ACTUAL CONDITION AND CHARACTERISTICS OF SLOPE FAILURE IN EAST TIMOR BY MULTIVARIATE STATISTICAL ANALYSIS By: 05KC051 Lourenco Soares A thesis submitted to Department of Civil and Environmental Engineering, Saitama University, Japan for the Requirements of Master’s Degree August 2007
Supervisor Professor Hidehiko KAZAMA
Department of Civil and Environmental Engineering Graduate School of Science and Engineering Saitama University, JAPAN
ACKNOWLEDGEMENT
At the outset, it is my duty to acknowledge with gratitude the generous help that I have received from my advisor, Professor Hidehiko KAZAMA. He is responsible for involving me in this master’s course in the first place. He taught me how to ask questions and express my ideas. He showed me different ways to approach a research problem and the need to be persistent to accomplish any goal. I also thank Mrs. Yumiko SHIRO and Mr. Masato IWAMA for their strong support in making me acquainted with Japanese life style in the past two years. I expresses with my deepest, heart felt gratitude to Mr. Kobayashi for being helpful person. Besides my advisor, I would like to thank the rest of my thesis committee: Professor Kunio Watanabe and Assoc. Professor M. Osada, who asked me good questions, gave insightful comments and reviewed my work on a very short notice. And most of all I would like to express my heart felt thanks to all staffs in Geosphere Research Institute of Saitama University (GRIS) which have direct and indirect value for finalizing this thesis. I would like to pass my great respect and Special thanks to Japan International Cooperation Agency (JICA) and Japan International Cooperation Center (JICE) for their support in funding my tuition and living expenses through out my stay in Japan to pursue the Master program in Saitama University, Japan smoothly. Especially, I would like to express my heart felt thanks to Mr. Mizuki MATSUZAKI, Ms. Yuri OSAWA, Mrs. WATANABE and Ms. Sayaka OSHIMI for their strong support, advice, suggestions, encouragement and their kindness cooperation for helping me in every aspect of my study and my life in Japan. Last, but not least, I thank my family (Amain sayang , Maun Du, Ina Noi, Alin Eqi), the late my father”† Salvador Soares” and my mother “ Andreza Soares, for giving me life in the first place, for educating me with aspects from both arts and sciences, for unconditional support and encouragement to pursue my interests, even when the interests went beyond boundaries of language, field and geography. My brothers (Maun Domingos, Enty, Rito, Abes and Aje) and friends: for sharing experience of life and dissertation to me, for listening to my complaints and frustrations, and for believing in me, most of all supporting.
i
CONTENTS
Acknowledgement Contents List of figures List of tables Abstract CHAPTER I Introduction 1.1 1.2 1.3 Background of study ………………………………………………….. Propose and scope of the study ……………………………………….. Data collection and methodology of research ………………………... 1 2 4
CHAPTER II Study Site Description and Literature review 2.1 Study site description ………………………………………………… 2.1.1 2.1.2 2.1.3 2.1.4 2.1.5 2.2 Geographical condition, location, and boundaries of study site.. Topography ……………………………………………………. Geology, landforms and soil ………………………………….. Climate ………………………………………………………… Vegetation ……… …………………………………………….. 9 9 13 14 19 24 31
Literature review ………………………………………………………
CHAPTER III Actual Condition, Characteristics and Distribution of Slope Failure in East Timor 3.1 3.2 Introduction ……………………………………………………………. Characteristics and distribution of slope failure in East Timor ……….. 3.2.1 3.2.2 3.2.3 Lithology ……………………………………………………… Vegetation …….. ……………………………………………… Inclination angle of slope ………………………………………. 36 40 41 44 46
ii
3.2.4 3.2.5 3.2.6 3.2.7 3.2.8
Direction of slope ……………………………………………… Landscape topography ………………………………………… Elevation ………………………………………………………. Slope width………. ……………………………………………. Slope length ……………………………………………………
49 51 53 54 58
CHAPTER IV Analyzing Method 4.1 4.2 4.3 Logistic regression analysis …………………………………………… Independent variables and sampling …………………………………. GIS application for slope failure mapping …………………………….. 63 67 70
CHAPTER V Analysis Result 5.1 5.2 5.3 Introduction ……………………………………………………………. All study site analysis result …………………………………………… Specific site Analysis ………………………………………………….. 5.3.1 5.3.2 5.3.3 Bobonaro site …………………………………………………. Cailaco site ……………………………………………………. Zumalai site …………………………………………………… 72 72 82 82 92 100
CHAPTER Conclussion and Future Subject 6.1 Conclusions…….…………………………………………………………. 6.2 Future Subject ……………………………………………………………... References ……………………………………………………………………. APPENDIX A: Physical Data of Slope Failure and Unfailure slope APPENDIX B: Logistic Regression Analysis 109 110 111
iii
LIST OF FIGURES
FIGURE 1.1 1.2 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 2.10 2.11 2.12 2.13 2.14 2.15 2.16 2.17 2.18 2.19 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 3.10 3.11 3.12 3.13 3.14 DESCRIPTION Flow chart of research methodology Slope failure location with aerial photograph Boundary of study site Study site and slope failure Study site and unfailure slope Topography of East Timor East Timor geological map physical types of East Timor Areas prone of landslide and flooding in East Timor Altitude and mean temperature correlation Monthly distribution of rainfall in East Timor (based on data from Ferreira 1965) The amount of daily rainfall from July – December 2006 in Dare station The amount of daily rainfall from July – December 2006 in Aileu station The amount of daily rainfall from July – December 2006 in Betano station Climate Natural distribution of forest in East Timor Actual forest covers Firewood cut by community as o source of income and used for cooking Cutting and burning the forest by community Sifting agriculture (slashes and burn agriculture) Category of land cover in East Timor Older landslide topography in East Timor Older landslide topography in East Timor Recent landslide topography in East Timor Recent landslide occurred on cut slope alongside road in East Timor Surface failure on hill slopes of mountainous in East Timor Surface failure on hill slopes of mountainous in East Timor Lithology Distribution of vegetation Distribution of inclination angle of slopes failure Distribution of direction of slope Landscape topography Distribution of elevation Width of landslide Width of surface failure iv PAGE 6 7 10 11 12 14 17 18 18 20 20 22 23 23 24 26 27 28 29 29 31 37 37 38 39 39 40 44 46 48 50 52 54 56 57
Width of surface failure and landslide Length of landslide Length of surface failure Length of surface failure and landslide mix Flow chart of logistic regression analysis Flow chart of Production of probabilities of slope failure maps based on GIS techniques Ranking of the top ten significant item and category based on influence ratio in all study site The top ten ranking of interaction term when combined with other variable based on influence ratio in all study site Observed groups and predicted probabilities of slope failure by logistic regression analysis Histogram of p redicting for probabilities of slope failure Map of relative slope failure susceptibility Ranking of the top ten significant item and category based on influence ratio in Bobonaro site The top ten ranking of interaction term when combined with other variable based on influence ratio in Bobonaro site Observed groups and predicted probabilities of slope failure by logistic regression analysis Histogram of predicting for probabilities of slope failure in Bobonaro site Ranking of the top ten significant item and category based on influence ratio in Cailaco site The top ten ranking of interaction term when combined with other variable based on influence ratio in Cailaco site Observed groups and predicted probabilities of slope failure by logistic regression analysis Histogram of predicting for probabilities of slope failure in Cailaco site Ranking of the top ten significant item and category based on influence ratio in Zumalai site The top ten ranking of interaction term when combined with other variable based on influence ratio in Zumalai site Observed groups and predicted probabilities of slope failure by logistic regression analysis Predicting for probabilities of slope failure in Zumalai site
LIST OF TABLES
TABLE 2.1 2.2 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 3.10 3.11 3.12 3.13 3.14 4.1 4.2 5.1 5.2 5.3 5.4 5.5 5.6 5.7 5.8 DESCRIPTION Land use in East Timor, Indonesia government estimation Land use, Alternative estimation, Saldanha Density of slope failure in study site Description of geological structures in each study area Lithology Distribution of vegetation Distribution of inclination angle of slope failure Distribution of direction of slope Landscape topography Distribution of elevation Width of landslide Width of surface failure Width of the mix of surface failure and landslide Length of landslide Length of surface failure Length of surface failure and landslide mix Classification of predicted the probabilities of slope failure from the logistic regression analysis Categories of the independent variables Classification table of cut value 0.50 Coefficient values and influence ratio of logistic regression of each item and category in all study site Cofficient values and influence ratio of logistic regression of interaction term with other item and category in all study site Classification of predicted the probabilities of slope failure from the logistic regression analysis Predicting for probability of slope failure Classification table of cut value 0.50 in Bobonaro site Coefficient values and influence ratio of logistic regression of each item and category in Bobonaro site Cofficient values and influence ratio of logistic regression of interaction term with other item and category in Bobonaro site Predicting for probability of slope failure in Bobonaro site Classification table of cut value 0.50 in Cailaco site Coefficient values and influence ratio of logistic regression of each item and category in Cailaco site Cofficient values and influence ratio of logistic regression of interaction term with other item and category in Cailaco site Predicting for probability of slope failure in Cailaco site Classification table of cut value 0.50 in Zumalai site vi PAGE 30 30 41 39 43 45 48 50 52 53 55 56 57 59 60 51 67 68 73 73 74 79 80 82 82
5.9 5.10 5.11 5.12 5.13 5.14
83 90 92 92 93 99 100
5.15 5.16 5.17
Coefficient values and influence ratio of logistic regression of each item and category in Zumalai site Cofficient values and influence ratio of logistic regression of interaction term with other item and category in Zumalai site Predicting for probability of slope failure in Zumalai site
101 102 107
vii
ABSTRACT
East Timor has risk number of natural hazards. Each year, heavy seasonal rain falling on steep slopes causes frequent flash flooding and slope failure, which are considered to be the two major natural hazards in the country. Apart from their potential to cause casualties and damage to rural communities, these events cause major disruption to the fragile road network, isolating communities and even whole districts for a long duration. Slope failures (i.e., landslide and surface failure) in mountainous terrain often occur as a result of heavy rainfall, resulting in the loss of life and damage to the natural environment. In this regard, slope failure hazard assessment as well as identify the characteristics and distribution of slope failure can provide much mitigation through proper project planning and implementation. Propose of this study are to know actual condition and characteristics of slope failure and to determine clearly the factors influencing of the slope failure occurrence in East Timor. The factors that influent to the slope failure in study area may be categories in the intrinsic variables that contribute for slope failure, such as geology, inclination angle of the slope, vegetation, elevation, direction and landscape topography of slope. Logistic regression analysis is a multivariate technique that considers several physical parameters that may affect probability. This modeling is intended to describe the likelihood of slope failure on a regional scale, and is very suitable for the assessment of slope failure actual condition and its characteristics because the observed data consist of item and category with a value of 0(absence of slope failure) or 1(presence of slope failure). The predicting and assess of slope failure occurrence for the training samples in this analysis. If we have a model that successfully distinguishes the two groups based on a classification cutoff value of 0.5. Result analysis shown that the model produced a concordance rate of 90 % with the use of 0.5 as a classification cutoff value. By examining this result to predict that’s factors influencing slope failure, we can see what a different classification rule should be adopted when applying the model analysis to each factor in the study area and obtain regression model composed of significant variables. The influence of the interaction among factors contributing for slope failure occurrence was examined. When the interaction term were introduce, the proportion of the observed all items and category predicted as high influence ratio increased by 1 to 4 times of individual category, which indicated a better prediction. viii
The comparison of the results from the analysis including the interaction terms among category and the individual category, the interaction term indicate that interactions among the variables of category were found to be significant for predicting probability of slope failure. From the result, slope failure would most possibly occur in area where cover by bare land and grassland and the elevation ranges from 200m to 800m, the surface slopes is steep and thin sedimentary rocks.
