis. Fig (4.1) shows the process of image preprocessing for analysis.

Fig (4.1): Flow Chart of Image Preprocessing

Raw Image

Radiometric Correction

Geometric Correction

Corrected Image

Image Calibration Atmospheric Correction

Ground Control Point - Polynomial

Rectification

4.1.2.1. Radiometric correction

The effect of atmospheric scattering caused by particles is a problem in imagery that should be removed or minimized to avoid bias to each spectral band. Histogram adjustment is one of the ways to minimize the bias. The radiometric correction algorithm is defined as below:

Output BV ijk = input BV – bias Where:

Input BV i,j,k = input pixel value at line i and column j of band k Output BV i,j,k= corrected pixel value at the same location

1989_ radiometric correction 1993_rediometric correction

Fig (4.2): The rediometric correction image of 1989 and 1993

4.1.2.2. Geometric Correction and Rectification

The geometric correction processes is normally implemented as a two-steps procedure. First, those distortions that are systematic, or predictable, are considered. Second, those distortions that are essentially random, or unpredictable, are considered (Lillesand and Kiefer, 1994). In the transformation process there is need to choose the control point type which must be polynomial --- (4.1)

for rectifying the image. Resampling is used to determine the pixel values to fill into the output matrix from the original matrix (Lilesand and Kiefer, 1994)

The image for the year of 1989,1993 and 1998 need to perform geometric correction in order to create the new dataset in the same geodetic datum (WGS 84) and map project (SUTM 48) as image in the year of 2003. The number and distribution of ground control point will influence the accuracy of the geometric correction and also root mean square error (RMS) should be less than 1.00. Re-sampling technique chosen is “nearest neighbour” and the cell seize is 30 meter. This process is very important for overlaying simultaneously both of vector and raster images.

1998 after geometric rectification 1993 after geometric rectification

Fig (4.3): Images after geometric rectification of the years 1998 and 1993

4.1.3. Satellite Data Classification

For unsupervised classification the ISODATA method was applied, and for supervised classification, the maximum likelihood classification (MLC) was preferred. To identify the sample areas for supervised classification, specific

procedures and information from topographic maps were used together with thematic maps and expert knowledge of the terrain after field checking.

The objective of these operations is to replace visual analysis of the image data with quantitative techniques for automating the identification of features in a scene. This involved the multi spectral image data and the application of statistically based decision rules for determining the land use identity of each pixel in the image.

In this classification, the interpreter knows beforehand what classes are present and whether each is in one or more locations within the scene. These are located on the image, areas the statistical analysis is performed on the multi band data for each such class. Instead of clusters then, one has class groupings with appropriate discriminate functions. It is possible that must be more that one class if only one band is used but unlikely when more than 3 bands are used because different classes seldom have similar responses over a wide range of wavelengths.

All pixels in the image which lying outside training sites are then compared with class discriminate, with each being assigned to the class it is closest to – this makes a map of established classes (with a few pixels usually remaining unknown) which can be reasonably accurate but classes present may not have been set up; or some pixels are misclassified. In this research primary image classification into (5) class, i.e.(1) forested area (2) agricultural area (3) industrial area (4) settlement (5) Bush/ grass land.

4.1.4. Satellite Data Interpretation and Vectorization of the Resulting Units

Landsat TM enhanced false color composites RGB (red, green, blue) 4,5,3; 5,3,2; 4,5,7 and 4,3,2 were used for the interpretation and delimitation of the land cover classes.

4.1.5. LCCS Classification

The Land Cover Classification System (LCCS) is a comprehensive, standardized a priori classification system, independent of the scale or method used to map. The classification uses a set of independent diagnostic criteria that allow correlation with existing classifications and legends, consequently the system could serve as an internationally agreed reference base for land cover. The methodology is applicable at any scale and is comprehensive in the sense that any land cover identified anywhere in the world can be readily accommodated. In this research, take the LCCS classification in to (10) classes which include (1) primary forest (2) Secondary forest or reforestation (3) mixed vegetable growing area (4) plantation (5) paddy field (6) bare land (7) bush land (8) extensively dense settlement (9) less dense settlement and (10) industrial area. Then extract the classified layer from the supervised classification image by using scatter gram. After doing limit to actual to the scatter gram, the classified layer appear as shown in Fig (4.4).