ix
CONTENTS OF APPENDIX
Appendix A : Physical data of slope failure and un-failure A.1 Physical data of slope failure A.1.1 Bobonaro site …………………………………………………. A.1.2 Cailaco site ……………………………………………………. A.1.3 Zumalai site …………………………………………………… A.1.4 Atsabe site ……………………………………………………. A.1.5 Maliana site …………………………………………………… A.1.6 Ainaro site ……………………………………………………. A.1.7 Hatolia site …………………………………………………… A.1.8 Hatobuilico site ……………………………………………… A.2 Physical data of unfailure slope A.2.1 Bobonaro site …………………………………………………. A.2.2 Cailaco site ……………………………………………………. A.2.3 Zumalai site …………………………………………………… A.2.4 Atsabe site ……………………………………………………. A.2.5 Maliana site …………………………………………………… A.2.6 Ainaro site ……………………………………………………. A.2.7 Hatolia site …………………………………………………… A.2.8 Hatobuilico site ……………………………………………… Appendix B : Logistic regression analysis result B.1 B.2 All study site ………………………………………………………. Specific site B.2.1 Bobonaro site ………………………………………………. B.2.2 Cailaco site ………………………………………………… B.2.3 Zumalai site ……………………………………………….. 165 178 189 151 134 139 143 145 147 148 149 150 117 122 126 128 130 131 132 133
CHAPTER I INTRODUCTION
1.1 Background of Study
East Timor is a rugged island with a narrow or non existent coastal plain along its northern coast and a southern coastal plain that varies from less than a kilometers wide in some areas to as much as 20 km in others. Highest mountain with a height of 2,963 meters is the Tatamailau or Ramelau in the Ainaro district. Slopes are steep, with as much as 44% of the country having a slope of 40% or more. Slopes this steep may need a zigzag path to climb. The soils are limestone-dominated. Such soils are prone to erosion, particularly on steep slopes and where vegetation cover has been degraded by poor agricultural practices or deforestation. This is the case in many parts of East Timor where the natural vegetation has been modified by human activity over centuries leaving sparse savannah woodland or grassland in most areas. East Timor is dryer than most equatorial islands, receiving most of its rainfall during the northwestern monsoon, which occurs from December to March. Southern slopes receive additional rain during the shorter southeast trade winds period between May and July. East Timor has risk number of natural hazards. Each year, heavy seasonal rain falling on steep slopes causes frequent flash flooding and slope failure, which are considered to be the two major natural hazards in the country. Apart from their potential to cause casualties and damage to rural communities, these events cause major disruption to the fragile road network, isolating communities and even whole districts for a long duration.. East Timor has two climate seasons are wet and dry season. From November to April, the country is risk of tropical cyclones and lesser tropical storms, which can cause coastal flooding and wave damage. In the dry season, drought conditions exist in large parts of East Timor. A delay in the onset of seasonal rains can become disastrous as fires can get quickly out of control. East Timor has a very fragile environment. It is particularly dried compared with other parts of the region, and is prone to regular droughts. Deforestation combined with steep slopes, thin soils and heavy seasonal rains have resulted in erosion and soil loss.
1
The situation has been exacerbated by deforestation, which has become more substantial during the last three decades. One of the country’s most valued forest resources is sandalwood which has now been reduced to just a few stands due to of over-exploitation. Another problem is that many rural communities rely on selling wood for fuel as source of family income and as a result, have contributed to deforestation. Geological hazards also threaten East Timor. Areas to the north of the island have experienced earthquakes of up to 6.9 on the Richter scale within the last 10 years. These can cause local tsunamis. A four-meter-high tsunami, caused by a major earthquake, struck the north coast of Timor in 1995. In addition, other hazards exist, including major transport accidents; urban fires and agricultural hazards. These risks are likely to increase as the nation develops unless necessary precautions are made and regulations put in place. Slope failures (i.e., landslide and surface failure) in mountainous terrain often occur as a result of heavy rainfall, resulting in the loss of life and damage to the natural environment. In this regard, slope failure hazard assessment as well as identify the characteristics and distribution of slope failure can provide much mitigation through proper project planning and implementation.
1.2 Propose and Scope of the Study
It is difficult to examine the natural hazard as well as slope failure hazard in East Timor because of the lack of consistent data, however little data has been collected to provide this study. The primary aims of this initial study are to identify the major influence factors for slope failure in East Timor. Logistic regression analysis is a multivariate statistical analysis has been used extensively at most of previously researcher to predict the factors influence to the slope failures occurrences. The purpose of this study is to present a method that utilizes and employs statistical analysis to define the physical parameters contributing to the occurrence of landslides. This method allows a series of statistically meaningful and independent variables to be included in the assessment of the analysis model. The procedure is based on the actual slope failure cases and is therefore representative of failure conditions and relatively objective. Logistic regression analysis describing in this study is to: • To know the actual condition and characteristics of slope failure in East Timor
2
• • •
Identify clearly the factors that are related to slope failures, Estimate the relative contribution of factors causing slope failures, and Establish a relation between the factors and slope failures.
The scope of activities with developing and applying the logistic regression analysis in this study consist of five main steps: • • • • • Pre-selection of variables based on a slope failure distribution analysis; Selection of statistically significant variables by a P-value significance test; Logistic regression modeling with those variables that passed the significance test; Logistic regression modeling with significant variables including the interaction terms; and Evaluation of the model results.
In the first step, a slope failure characteristics analysis is used to pre-select the variables that are relevant for the regression. This analysis involves overlaying the variables of category of slope failure occurrences and the variables of category of a factor (such as lithology), then calculating the percentage of coverage of the slope failure occurrence on each class for each input factor, such as slope inclination angle within elevation factor. By comparing the slope failure distributions, a preliminary ranking of the variables can be developed. Important variables will be considered in the following significance tests. In the second step, the significance p-value of 0.05 is specified as the cut-off value to choose the variable for further analyses and 0.10 is chosen as the value for elimination of insignificant variables. The variables that passed the significance test can be entered into the logistic regression modeling in the next step. After the steps of pre-selection and significance test, we can know the total of the independent variables were selected for the regression analyzing. In the third step, the model is checked for its goodness of fit by entering a variable or removing a variable. Following the SPSS procedures, 20 iterations are preferred to obtain optimal models of analysis. The final suitable logistic regression analysis is based on the variables presented in the final step of the statistical calculation in the SPSS program, and the regression coefficients are obtained.
3
In the fourth step, the interaction terms representing the interactions among variables are entered into the logistic regression analysis. In particular, the interactions among variables from six category factors (i.e., lithology, vegetation cover, slope aspect, elevation inclination angle and landscape topography) are selected to form the interaction terms for the
regression. The interactions among two, three, and four variables at one time were tested. Only significant interaction terms are retained for analyzing. When interaction terms are introduced into the model, the ranking of the significance of some of the variables will change. Some of the variables showing significance in the previous step may become insignificant, and some of the interaction terms showing significance are added into the model. After many tests with the interaction terms, the model that produces the best prediction result is adopted as the final optimal model. In the fifth step, the models obtained from above and the factors influence to the slope failure occurrences generated from the models are evaluated. Slope failure probability values between 0 and 1 at each unique-condition unit are obtained from the final regression.
1.3 Data Collection and Methodology of Research
Slope failure often occurs at specific locations under certain topographic and geologic conditions. Therefore it is important to utilize existing data (history of the problem and data review) in order to understand the topography, geology, and properties of similar slope failure. It is also important to understand their relationship with meteorologist factors, chronology of topographic change or erosion by rivers, earthquakes, and other factors which may have a relationship with the slope deformation surrounding the study site prior to the detailed investigation. In this study, data collections to provide this research are: • • • • Aerial photograph with magnitude scale 1:13,000 Topography map with magnitude scale 1:15,000 geology map with magnitude scale 1:350,000 Rainfall Data (July 2004 – December 2006)
4
Investigations of Aerial photographs are used to understand the chronologic and topographic changes over the country. Furthermore, in order to be able to effectively interpret the phenomena related to slope failure. By utilizing aerial photographs, it is possible to interpret landslide phenomena and warning signs, geology structure, topography and distribution of vegetation type. Topographic investigation is necessary to identify any changes in the site topography. That can be accomplished by recognizing; 1) the overall topographic feature of the site; 2) understanding the topographic characteristics of the site slopes; and 3) estimating the regional geologic structure of the site. Such methods include comparing the aerial photographs of the site and vicinity taken prior to and after the sliding, and interpreting the topographic maps and aerial photographs. Geological map is necessary to investigate geologic structure, however to identify the bedrock distribution, rocks types and rock mass engineering properties in the surrounding study site. Based on aerial photograph and topographic map in the study area, there are 506 number of slope failures from the inventory. For each slope failures inventory, it includes information such as location, slope geometry (slope inclination angle, direction, width and length), geology factor (rocks types), vegetation cover (high tree, low tree, grassland and no vegetation), landscape topography (valley, ridge and flat) and slope aspect (direction) are used for actual condition and characteristics of slope failures analysis. Considering the regional variations identified and data availability in the above, six factors were considered in this study: geology factor, vegetation cover, slope gradient (i.e., slope inclination angle), elevation, landscape topography and slope aspect (i.e., direction). Detail research methodology in this study has shown in Figure 1.1.
Map of Locations Representing of the Selecting Area of Slope Failure and Unfailure Slope by Random - Lothology - Slope Gradient - Vegetation Cover - Slope Aspect - Elevation - Landscape Topography
Extraction of Independent Variables for points representing of Slope Failure and Unfailure Multivariate Statistical Analysis by Logistic Regression Analysis
Stepwise of logistic Regression Analysis
Development of Logistic regression Analysis
Verification of the probabilities and Susceptibilities of Slope failure mapping
Result of Analysis and Discussion
Figure.1.1 Flow chart of research methodology
6
Figure 1.2 Slope failure locations with aerial photograph.
7
A key assumption using this approach is that the potential (occurrence possibility) of slope failure will be comparable to the actual frequency of slope failure. After the study area was selected, slope failure areas were detected in the study area by investigation of Aerial photograph. The maps of aerial photograph used were these from January 2000 (Figure 1.2), after slope failure. This air photograph, in combination with logistic regression analysis result and GIS was used to evaluate and predicted the probability of slope failure in the study area. A GIS database has been developed using ArcGIS version 3.3 software. The slope failure in the study area and the factors contributing for slope failure have been recorded and saved as separate layers in the database. All the data layers were in vector format, transformed in grids with cell size 30x30 meters.
8
CHAPTER II STUDY SITE DESCRIPTION AND LITERATURE REVIEW
2.1 Study Site Description 2.1.1 Geographical Condition , Location, and Boundaries, of Study Site
East Timor is approximately the eastern half of the island of Timor, and part of the Lesser Sunda Island chain, distant from Australia by only 500 km. It is between longitudes 1270 22” and 1320 25” and latitude 80 17” and 100 22” with a general orientation of southwest to northeast. The area of East Timor as a whole is only about 15,007 km2 and the coastline is 706 km. Timor’s boundaries are as follows: • • • • In the north, the boundary of Wetar Strait with Ombai Strait. In the east, the boundary with the Maluku Strait. In the south, the boundary with the Timor Sea. In the west, the boundary with Nusa Tenggara Timor, the eastern region of Indonesia.
In this study, the study site at the western part of East Timor. There are lies between latitude 080 52’’ 30’’ and 090 15’’00’’ to South and longitude 1250 15’’ 30’’and 1260 15’’00’’to East, and has area about 1448 km2 with elevations ranging from 200m to 2100m. The study area is mountainous area, which is also landslide prone, and is quite flat in the south. The underlying bedrock is limestone, siltstone, sandstone, shale and conglomerate. Most folds are developed in the western mountainous area and a thrust fault extends from north to south of the study area. The study site are covering eight sub district in western part of East Timor, there are Bobonaro, Cailaco, Zumalai, Atsabe, Maliana, Ainaro, Hatolia and Hatobuilico (Figure 2.1, Figure 2.2 , and Figure 2.3 ).
9
N
Firuge 2.1 Boundary of study site
Boundary of study site
Atsabe Cailaco Hatolia Hatobuilico
Maliana Zumalai Bobonaro
Ainaro
10
Atsabe Hatobuilico
Cailaco
Hatolia
Maliana Zumalai Bobonaro
Ainaro
Figure 2.2 Study site and slope failure 11
Atsabe
Cailaco Hatobuilico Hatolia
Maliana Bobonaro Zumalai
Ainaro
Figure 2.3 Study site and unfailure slope
12
2.1.2 Topography
A mountain range runs from the east to the west of East Timor. The mountainous terrain results in many watersheds and streams, making transportation very difficult. The land is made up of limestone, coral, thick clayey soil, sand and a small amount of volcanic origin. In East Timor there are seven mountains with heights over 2000m as seen in the following table. The highest mountain with a height of 2,963 metres is the Tatamailau peak of the Ramelau Range in the Ainaro district. Name of District Height Mountain Above Sea Level 1.Tatamailau Ainaro 2,963 metres 2.Sabiria Aileu 2,495 metres 3.Usululi Baucau 2,620 metres 4.Harupai Ermera 2,293 metres 5.Cablake Manufahi 2,495 metres 6.Laklo Manatuto 2,050 metres 7.Matebian Baucau 2,373 metres As a broad outline, the watersheds of East Timor can be divided into two areas; northern and southern. Of the many rivers in East Timor, the following rivers flow all year round; the Laklo river in the district of Manatuto, the Seical river in Baucau district, the Bulobo, Marobo, Malibaka and Nunura rivers in Bobonaro district, Gleno river in Ermera district, Karau Ulun in Manufahi district, the rivers of Dilor, Uca, Uwetoko, Bebui and Irabere in Viqueque district, the Loes river in Liquica, and the Tono river in Oecussi. Overall the climate in East Timor is classified as tropical. The minimum temperature range is 18-21ºC while the maximum temperature range is 26-32ºC. In the north (as far east as Baucau) the rainy season begins in November and is usually accompanied by a westerly monsoon; the months of May and October are months of change from dry to wet season. In the east and the south the situation is different - the rainy season is at its height in April. The dry season occurs during May, and the rainy season returns at the beginning of June until August. When it is winter in Australia (August to October), sometimes the temperature in East Timor can be as low as 18ºc. This is also true of the opposite scenario. When it is summer in Australia, the temperature is high on the coast of East Timor, even in the rainy season.
13
Figure 2.4 Topography of East Timor (Source from Internet)
2.1.3 Geology, land forms, and soil
Timor is a continental fragment, not a volcanic island. The foundation is largely made up of limestone and other sedimentary deposits. This differentiates it from its neighbors to the north and west in the Sunda chain which are volcanic. It is theorized that Timor, in fact, is a piece of the Australian geological plate which, separated from the mainland, has been pushed into the Indonesian plate. (Monk et al. 1997:23) That it has been repeatedly uplifted and submerged over the millennia accounts for the presence of coral layers in the soil at heights of up to 2,000 meters above sea level. The erosion of these rocks is considerable. The topography of East Timor is dominated by a massive central backbone of up to 3,000 meters, the Ramelau mountain range, which is dissected by deep valleys prone to flash floods. Toward the northern side, the mountains extend almost to the coast without extensive
14
plains. To the south, on the other hand, mountains taper off some distance from the sea leaving a wide littoral plain, more propitious for agriculture. The plain is 20 km and even 30 km wide running almost the length of East Timor and widens at the eastern end. There are more perennial streams flowing to the southern coast which allow for more agriculture and irrigation. The Fuiloro plateau, in the far East, descends in altitude southwards, from 700 meters to 1500 meters. The slope is almost unnoticeable due to the large area, which may have been the primitive lagoon of a big fossil atoll. Three other main planaltic formations surround it: Nári in the north, Lospalos to the center-west and Rare to the south. Nestled in the mountain range near the border with West Timor lies the low plateau of Maliana in what was once a gulf. This area is better suited to irrigated agriculture than the rest of East Timor. As much as 44 percent of East Timor may have a slope of land of more than 40 percent. (Monk et al. 1997:52; Dick 1991) A slope of 40 percent is difficult to descend and may need a zigzag path. Bierenbroodspot (1986 in Monk et al. 1997:107) suggested the following erodibilty classification and appropriate uses for sloping land on Timor: • • • Land with less than 17 percent slope tends to be suitable for cultivation provided that any incipient soil erosion is controlled; Land between 17 percent and 30 percent is best used for grazing as soil erosion cannot be controlled on such steep slopes under permanent or shifting cultivation; Land over 30 percent suffering from soil erosion is unsuitable for sustainable agriculture and can require reforestation or conversion to suitable tree or perennial cover crops. Soils are ultimately the combination of base rock, topography, climate, vegetation and, to some extent, the fauna which is present in any one place. Topography influences the weathering, depth, erodibility, infiltration, and leaching of a soil. The major limitations to plant production, and therefore to agriculture, are steep slopes and shallow soils. The outerarc islands, dominated by limestone, generally have lower, rounded hills with relatively infertile, alkaline soils. Often the better soils are only on the alluvial deposits along the coasts and in depressions such as lake or lacustrine basins surrounded by steeper, eroded land. Such a lacustrine basin occurs in north central Timor (Maliana).
15
Climate is perhaps the most important factor affecting the development of tropical soils (Mohr et al. 1972). The most important climatic factor affecting tropical soil fertility and structure is temperature. Up to 20°C, humus forms faster than it is broken down, enriching the soils with nutrients and improving its structure (Chambers 1983). Above 20°C, and particularly in hot, arid Conditions, bacteria decompose dead vegetation faster than it accumulates, with the result that humus and fertility levels diminish. Thus, many tropical soils have a low organic content and inherent low fertility. Tropical soils can maintain natural fertility where climatic conditions favor the accumulation of humus. This occurs in continuously moist soils found in wetter regions or higher altitudes; or when nutrients are resupplied from outside the system, such as when a volcanic eruption spreads mineral-rich ash deposits over the land. A second important climatic factor affecting fertility and structure is the soil moisture regime, that is, the relationship between the length of the dry season and total rainfall. Most of the area experiences a seasonal climate. Prolonged droughts are followed by total annual precipitation which falls within a few months or even days. This strongly affects the movement of salts and minerals through the soil. Soils may bake hard and crack during a prolonged dry season. These conditions are intensified in savannas, because the annual fires remove the supply of new organic matter and, at the end of the rainy season with ground cover at a minimum, heavy rainfall may result in surface runoff with potential for rill and gully erosion. The soils of the outer-arc islands tend to have less clay and, as a result, lower water holding capacity (WHC) than the inner volcanic arc islands (Carson 1989). Shallow, calcareous soils on raised coral reefs on islands such as Timor have a limited WHC; Timor's soils are 20-30 cm deep over the island (Mahadeva and Laksono 1976), except where there are lake deposits. The area with steep slopes and thin soils is naturally biased toward high rates of erosion. Some local farmers have an understanding for the fragility of the soil and have developed a sophisticated indigenous method of soil conservation. In other areas, however, soil is being lost at high rates through inappropriate land management. In particular, high losses of organic matter occur during and shortly after clearing, and before establishment of suitable cover crops. Under such conditions, intense bombardment of the soil surface by rain can quickly break down soilorgano aggregates, thus permitting high erosion losses. In addition, surface temperatures
16
increase on cleared land, thus increasing oxidation and loss of organic matter. As it is difficult to restore organic matter, conservation measures such as early planting of cover crops, incorporation of plant residues and erosion control should be strictly followed (FENCO 1981).