Fig (4.4): The scatter gram layer of each class of classified image It is need to make region to the classified scatter gram pixel, for extracting the classified layer. The following region represents the classification layer “forest” and all the forest land use classification turned to the selected color. Save the image as visual dataset then change to vector format.

Interpretation and vectorization on the screen, available in ArcView format was the preferred methodology because polygons created have vector format and can be directly transformed to a land cover map.

4.1.6. Field Checking

Field visits were undertaken in all areas under study to collect terrain information and interpretation keys useful for image interpretation. Later, field checking was carried out to test accuracy of image interpretation at selected sites and to clarify interpretation assumptions.

Forest layer

Settlemen t layer

Industrial layer Bare land

layer

Plantation layer

Bush land layer

Orchard layer

4.1.6.1. Accuracy of Land Use and Cover Mapping

The accuracy of the map was assessed by making use of ground truth samples that were collected during the fieldwork. Using the location information of these samples, a point map was created in Arc View. This map was rasterized and then combined with the land use/cover map created from Landsat ETM, acquisition date 6 December, 2003. This gave rise to confusion matrix (Table4.2) giving information on the overall accuracy of the land use/cover map. The overall accuracy of the map was 74.5% which was considered to be satisfactory for the purpose of this study (Table 4.4). The map was imported into in raster format and each class on the map was assigned a unique identifying number.

Table (4.2): The Accuracy assessment of each class of land cover classification (2003)

Over all accuracy (%) 74.5 Average accuracy of each class (%) 73 Average error reliability (%) 72

4.1.7. Composition of Final Land Cover Maps

Vector shape files were created manually, in ArcView using both the original image and the results of the supervised classification in the background to provide a basis for visual interpretation. The result of land cover classification map of the year 2003, 1998, 1993 and 1989 is shown in Appendix (6) to (9).

4.2. Determining RUSLE factors

4.2.1. Rainfall Erosivity (R factor)

The generation of the rainfall erosivity layer is based on daily rainfall amounts for Bandung regency was aggregated into annual rainfall amount, and the isohyet map of West Java. As is common to many weather stations in developing countries, the rain gauges measuring precipitation in the vicinity of rainfall station and automated rain gauge station do not collect data pertaining to rainfall intensity. The weather stations at Bandung regency which is included in upper Citarum watershed daily recorded rainfall amounts, but storm duration data as well as storm power needed to calculate rainfall intensity were not available. Because of this limitation, a relational equation that estimates rainfall erosivity using annual rainfall totals was applied.

There are (7) rainfall stations situated in upper Citarum watershed which lie around the study area. The location of the rain guage station is presented in (Appendix -1). The resulted map of rainfall erosivity (R) factor is presented (Appendix-2).

4.2.2. Soil Erodibility (K factor)

According to the soil classification map, scale 1:50,000 (Soil Research Center,1976) the study area is covered by six soil types, i.e: Alluvial Eutric, Andosal Distic, Gelosols Humus , Lithosol, Regosol Audic and Podzolic. These soil types generally are characterized by silty clay to sandy loam texture, low to high permeability, and low to very high erodibility. The soil depth is more than 90 cm, only in the small part of the area the soil depth of less than 30 cm.

In Lembang district, most of the agricultural land is grown without soil mulching type in high elevated area where the erosion rate must be high at the time of heavy raining period. Land management factor is the effective measure for reducing erosion in agricultural land. So land management in the district need to consist of contour tillage, good terracing practice and good mulching practice. Contour cultivation is highly recommended for elevation more than 1000 meter because contouring reduces soil erosion and increases soil moisture by reducing runoff losses. The effectiveness of contouring is also impaired by changes of infiltration capacity of the soil owing to surface sealing.

The area around the lake Saguling which include Batujajar, Cililin and Pajalenjan districts must be maintained as soon as possible because of high rate of erosion around that area caused by agricultural intensification and dense industrial area. The erosion has resulted in the loss of fertile top soil, the formation of rills and even gullies in some parts of the area.

V. CONCLUSIONS AND RECOMMENDATONS

5.1. Conclusions

Land cover change in image classification analysis can be divided into two, seasonal and annual change. Agricultural land and deciduous forest change seasonally as can be seen clearly in 1989 and 1993 Landsat TM+ images. The most clearly point in 2003 Landsat ETM satellite image is permanent change or annual change because of deforestation which is caused by agricultural intensification and settlement increase.