Figure 2.5: East Timor geological map (Instituto Superior Tecnico, Portugal, 2000)
17
Figure 2.6 Physical types of East Timor
Source: Monk et al. 1997: Figure 2.10, originally from RePPProT 1989b
The physical types present in East Timor are 2 - tidal swamps; 4 - meander belts; 7 - Fan and lahars; 8 - terraces; 9 undulating rolling and hillocky plains; 10 - hills; and 11 - mountains. (Monk, et al. 1997:50; original RePPProT). A revised draft map is in preparation for East Timor by the Geological Research and Development Centre, Bandung -GRDC. The geology of East Timor was mapped previously by Audley-Charles (1968).
Fig.2.7 Areas prone of landslide and flooding in East Timor (Source: Monk et al. 1997: Figure 2.13originally RePPProt 1989a.
18
2.1.4 Climate
Knowledge of climatic conditions is of great importance for environmental management. Climatic maps showing the amount of rainfall, including dry or drought periods indicate what crops that will grow on an island or in a particular valley, or what pests may migrate into the area if particular crops are cultivated. Much historical data exists for both temperature and rainfall from the Portuguese colonial period. East Timor continues to have more stations for measuring these and other factors than do the neighboring areas in Indonesia. Climate is a function of the latitude, wind patterns bringing rain, rainfall volume, seasonality, and intensity, soils, and the altitude above sea level. There is a clear correlation for East Timor between altitude and average temperature and seasonal variations as shown by Felgas (Figure 2.8, reproduced in Monk 1997). While the general climate in East Timor can be classified as hot (average temperature 210 C) and humid (70-80 percent), the geographic position and the topography is such that climatic conditions differ substantially between mountainous regions and lower altitudes. Even regions of the same altitude have very different climates when separated by high mountains which act like a wall. Therefore, since topography is not equal to climate, a system that separates lowlands, mountains, and plains is a useful first step to classifying climactic conditions. On the southern coast rainfall is high, with volumes of 2,000 mm or more per year spread over a longer period of months. On the northern coast, at the same altitudes, rainfall could be as little as 500-1,000 mm per year and concentrated in a shorter period of months. The Indonesian government, (RePPProT) used the Schmidt and Ferguson method of counting and comparing months with more than or less than 100 mm rainfall each and the Fontanel and Chantefort method of combining this with temperature data. The result is that the northern coast is basically seasonally dry except on the coast which is permanently dry. The southern coast is permanently moist (Monk et al. 1997:75-77). A permanently moist climate might allow for the growing of two annual harvests of crops, such as rice. However, for the purpose of land use planning, a more detailed discrimination of climate is necessary. (See sections on rainfall, vegetative cover, and agriculture, below.)
19
Figure 2.8: Altitude and mean temperature correlation
Source: Monk et al. 1997: Figure 2.19, originally from Felgas 1956
Figure 2.9 Monthly distribution of rainfall in Timor Leste (based on data from Ferreira 1965).
20
It is difficult to examine present climate risk in East Timor because of the lack of consistent climate data. During the Portuguese period several stations measured rainfall/climate data for varying periods from 1914 to 1975, but many of these records are incomplete. It is unclear how much data was recorded during the period of Indonesian control from 1975 to 1999. Since 1999 there have been no meteorological or hydrological services available in East Timor. In November 2000, 50 rain gauges were distributed around the country by the Department of Agriculture and funded by AusAID, however little data has been collected from these gauges (Ongoing monitoring and educational activities are important to establish continuity in such programs). Automatic weather stations have been installed at the main airports (Dili, Baucau and Suai) by the Australian Bureau of Meteorology (Darwin). Weather or seasonal climate forecasts have only been used sporadically by the National Disaster Management Office, and these were based on information available from the internet. Furthermore, there are currently no means to communicate this information to the users that require it. The Australian Bureau of Meteorology will be providing weather forecasts for East Timor for as long as Australian forces are present in the territory (see http://www.bom.gov.au/reguser/by_prod/aviation/). As well as this lack of temperature and rainfall data, there is a lack of consistent data on a range of climate-related processes like river runoff, tides, floods, and groundwater levels. This lack of data makes it difficult to assess whether climate is changing in East Timor. There is also insufficient data on which to base scenarios of future climate changes and its impact on environmental and social systems. Nevertheless, some broad conclusions about climate change in East Timor can be drawn and these will be discussed in the following pages. East Timor is predominately influenced by the monsoon climate. There are two distinct rainfall patterns: the Northern Monomodal Rainfall Pattern produces a 4-6 month wet season beginning in December which affects most of the northern side of the country and tapers to the East; and the Southern Bimodal Rainfall Pattern which produces a longer (7-9 month) wet season with two rainfall peaks starting in December and again in May which affects the southern side of the country (Keefer 2000: 11). Rainfall can be broadly described as being low to very low along the northern coast of East Timor (<1000mm/annum), low to moderate throughout the central and elevated areas (1500-2000mm/annum), and relatively
21
high (>2500mm/annum) in high altitude areas which are mostly in the west. In common with most tropical locations, extremely heavy rainfalls occasionally occur over East Timor during relatively short time intervals (Figure 2.9). The general climatic conditions define two zones: northern areas and southern areas, divided by mountains into: • The northern area characterized by one rainfall peak within four to six months in the wet season. The northern coastal areas have an average yearly rainfall from 500 to 1500 mm, while higher altitudes above 500 m receive abundant rainfall from 1 500 to 3 000 mm, an average of monthly rainfall from 50mm to 150mm (Figure 2.10). • The southern areas characterized by two rainfall peaks that appear within seven to nine months in the wet season. The first peak appears between December and February and the second peak appears between May and June. The southern coastal areas have an average annual rainfall from 1 500 to 2000 mm. The areas above 500 m receive more abundant rainfall from 1 700 to 3 500 mm, an average of monthly rainfall from 70 mm to 150mm (Figure 2.11 and 2.12).
The amount of daily rainfall in Dare station
July-Dec. 2004 240 220 200 180 160 140 120 100 80 60 40 20 0
5.1 - 10 15.1 - 20 25.1 - 30 35.1 - 40 45.1 - 50 55.1 - 60 65.1 - 70 75.1 - 80 85.1 - 90 95.1 - 100 0
2005
2006
Days
Intensity rainfall (mm/day)
Figure 2.10 The amount of daily rainfall from July 2004 – December 2006 in Dare station
Figure 2.11 The amount of daily rainfall from July 2004 – December 2006 In Aileu station
The amount of daily rainfall in Betano station
July-Dec. 2004 260 240 220 200 180 160 140 120 100 80 60 40 20 0
0 0.1 - 5 5.1 - 10 10.1 - 15 15.1 - 20 20.1 - 25 25.1 - 30 30.1 - 35 35.1 - 40 40.1 - 45 45.1 - 50 50.1 - 55 55.1 - 60 60.1 - 65 65.1 - 70 70.1 - 75 75.1 - 80
2005
2006
Days
Intensity rainfall (mm/day)
Figure 2.12 The amount of daily rainfall from July 2004 – December 2006 In Betano station
23
Figure 2.13: Climate
Source: Monk et al. 1997: Figure 2.17, originally from RePPProT 1989a
2.1.5 Vegetation
The present vegetation cover is a combination of what could be there given the climate and the particularities of each area, and anthropic actions of settlements, clearings, agriculture, grazing, plantations, etc. This section speculates what the natural distribution of forests and grasslands in East Timor would have been. It also assesses what is known of the historical distribution of vegetation cover. It was noted previously that East Timor suffers from an exceptionally dry climate, especially in the northern half. This condition directly affects the likely historical distribution of forest. Monk suggests that classification of forests in this area is particularly difficult because of the extreme influence of altitude and rainfall patterns on forest types. These vary widely in small areas and along steep slopes. Not enough work has been done on classification specifically for East Nusa Tenggara, Maluku and East Timor. Figure 2.14 shows the types of forest which would be naturally occurring in eastern Indonesia based on the number of dry months and annual rainfall. According to the classification utilized in Monk et al. (1997), the natural vegetation for East Timor would be various kinds of forest from evergreen in the mountains, especially the southern slopes, to thorn forest along the northern coasts. Because of the influence of the mountains on rainfall in the southern part of East Timor, by the 1950s rainforest originally
24
occurring on the south escarpment of the Fuiloro limestone plateau had been extensively replaced by secondary forest (Felgas 1956; van Steenis, un-publicized. in Monk et al. 1997:234). All land would be covered by different types of forest. Savanna and grassland are assumed to be secondary vegetation (Monk et al. 1997:197). This vegetation distribution would be before the indigenous people or the Portuguese began to occupy the land. Monsoon forest, one of the most sensitive and vulnerable of the tropical forest formations, is easily lost. The original monsoon forests of the dry regions have been extensively replaced by savanna and grassland. Generations have repeatedly burnt the dry forests for hunting and to accommodate shifting cultivation. (Monk et al. 1997:202) When these forest types are disturbed, principally by burning, then secondary vegetation, savanna or grasslands emerge. Figure 2.15 indicates there are very few areas of forest left. Deforestation is not a phenomenon confined to the eastern part of the island. When Crippen International carried out a detailed survey of forests in West Timor, it found that the majority of this part of the island was also covered with savannas and grasslands (Crippen International 1980 vol.14 - Forestry). It is also worth noting that when RePPProT used Landsat images from 1972 to 1986 to update aerial photos and coverage estimates, there were no aerial photos available for East Timor. Official numbers exist for the location and distribution of forest types on East Timor but these are of uncertain accuracy because of both their source and their age. Up-to date information gathered from remote sensing satellites or aerial photography, and actual in-thefield observations will be of critical importance. Monk et al. (1997: 211) concludes: "The accuracy of historical data available for East Timor is even more difficult to assess as no official survey seems to exist.” Felgas (1956) quotes estimates by Ruy Cinatti, the head of the Portuguese Timor Agricultural 17 and Veterinary Technical Department indicating that there were 74 km2 of mangroves; 2149 km2 of primary forest and 2646 km2 of savanna and grassland. This suggests that closed forest cover in East Timor rose from 16 percent in the 1950s to 29 percent in the 1980s. It is, however, not likely that such extensive reforestation occurred either naturally or through human activity. This casts doubt on any forestcover figures for East Timor. Scrub forest, savannas, and grasslands areas now make up as much as three
25
fourths of the land. Various grasses, xerophytic shrubs in the driest areas, and other shrubs are present including evergreens, small trees, and vines interspersed with stands of casuarina, eucalyptus, bamboo, acacia, or even palms. (Metzner 1977:104-114) Although much anecdotal information on the savannas exists, detailed quantitative descriptions are lacking. There are three ecological descriptions including two prepared by consultancy companies on West Timor (ACIL Australia Pty. 1986m; Crippen International 1980F).
Figure 2.14: Natural distribution of forest in East Timor
Note: A = Evergreen rain forest; B= Semi-evergreen rain forest; C= Moist deciduous forest; D= Dry deciduous forest;E= Thorn forest Source: Monk et al 1997: Figure 4.4
26
Figure 2.15 Actual forest
Source: Monk et al. 1997: Figure 4.5 Based on data and maps from Collins et al. 1991 with permission from N.M.Collins of World Conservation Monitoring Centre; The National Forestry Inventory Project, from the Directorate General of Forest Inventory and Land Use Planning and Information System Development Project for the Management of Tropical Forests; RePPProT 190b; K.A. Monk pers. obs.
The main consequences of deforestation are loss of genetic resources and increased risk of erosion and flash floods resulting from bare hillsides. Even before the era of Portuguese colonization, the original forest area of East Timor was shrinking as agriculture expanded through plantations or household production. Particularly in a landscape not endowed with fertile soils and regular and bountiful rainfall, the productivity of newly cleared lands quickly falls and farmers are forced to burn and clear new lands. Particularly in a landscape not endowed with fertile soils and regular and bountiful rainfall, the productivity of newly cleared lands quickly falls and farmers are forced to burn and clear new lands. If this occurs before the soil is entirely exhausted, the area will quickly return to a secondary forest lacking the species and complexities of the primary forest. In 1994, the GOI estimated actual land use (Table 2.1). The term “light forest lands” is used for much of the shrub or savanna. Saldanha, (1999) describes a forest component distinct from the majority shrubs (Table 2.2). As many as 70,000 hectares of forest were burned in the last decade by official estimates but some analysts believe that the real number is higher (Gomes 1999; 65). There is not adequate information on the actual extent and conditions of the various forests and forest types given the deforestation that has occurred in recent years. From the time of the first settlers on the island there has been shifting cultivation with negative but not 27
disastrous consequences. However, in recent years with the high increase in population in certain areas, there is increased pressure on the land. Many Timorese have been displaced to more marginal lands and their former lands occupied by migrant farmers whose practices may not be adapted to Timorese conditions. The situation has been exacerbated by deforestation, which has become more substantial during the last three decades. One of the country’s most valued forest resources is sandalwood which has now been reduced to just a few stands due to of over-exploitation. Another problem is that many rural communities rely on selling wood for fuel as source of family income and as a result, have contributed to deforestation (Figures 2.14, Figure 2.15 and Figure 2.16).
Figure 2.16 Firewood cut by community as a source of income and used for cooking.
Tabel 2.1 Land use in East Timor, 1994, Indonesia government estimated Land use
Human settlement Irrigated rice field Non-irrigated rice field Plantation Mixed framing Light forest Bush land Lakes, ponds, swamps Critical land Others Source: Brahmana and Emmanuel, 1994
%
1 3 3 3 2 76 9 0 0 1
Table 2.2 Land Use, Alternative estimation Land use %
Village Rice paddies Rain fed paddies Plantation rice paddies Mix plantation Homogeneous mix Shrubs Forest Swamps, lakes Roads, rivers 1 3 4 1 1 8 81 1 0 1
Source: Saldanha, 1999
East Timor is a comparatively small but mountainous territory, extending roughly 300 km in length and 100 km at its widest point. Estimates of the extent of forest cover over East Timor are notoriously variable. One respected study using LandSat imagery established a figure of 41 per cent for the eastern half of the island, with just 29 per cent as closed forest; this figure was adopted by the Indonesian government, which recorded forest cover as 40.6 per cent.3 These totals cover a wide range of forest types, including predominantly open and mixed savanna along the drier northern coast and hinterland, extensive eucalyptus and moist upland forests in the central highlands and semi deciduous monsoon and tropical lowland forest blocks along the southern coast and hinterland (Figure 2.19)
30
Figure 2.19 Category of land cover in East Timor
(Modified version of map produced by GIS unit, Ministry of Agriculture and Fisheries and Japan International Cooperation Agency [JICA], Dili, East Timor,2001)
2.2 Literature Review
A slope failure (i.e., landslide, surface failure, debris flow, rockfall and erosion) is define by Cruden (1991) for the working party on world slope failure inventory, as “a movement of a mass of rock, earth or debris down a slope”. Varnes (1978) indicated that slope movement would be a better comprehensive term as it does not infer process. His definition is “a downward and outward movement of slope forming materials under the influence of gravity”. In both the mining and civil aspects of engineering, slope failures can take lives and negate all the hard design and development processes involved in completion of a project. Slope failures can occur at any time of the year and sometimes can happen without any obvious warning signs. They can range from sinkholes to rockslides or avalanches. There are many effective ways to prevent of slope failures but uncertainties about surrounding environmental conditions that may cause a slope failure must be investigated to find the proper way to handle the potential problem.
31
From the aerial photo investigation in the study area, the slope failures were mainly landslide and surface failure, and most landslide features is subdivides in recent and older landslides. Cruden and Varnes, 1996 defined that Landslide described as “recent” have distinct features, clearly define boundaries and have moved in the past several years. They include active, suspended, and dormant earth flows and earth slides. Older landslides have hummocky topography, muted features, and indistinct boundaries. This category includes dormant, relic, and ancient earth flows and earth slides. Data on recent and older landslides have been used to develop the landslide hazard analysis in this study. These landslides are predominantly shallow failures with basal failure planes in the soil or weathered bedrock. Although deeper earth and rock slides also occur, such deep landslides overlap areas with shallow landslides. Slope failures have caused large numbers of casualties and huge economic losses in hilly and mountainous areas of the world. In tropical country like East Timor where heavy rainfall occasionally occurred and high temperatures around the year, cause intense
weathering and formation of thick soil and weathered rock profile. With these set of climate and geological condition, combined with other causative factors, slope failure is one of the most destructive natural disasters in East Timor. Each year, a number of major slope failures were reported in East Timor, involving fill and cut of natural slopes, which results in death of people and have posed serious threats to settlements and structures that support transportation. Most of these slope failure occurred on natural slope and cut slopes or embankments alongside roads in mountainous areas. Richard Dikau at al. (1996) stated as the probability of slope failure changes, due to changing climate or increasing human activity it becomes more important to recognize the potential event as well as geomorphologies and geology and these can be catalogued, classified and mapped. A primary task, therefore, is to develop a manual of such indicators and mapping techniques, providing a basic understanding to slope failure recognition. However, potential sites that are slope failure-prone should therefore be identified in advance to reduce such damage. In this regard, actual condition, distribution and characteristics of slope failure will be known to provide much of the basic information essential for hazard mitigation through proper project planning and implementation.
32
Slope failure hazard was defined by Varnes (1984) as the probability of occurrence of a potentially damaging slope failure phenomenon within a specified period of time and within a given area. The factors that determine the slope failure hazard of an area may be grouped in two categories: (1) the intrinsic variables that contribute to slope failure occurrence, such as geology, slope inclination angle, slope aspect, elevation, soil geotechnical properties, vegetation cover, and a long-term drainage patterns; and (2) the extrinsic variables that tend to trigger slope failure occurrence, such as heavy rainfall, and earthquakes (Wu and Sidle 1995); Atkinson and Massari 1998). Obviously, the probability of slope failure occurrence depends on both the intrinsic and extrinsic variables. However, the extrinsic variables may change over a very short time span, and are thus very difficult to estimate. If extrinsic variables are not taken into account, the term of “actual condition, characteristics and distribution slope failure” could be employed to define the likelihood of occurrence of a slope failure event. The spatial distribution of the intrinsic variables within a given area determines the spatial distribution of relative slope failure occurrences in that region (Carrara and others 1995). A variety of techniques, such as heuristic, statistical, and deterministic approaches, has been developed to predicted probabilities of slope failure occurrences. In heuristic approaches, expert opinions are used to estimate slope failure potential from data on intrinsic variables. They are based on the assumption that the relationship between probability of slope failure occurrences and the intrinsic variables are known and are specified in the model of analysis. A set of variables are then entered into the analysis model to estimate probability of slope failure occurrence (Niemann and Howes 1991; Anbalagan 1992; Pachauri and Pant 1992; Atkinson and Massari 1998). One problem with the heuristic models is that they need long-term information on the slope failure and their causal factors for a similar geo-environmental condition or for the same site, and these are, in most cases, not available. Statistical analysis models involve the statistical determination of the combinations of variables that have led to slope failure occurrence in the past. Quantitative or semi-quantitative estimates are then made for areas currently free of slope failure, but where similar conditions exist. Logistic regression analysis is one of the multivariate statistical analysis models, is useful for predicting presence or absence of a outcome based on values of a set of predictor
33
variables.Klein-baum (1994), stated that the advantage of logistic regression analysis over other multivariate statistical technique, including multiple regression analysis and discriminant analysis, is that the dependent variable can have only two values an event occurring or not occurring, and that predicted values can be interpreted as probability because they are constrained to fall in the interval between 0 and 1. Mark and Ellen (1995) used logistic regression to predict the sites of rainfall induced shallow landslides that initiate debris flows in San Mateo County, California. In this study, the dependent variable is a binary variable representing of the slope failure or un-failure of slopes. Recently, there were studies on slope failures hazard evaluation using GIS, and many of these studies have applied probabilistic methods (Rowbotham and Dudycha 1998; Guzzetti et al. 1999; Jibson et al. 2000; Luzi et al. 2000; Parise and Jibson 2000; Rautelal and Lakhera 2000; Baeza and Corominas 2001; Lee and Min 2001; Temesgen et al. 2001; Clerici et al. 2002; Donati and Turrini 2002; Lee et al. 2002a,b; Rece and Capolongo 2002; Zhou et al. 2002; Chung and Fabbri 2003; Remondo et al. 2003; Lee and Choi 2003c; Lee et al. 2004b). The logistic regression method has also been applied to slope failure hazard mapping (Atkinson and Massari 1998; Dai et al. 2001; Dai and Lee 2002; Ohlmacher and Davis 2003). There are other methods for hazard mapping, such as the deterministic (or safety factor) approach used by Gokceoglu et al. (2000); Romeo (2000); Carro et al. (2003); Shou and Wang (2003), and Zhou et al. (2003). Fuzzy logic and artificial neural network methods have also been applied in various case studies (Ercanoglu and Gokceoglu 2002; Pistocchi et al. 2002; Lee et al. 2003a, b; Lee et al. 2004a). To represent the distinction quantitatively, logistic regression analysis were used. For this analysis, the calculated and extracted factors were mapped to a 30-m-resolution grid. The raster data were converted for the statistical program used. Then, using the logistic regression analysis models, the spatial relationships between the slope failure location and each slope failure-related factor, such as geology, vegetation cover, slope gradient (i.e., slope inclination angle), elevation, landscape topography and slope aspect (i.e., direction), were analyzed in the statistical program, and a formula of slope failure occurrence possibility was extracted using the relationships. The formula was used for calculating the probability of slope failure
34
occurrence, which was mapped to each grid cell. Finally, the susceptibility and probabilities occurrence map was verified using known slope failure locations and success rates were calculated (Chung and Fabbri 1999) for quantitative verification. In this study, GIS software, ArcView 3.3 and statistical software, SPSS 10.0, were used as the basic analysis tools for spatial management and data manipulation.