According to figure (4.5), comparisons between the classification of Landsat TM+ and ETM+ images for the year 1989, 1993,1998 and 2003 could be concluded as fallows.

• The agricultural area is in barren land condition according to the cropping schedule. Most of the barren land is intended for agriculture purpose in 1989 and 1993.

• For 2003, most of the forest areas and agricultural areas such as paddy fields and plantations are converted to residential and industrial area. • Secondary forest (reforestation) changed to bush land because of

deforestation and population increase between 1993 and 2003.

From the comparison between four different periods of land cover (2003, 1998, 1993 and 1989) and the analysis could be made conclusions. The decrease of area in percentage was more than five percent for paddy fields and secondary forest areas between year 2003 and year 1998 land cover type. At the same time

the settlement and industrial area increase dramatically. The increase in industrial and settlement areas was approximately 10 to 15 percent from 1998 to 2003.

The comparison between 1989, 1993, 1998 and 2003 soil loss rate due to erosion, there are two ways in approaching the change of soil erosion analysis: absolute and relative value of soil loss. The absolute soil loss value refers to the quantity of soil loss change while the relative value of soil loss refers to the quality or percentage of soil loss change.

• The potential for absolute soil loss in Bandung regency for the districts of Batujajar, Padalarang, Pagalenjan, Gununghalu, Pacet, Kertasari and Arjasari are described. In those districts the total soil loss areas was more than 50,000 hectare.

• The potential for relative soil loss districts are Cisarua, Padalerang, Batujajar, Lembang, Cililin, Gununghalu, Pangalengan, Pacet and Pasirjambu districts. In those areas the soil loss rate was more than (30 ton/ha/yr). Very severe erosion occurring in some areas, while the soil loss rate was more than (90 ton/ ha/yr) in those districts.

• The comparison of erosion rate in Bandung regency was based on four different land use scenario (2003, 1998, 1993 and 1989), there was less than 1 percent of very high and high erosion rate in the whole study area. Moderate and light erosion area significantly increased from 5 to 10 percent during 1998 and 2003 (Appendix-14).

In relation to the future management point of view based on Forest Stewardship Council “High Conservative Value (HCV) Principles 9” and the

combination of “FAO Land Capability Mapping and Mekong Commission's Watershed Mapping”, some specific conclusions are drawn as follows:

Conservative forest area in Pangalengan, Pasirjanbu, Gununghalu, Kertasari, Arjasari and Pacet districts which location have more than 40 percent slope gradient must be maintained as highly conservative value forest area that comprise in (HCVF 4.2) as shown in the figure(5.1).

Areas not well covered by forest vegetation and also having slope of well over 40% need to be reforested. Land cover like plantations established here cannot be harvested for human use. These areas are suitable for watershed plantations and natural regeneration area.

According to the conclusion from four different periods of soil loss by erosion in study area, there was 1825 hector of land area was need to be considered as critical as rehabilitant area for soil conservation was identified. In that area the slope percent more than 25 percent and continuously suffered the high erosion rate (60 to 75 tons/ha/yr) and very high erosion rate (more than 75 tons/ ha/yr) as shown in figure (5.1). The detail location of rehabilitant area was shown in Appendix-15. Those areas are necessary to be selected for establishing forest plantations. Community forestry sites, which will be using clear-felling practices, can be selected from such areas.

Established or permanent agriculture areas like Lembang and intensified residential and industrial area in Cililin and Batujajar are difficult to reclaim under forestland. If they are on areas over 25% slopes, there exists erosion danger. These areas are suitable for sound soil conservation practice such as agro-forestry

5.2. Recommendations

Landsat Thematic Mapper and Enhanced Thematic Mapper data provide excellent ways for discriminating natural and man-made features mapped at 1:50, 000 scales. Therefore, it should be recommended for land cover classification of land use planning. Further research should be conducted on sediment loading rate in Lake Saguling and along the water way of Citarum River. There are some hard decisions for farmers to make in preserving soil from land degradation. Soil conservation is critical to creating sustainable environment. The heavy doses of pesticides effect the sustainable environment. Further research can be done considering the effect of pesticide on land degradation. The future steps should be taken of land use and their interrelationship with land management decisions and resource policies to develop projections of future land use and management decision outcomes under a range of economic, environmental, and social scenarios.