35
CHAPTER III ACTUAL CONDITION, CHARACTERISTICS AND DISTRIBUTION OF SLOPE FAILURE IN EAST TIMOR
3.1 Introduction
According to aerial photograph investigation, the actual slope failure distribution will established in this study are landslide, surface failure and mix of landslide and surface failure. From the aerial photo investigation in the study area most landslide features is subdivides in recent and older landslides (Figure 3.1, Figure 3.2, and Figure 3.3). Cruden and Varnes, 1996 defined that Landslide described as “recent” have distinct features, clearly define boundaries and have moved in the past several years. They include active, suspended, and dormant earth flows and earth slides. Older landslides have hummocky topography, muted features, and indistinct boundaries. This category includes dormant, relic, and ancient earth flows and earth slides. Data on recent and older landslides have been used to develop the landslide hazard analysis in this study. These landslides are predominantly shallow failures with basal failure planes in the soil or weathered bedrock. Although deeper earth and rock slides also occur, such deep landslides overlap areas with shallow landslides. In East Timor, slope failures are common in the mountainous areas and in many regions. The high occasional rainfall, steep slopes, high weathering rates and slope material with a low shear resistance or high clay content are often considered the main preconditions for mass movement in East Timor, turning it in an inherent susceptible area of slope failure. The main causal factors for slope failure in highlands, as found in international literature, can be divided into preparatory and triggering causal factor (Glade and Crozier, 2004). Preparatory causal factors, i.e. factors making slopes susceptible to movement over time without actually initiating it, often reported for this region include the increasing population pressure with slope disturbance and deforestation as a consequence and the reduction in material strength by weathering. Triggering causal factors on the other hand can be seen as external stimuli responsible for the actual initiation of mass movements. The triggering causal factors in the region can be earthquakes, excessive rainfall events and human disturbance such as slope excavation and terracing, inconsiderate irrigation and water leakage.
36
Figure 3.1 Older landslide topography in East Timor
(Source:Prof. H. Kazama documentation,August 2005)
Figure 3.2 Older landslide topography in East Timor
(Source:Prof. H. Kazama documentation, August 2005)
37
Figure 3.3 Recent landslide topography in East Timor
(Source:Prof. H. Kazama documentation,August 2005)
In many regions of the East Timor highlands, a clear insight into the local causes for mass movement is lacking. Therefore, the search for region-specific solutions is hampered. In East Timor, slope failure i.e., landslides, surface falure, erosion, and rock fall are common in the mountainous areas of all districts but so far no systematic scientific research has been conducted on this topic. Western part of East Timor, situated on the southwestern foot slopes of the mountainous of Tatamailau (Ainaro), Sabiria (Aileu), Harupai (Ermera), Atubuti (Bobonaro) is the most sensitive area for slope failure in East Timor. As a broad outline, the watersheds of East Timor can be divided into two areas; northern and southern. Of the many rivers in this study site, the following rivers flow all year round; the Bulobo, Marobo, Malibaka and Nunura rivers in Bobonaro , Gleno river in Ermera,ladibau in Hatolia, Aimera in Cailaco, Belulik in Ainaro and Mola in Zumalai. Mass movements associated with intense rainstorms are reported to have occurred sporadically in mountainous since the twentieth century but the increase in fatalities and losses as a consequence of the enormous population growth draws attention to the phenomenon nowadays. By studying the causal factors for slope failure in these mountainous areas of western part of East Timor, this study tries to contribute to the restricted knowledge
38
on slope failure in East Timor. After a brief introduction of the study area and the spatial distribution and characteristics of its landslides, the preconditions, preparatory and triggering causal factors for mass movement affecting slope failure will be discussed with attention to their spatial variation.
Figure 3.4 Recent landslide occurred on cut slopes alongside road in East Timor (Source:Prof. H. Kazama documentation,August2005)
Figure 3.5 Surface failure on hill slopes of mountainous in East Timor
(Source:Prof. H. Kazama documentation, August 2005)
39
3.6 Surface failure on hill slopes of mountainous in East Timor
As state in above, According to aerial photograph investigation, the actual slope failure distribution will established in this study are landslide, surface failure and mix of landslide and slope failure. The distribution and characteristics of slope failure in East Timor has shown in Table 3.1 to Table 3.13 and Figure 3.7 to Figure 3.18 .
3.2 Characteristics and Distribution of Slope Failure in East Timor
Two types of slope failures were identified based on aerial photograph and topography map are surface failure and landslide. This study covers four of topographic and air photograph sheets, and 506 number of slope failure and 506 number of unfailure slope with area 1448 km2 are already mapped in the region. A significant number of these slope failures were reactivations of old slope failures. There are density of the distribution of slope failure in East Timor will describe in Table 3.1, and show that landslide and slope failure are common in East Timor with highest density in Bobonaro , Cailaco and Zumalai site, and moderately density in Hatolia and Atsabe site and the lowest density in Maliana, Ainaro and Hatobuilico study site. Types of slope failure occurred in East Timor dominantly by landslide 56% with density 0.28 Number/km2 ,surface failure are 37% with density 0.16 Number/km2 and mix of landslide and surface failure are 7% with density 0.08 Number/km2.
40
Table 3.1 Density of slope failure in study site Site Area (Km2) Landslide Density (N/km2) Type and density of slope failure Surface failure Density (N/km2) Mix Density Total (N/km2) density (N/km2) Bobonaro Cailaco Zumalai Atsabe Maliana Ainaro Hatolia 259 88 150 89 75 385 255 88 93 42 21 16 5 13 5 283 0.34 1.06 0.28 0.24 0.21 0.01 0.05 0.03 0.28 61 30 33 12 15 18 7 13 189 0.24 0.34 0.22 0.13 0.20 0.05 0.03 0.09 0.16 18 10 0 6 0 0 0 0 34 0.07 0.11 0 0.07 0 0 0 0 0.08 0.64 1.51 0.50 0.44 0.41 0.06 0.08 0.12 0.35
Hatobuilico 147 Total 1448
3.2.1 Lithology
Lithology exerts a fundamental control on the geomorphology of a slope failure. The nature and rate of geomorphological processes, including the slope failures process, is partially on the lithology and weathering characteristics of the underlying materials. Based on the East Timor geological map with 1:350,000- scale solid and superficial geological map covering the study area were used to identify the geological groups, with each group comprising units of broadly similar lithology. For analysis, the groups were further reclassified into three categories of geological materials with similar engineering properties. They are: sedimentary rocks with a few volcanic and igneous rocks, Sedimentary rocks and littoral deposit and Sedimentary rocks and a few metamorphic rocks and volcanic rocks. Detail of lithology analysis for this study are categories in five dominant lithology based on geological map with attributes of geology in study area, namely: Sedimentary rocks (Sr), Littoral deposit rocks (Ld), Metamorphic rocks (Mr) Igneous rocks (Ir) and volcanic rock (Vr). Description of geological structures in each study area has shown in Table 3.2 and Table 3.3, and Figure 3.7. It can be seen that many number of slope failure relatively highest 41
density in sedimentary rocks and littoral deposit rocks and lowest in igneous rocks, metamorphic rocks and volcanic rocks. The Tatamailau, Sabiria , Harupai , Atubuti mountains hills located in the study area are several hundreds to two thousands meters high in elevation is the most sensitive area for slope failure. There are composed of Miocene to Pliocene sedimentary rocks such as sandstones, limestones and siltstones, in part associated with a small amount of Mesozoic volcanic rocks. These sedimentary rocks and associated volcanics make up the Bobonaro and Lolotoe formation that are arranged chronological in other. Like many of the mountain ridges in the western region, these mountains hills correspond to folded structures of anticlines and synclines, and are elongated toward the north – northeast direction following the fold axes.
Table 3.2 Description of geological structures in each study area
Study Area
Bobonaro Tertiary
Age
Pliocene and Miocene
Lithology
Bobonaro Complex and Lolotoe Formation Bobonaro Complex and Viqueque Formation Bobonaro Complex, Cablaci Limestone and Cribas Formation Bobonaro Complex, Cablaci Limestone and Wailuli Formation Sedimentary Rocks and a few of Volcanic and Igneous rocks Sedimentary rocks and littoral deposit.
General Lithological Description
Mainly composed by chaotic rock with scaly matrix and blocks of older rock ; doleritic lava, volcanic breccia, tuff, green sandstone, metagabro a,d metadiorite Mainly composed by chaotic rock with scaly matrix and blocks of older rock; alternating conglomerate, conglomerate sandstone, sandstone, a lot of foraminifera in marl and sandstone Mainly composed by chaotic rock with scaly matrix and blocks of older rock ;contains marine foraminifera, Clastic limestone, crustaline, fine coarse grained, shale, claystone, siltstone and micaceous quarts sandstone Mainly composed by chaotic rock with scaly matrix and blocks of older rock ;contains marine foraminifera, and also dominant by sandstone, shale siltstone and limestone.
Cailaco
Tertiary
Pliocene and Miocene
Zumalai
Palaozoic and Mesozoic
Miocene And Permian
Sedimentary rocks and littoral deposit
Atsabe
Tertiary and Mesozoic
Miocene and middle to Jurassic
Sedimentary rocks and littoral deposit
42
Table 3.2 ( …continued) Maliana Tertiary
Late to Pleistocen e and Miocene
Ainaro
Quaternar y and Mezosoic
Early Miocene and middle to Jurassic
Viqueque Formation , Bobonaro Complex and Ainaro Formation Bobonaro Complex, Ainaro Formation and Cablaci Limestone Wailuli Formation and Aileu Formation Wailuli Formation , Lolotoe Formation anf Dartollu Limestone
Sedimentary rocks and littoral deposit
Mainly composed by chaotic rock with scaly matrix and blocks of older rock; alternating conglomerate, conglomerate sandstone, sandstone, a lot of foraminifera in marl and sandstone; mixture sand and clay Mainly composed by chaotic rock with scaly matrix and blocks of older rock ;contains marine foraminifera, mixture sand and clay
Sedimentary rocks and littoral deposit
Hatolia
Mezosoic
Early Jurassic and late to Jurassic Early Jurassic and late Eosin
Hatobuilico
Mesozoic
Sedimentary rocks and a few of Metamorphic rocks and volcanic rocks Sedimentary Rocks and a few of metamorphic rock and volcanic rocks
Dominanted by sandstone, shale silttone, limenstone; phylite, schist, amphibolite, slate, metasandstone, sandstone, shale and a few of volcanic rocks Dominanted by sandstone, shale silttone, limenstone; doleritic lava, volcanic breccia, tuff, green sandstone, metagabro a,d metadiorite
Failure
110 100 90 80 70 60 50 40 30 20 10 0
Num ber of Slopes Failure
Maliana
Failure
N u m b e r o f s lo p e
SR
LR
IR
Lithology
MR
VR
1
LR
2
IR
3
MR
4
VR
5
6
Bononaro Malia na
90
Lithology of Unfailure Slopes
Lithology
Ca ila co Aina ro
Zuma la i Ha tolia
Atsa be Hatobuilico
SR: sedimentary rocks
N m e o U -f il reS p s u br f n au lo e
80 70 60 50 40 30 20 10 0 SR LR IR Lithology MR VR
LR : Littoral deposit rocks IR : Igneous rocks MR: Metamorphic rocks VR: Volcanic rocks
Unfailure
Figure 3.7. Lithology
3.2.2 Vegetation
Many studies have revealed a clear relationship between vegetation cover and slope stability, especially for shallow landslides. Parameters, such as cohesion, internal friction angle, weight of the soil and pore-water pressure, all tend to be substantially modified by the presence of vegetation. Vegetation can both enhance effective soil cohesion due to root matrix reinforcement and soil suction or negative water pressure through evapotranspiration and interception. According to Selby (1993), tree-covered hillslopes are thought to increase soil shear strength by about 60% depending on the tree type. Mehrotra et al. (1996) show that landslide activity increases by up to 15% in those places where the original vegetation cover
44
has been removed or altered. In order to correlate vegetation cover with other factors affecting slope failure, a vegetation classification was carried out in this study. The intention was to discriminate between different vegetation cover types. Indeed, many studies have pointed out that the degree of soil stability provided by vegetation decreases in the following order: trees, shrub, grass and bare soil (Coppin and Richards, 1990). The presence or absence of thick vegetation may affect slope failure. Due to the characteristics of the study area, where land cover is not homogenous with the presence of natural vegetation and for the purpose of this study and based on aerial photograph interpretation these vegetation types were then simplified in to four types, namely woodland or high tree (HT), scrublands or low tree (LT), grassland (G) and bare land or no vegetation (NV). To assess the effect of vegetation cover on the slope failure, the correlation between vegetation type and number of slope failure is shown in Table 3.4 and Figure 3.8. It can be seen that the number of slope failures on bare land and grassland is highest, and is lowest on woodland and scrubland. This is in agreement with the fact that vegetation cover, especially of a woody type with strong and big root systems, help to improve the stability of slopes. Other cause of this agreement is many Timorese have been displaced to more marginal lands
and their former lands occupied by migrant farmers whose practices may not be adapted to Timorese conditions. The situation has been exacerbated by deforestation, which has become more
substantial during the last three decades. Another problem is that many rural communities rely on selling wood for fuel as source of family income and as a result, have contributed to deforestation. Under such conditions, intense bombardment of the soil surface by rain can quickly break down soil-organo aggregates, thus permitting slope failure.
Table 3.4 Distribution of vegetation
Study Area Bobonaro Cailaco Zumalai Atsabe Maliana Ainaro Hatolia Hatobuilico Total Types of vegetation cover and Number of Slope failures High Tree Low Tree Grassland No Vegetation 8 21 76 62 0 10 50 73 7 33 20 15 0 8 25 6 1 8 10 12 4 5 10 4 0 6 11 3 1 3 2 12 21 94 204 187 Total 167 133 75 38 31 23 20 18 506
45
All Site
Un-failure 300 Failure
80 Number of Slopes Failure
70 60 50 40 30 20 10 0
High Tree Low Tre e Grassland No Vege tation
Failure
Vegetation
HT
1
LT G Vegetation
2
3
NV
4
5
Bononaro Maliana
Vegetation Cover of Unfailure Slopes
Cailaco Ainaro
Zumalai Hatolia
Atsabe Hatobuilico
HT : High tree
N m e o U -F ilu slo e u b r f n a re p s
90 80 70 60 50 40 30 20 10 0
LT : Low tree G : Grassland NV: No vegetation
Unfailure
Figure 3.3 Vegetation
High Tr e e
Low Tr e e
Gr ass land
No Vege tation
Vegetation
Figure 3.8 Distribution of vegetation
3.2.3 Inclination angle of slope
Slope is the angle formed between any part of the surface of the earth and a horizontal datum. It is the means by which gravity induces stress in the slope rocks, flux of water and other materials; therefore, it is of great significance in hydrology and geomorphology. In fact, slopes affect the velocity of both surface and subsurface flow and hence soil water content, soil formation, erosion potential and a large number of important geomorphic processes. It has been widely shown that landslides tend to occur more frequently on steeper slopes (McDermid and Franklin, 1995; Cooke and Doornkamp, 1990). Slope failure tends to
46
increase with slope angle but when the slope becomes near vertical, landsliding is scarce or absent altogether. The reason is the lack of soil development and debris accumulation in such topographic conditions (Selby, 1993; Derruau, 1983). A long slope may include sections that can be affected by large movements originating further up the hills slope. The estimation of the slope angle for this study was implemented using by topographic map investigation in which slope is considered as the change in elevation over a fixed distance. Inclination angle of slope is an essential component of slope stability and an important control on slope failure. As slope inclination angle increases, the level of gravitation-induced shear stress in the residual soil increases as well. Gentle hill slopes are expected to have a flow frequency of slope failures because of generally lower shear stresses associated with low inclination angle. In this study, inclination angle of slope has categories with ranges: 60 – 120, 120 – 180, 180 – 240, 240 – 300, 300 – 360, 360 – 420 and 420 - 480. In regional slope failure (i.e., landslide and surface failure) susceptibility or hazard assessment, slope inclination angle in terms of slope failure activity in taken into consideration as an conditioning factor(Y. Duman et all, 2006). In the study site, the distribution number of slope failure occurred with inclination angle of slope has shown in Table 3.5 and Figure 3.10 . It can be seen that examination of the distribution of number of slope failure with corresponding slope inclination angle ranges shows that most of slope failures with inclination angle do have ranges increase in the 120 – 300 and gradually decrease in the ranges 60 – 120 and 300 – 480. This is refection that steep natural slope with outcropping bedrock and hence much higher shear strength may not susceptible to shallow landslide.
47
Table 3.5 Distribution inclination angle of slope Site
60~120 1. Bobonaro 2. Cailaco 3. Zumalai 4. Atsabe 5. Maliana 6. Ainaro 7. Hatolia 8. Hatobuilico Total 17 19 5 9 3 0 2 0 55
Inclination angle ( o )
Nm u ber of U n-fa re s ilu lope s
60 50 40 30 20 10 0
Unfailure
6~12
12~18
18~24
24~30
30~36
36~42
42~48
Slope Inclination angle
(o)
Figure 3.9 Distribution of inclination angle of slopes
48
3.2.4 Direction of Slope
Aspect is often expressed as a compass direction .The aspect of slope failures ( i.e., the direction) has the potential to influence its physical properties and its susceptibility to slope failure. The processes that may be operating include exposure to sunlight, drying winds, rainfall, earthquake and groundwater behavior. Although, the relation between slope aspect (i.e., direction) and mass movement has long been investigated, no general agreement exists on the effect of the aspect on slope failure occurrence (Carrara et al. 1991). However, slope aspect is related to the general physiographic trend of the area and/or the main precipitation direction, and direction of the slope failure is roughly perpendicular to general physiographic trend. Several researchers have reported a relationship between slope orientation and landslide occurrence. For example, DeGraff and Romesburg (1980) point out that, to some extent, aspect gathers the structural and organic basic conditions of a slope including fault planes and climatic factors, respectively. It is reported by Lineback et al. (2001) that larger numbers of landslides occur in the wetter north-facing aspects than in drier, south facing aspects. Marston et al. (1998) report a similar finding and highlight that soil exposed on south-facing slopes are subject to several wetting and drying cycles, thus increasing landslide activity in the Himalayas. The distribution of direction among the aerial photograph and topography maps show that the general physiographic trend of the study site is East to West and an important part of slope failures in most of study area was highest number on North – Northeast and Northwest facing slope, indicating that natural terrain slope failures is more common on these slopes. The frequency of slope failures was lowest on those slopes facing, south and west, while the frequency of slope failures remained moderate on the east, southeast and southwest facing slopes. The distribution of slope failures direction has shown in Table 3.5 and Figure 3.10.
49
Table 3.6 Distribution of direction of slope
Direction and number of slope failure Site
1. Bobonaro 2. Cailaco 3. Zumalai 4. Atsabe 5. Maliana 6. Ainaro 7. Hatolia 8. hatobuilico Total
N
42 19 4 3 0 0 1 5 74
NE
49 53 28 4 0 2 0 3 139
E
10 21 5 0 9 0 0 0 45
SE
25 11 11 0 7 1 5 9 69
S
5 0 0 0 2 0 7 0 14
SW
16 0 17 6 10 0 3 0 52
W
12 0 3 0 0 3 4 0 22
NW
8 29 7 26 3 17 0 1 91
Total
167 133 75 39 31 23 20 18 506
All Site
Bononaro
Direction of Slopes Failure Cailaco Ainaro Zumalai Hatolia Atsabe Hatobuilico
Un-failure
Failure
60
Maliana
Number of Slopes Failure
150 N u m b e r o f s lo p e 100 50 0 0 N NE2 E 4 SE
S
50 40 30 20 10 0 N NE E SE S SW W NW
Failure
Direction
6 SW W
8 NW
10
Bononaro Maliana
45 40 Num ber of Unfailure Slo pes 35 30 25 20 15 10 5 0 N
Direction of Unfailure Slopes
Cailaco Ainaro
Zumalai Hatolia
Atsabe Hatobuilico
Direction
Unfailure
N
: North
S
: South
NE : Northeast E : East SE : Southeast
SW : Southwest W : West
NW : Northwest
NE
E
SE
S
SW
W
NW
Direction
Figure 3.10 Distribution of direction of slope
50
3.2.5 Landscape topography
Landscape topographic represents a theoretical measure of the accumulation of flow at any point within a river basin. The landscape topography can be thought of as an abstract parameter to be used as a basis for estimating the local soil moisture status and thus slope failure areas due to surface topographic effects on hydrologic response. Soil moisture plays an important role in slope instability, particularly for shallow landslides and surface failure. Water operation may be through the accumulation of rainfall, as an agent of weathering, hydration of fine soils (i.e. clayey soils), undercutting of slopes and spontaneous liquefaction. In fact, according to Lamb (1996), hallow landslides can occur on slopes when water from precipitation infiltrates the soil and eliminates the suction and lowers the apparent cohesion. Modeling water in soil slopes in extensive areas is a difficult task as soil water content is governed by a number of factors, some of which are estimated from laboratory tests. Since landscape topographic is intended to represent the topographic control on soil wetness, it is considered in this study as an indirect measurement of soil water content. According to the topographic map investigation in this study, the landscape topographic index three variables are required, which are namely Valley (V), Ridge (R) and Flat (F). We interpret the flat, ridge and valley topography as the result of subaerial erosion. Most of study areas located in hilly lands of mountainous landscape, while there is covered with loose soil mantle of variable thickness. In such a slope failures (i.e., landslide and surface failure) of soil mantle ridge, valley and flat topography, shallow slope failures typically only involved the soil mantle and commonly occurred at or near the soil- bedrock boundary. The distribution of landscape topography of slopes has shown in table 3.7 and figure 3.11. It can be seen that landscape topography of study site where slope failure occurred with ridge and valley topography that often dominates shallow landslide and surface failure location. This assessment indicated that most of slope failure occurred in the hilly and mountainous terrain of ridge and valley of study site. Overall distribution of the slope failure was determined
primarily by the intensity of ground shaking; the local density of slope failure reflected differing local geology. Where ridge and valley tops have been severely fractured, abundant landslides may develop later when saturated with water.
51
Tabel 3.7 Landscape topography
Site 1. Bobonaro 2. Cailaco 3. Zumalai 4. Atsabe 5. Maliana 6. Ainaro 7. Hatolia 8. hatobuilico Total Landscape topography and number of slope failure Valley Ridge Flat 51 106 10 90 75 25 13 23 11 18 306 30 0 14 18 0 9 0 177 13 0 0 0 0 0 0 23 Total 167 133 75 39 31 23 20 18 506
3.2.6 Elevation
Some researches have found that landslide activity, within a specific basin, occurs at certain elevations (Greenbaum et al., 1995; Jordan et al., 2000), the relationship between landslide activity and elevation is still unclear, hence it requires further studies. Pachauri and Pant 1992 report that the elevation is a good indicator of slope failure susceptibility to occurred. However, it is well known that elevation influences a large number of biophysical parameters and anthropogenic activities. In turn, these conditions are likely to affect slope stability and generate slope failure (Vivas, 1992). Elevation also affects soil characteristics significantly. Ochoa (1978) relates the influence of elevation on physical–chemical soil properties in the Cordillera de Me´rida. He argues that soil texture varies with elevation, as the grain size increases with the altitude. Although, in the study site, there is a considerable difference between the lowest and the highest elevation values has shown in Table 3.8 and Figure 3.12. It can be seen that hill slopes between 200m to 800m in elevation had frequencies of slope failure that were 2 times greater than those on hill slopes that are less than 1400m to 2100m and greater than 800m to 1400m in elevation. At intermediate elevations there are mountain summit, which is more prone to landslide that are usually characterized by weather rocks, and the shear strength of these is much higher. At very low elevations, the frequency of slope failure is low because the terrain is gentle, and is covered by residual soil, and higher perched water table will required initiating slope failure (F.C Day et al. 2000). Table 3.8 Distribution of elevation
Site 200~500
1. Bobonaro 2. Cailaco 3. Zumalai 4. Atsabe 5. Maliana 6. Ainaro 7. Hatolia 8. hatobuilico Total 29 68 39 6 10 4 7 0 163
3.2.7 Slope width
Table 3.9 and Figure 3.13 shows that most of landslide occurred in all study sites frequently do have width with ranging from 31 m to 90 m. Some site like Bobonaro, Cailaco, Atsabe, Maliana, hatolia and Hatobuilico study site, landslide occurred moderately with ranges greatest than 90m to 150m, and especially Bobonaro and Cailaco site, a few number of landslide occurred with ranges greatest than 150m. Table 3.10 and Figure 3.14 shows that most of surface failure occurred in all study sites frequently do have width with ranging from 31m to 150m, especially in case of Bobonaro site a few number of surface failure occurred have ranges greatest than 150m to 360m. Table 3.11 and Figure 3.15 shows that in some site
54
like Bobonaro, Cailaco and Atsabe study site, most number of surface failure occurred in older landslide area do have width with ranging 31m to 120m. Based on aerial photograph and topographic maps investigation, most of slope failure occurred in grassland and bare land areas where lateral roots strength of grassland could not provide help to improve the stability of slopes. However, we assess that estimated slope failure width (i.e., width of landslide and surface failure), assuming that root strength acts primarily through a perimeter boundary. Reneau and Dietrich (1987) derived an expression relating landslide width to length by assuming that the soil was saturated and that its strength was composed of frictional term acting on a basal slide area and a root strength acting on the perimeter of the slide. In this study, we predict that slope failure width increases with decreasing root strength, its means that as larger masses of soil are needed to overcome resisting forces. Perhaps surprisingly, the drier the soil, the larger of slope failure mass and width, whereas the water table rise reduces the size needed for failure. The comparison of this result with field data suggests that slope failure size is controlled by the local patchiness of soil thickness, root strength and topographically-driven relative saturation.
3.2.8 Slope length
Slope length is the distance along a slope subject to uninterrupted overland flow, from of the point at which overland flow begins to where deposition starts, or where flow enters a well-defined channel (Wischmeier and Smith, 1978). This distance is computed on the horizontal and normal to the contours of the surface of the slope. It is, in fact, the horizontal projection of the slope distance, which is measured along the slope surface. In this study, slope length was estimated using by topography map investigation which generates real slope length values. Table 3.12 and figure 3.16 shows that most of landslide occurred in all study sites frequently do have length with ranging from 31m to 150m and a few numbers of landslides were occurred with length 180m to 210m. In some site as well as Bobonaro and Cailaco study site, a few number of landslide occurred with ranges greatest than 240m to 510m. Table 3.13 and figure3.17 shows that most of surface failure occurred in all study site frequently do have length highest with ranging from 31m to 90m and moderately number of surface failure were occurred with length 91m to 180m in Bobonaro, Cailaco and Zumalai study site, and a few number of surface failure occurred in Bobonaro , 58
Cailaco and Hatobuilico site do have ranging greatest than 180m. Carrara et al. (1995) argue that field and laboratory analyses show that slide density increases linearly with slope length up to a threshold value of about 500 m. However, slope length may be considered an important factor in landslide activity since longer slope lengths increase the potential of erosive agents to dislodge and transport materials downslope. Moreover, downslope water velocity is greater on longer slopes. The slope length is of paramount importance for the travel distance of materials.
CHAPTER IV ANALYZING METHOD
4.1 Logistic Regression Analysis
Landslide susceptibility evaluation involves a high level of uncertainty due to data limitation and model shortcomings (Zezere 2002). For this reason, the landslide researchers have considered different techniques for analyzing factors contributing of landslide occurrence and preparation landslide susceptibility maps. One of these techniques is statistical analysis. In this study, a multivariate statistical analysis in the form of logistic regression was used to analyzing the factors contributing of slope failure in East Timor. The fundamental principle of logistic regression is based on the analysis of a problem, in which a result measured with dichotomous variables (such as zero and one or true and false) is determined from one or more independent factors (Menard 1995). Logistic regression analysis is a multivariate technique that considers several physical parameters that may affect probability. Logistic regression can be used to determine the relation of slope failure occurrence and the related factors. The dependent variable (y) for this analysis is the failure or un-failure of a slope. Considering P independent variables, x1 , x 2 ,......, x p , affecting slope failure occurrences,
we define the vector X = ( x1 , x 2 ,......, x p ). In this study, the independent variables correspond to the classes of the independent variables categories in Table 1; each of these variables is binary, with values of 1 (failure) or 0 (un-failure). The reason we consider each class as an independent variable is that we are interested in detailed relationships of these classes and not just the relationships between the broader independent variables (or factors). The conditional probability that a slope failure occurs is represented by ( P( y = 1 X ) . The
logit of the multiple logistic regression models (Hosmer and Lemeshow, 2000) is: Logit(y)= b0 + b1 x1 + b1 x2 + ......b p x p (1)
Where b0 is the constant of the equation, and b1, b2,……,bp are the coefficients of variables x1 , x 2 ,......, x p . The probability P( y = 1 X ) can be expressed in the logistic regression model :
63
1
P( y = 1 X ) =
1+ e
−(b0 +b1x1 +b2 x2 +...,+bp x p )
(2)
where e is the constant 2.718 Assume that we obtain a sample of n observations (Xj,xj), j = 1,2, …, n, Xj= (x1j,x2j,…xpj), the yj is either 1 or 0, yj = 1 for a slope failure event, and yj = 0 for non event (un-failure). By fitting the binary logistic regression analysis using the sample observations, we estimate the logistic regression coefficient bi, i = 0, 1, ... , p. Based on this model, the probability of slope failure occurrence in the future can be estimated using Eq. (2). By examining the sign of a variable’s coefficient estimate, the effect of that variable on the probability of slope failure occurrence could be determined. A positive coefficient estimate indicates that the independent variable increases the probability of a slope failure, assuming that the other variables in the model are held constant. Another method that can be used to interpret the regression results and examine the significance of a variable in the model involves determining the influence ratio. The influence ratio is odds ratio of statistics used to assess the risk of a particular outcome if a certain factor is present. The influence ratio is a relative measure of risk, telling us how much more likely it is that some item of factor is exposed to the category under study will develop the outcome as compares to some item of factor who is not exposed. Odds are a way of presenting probabilities, but unless you know much about betting you will probably need an explanation of how odds are calculated. The influence of an event happening is the probability that event will happen divided by the probability that the event will not happen. By definition, the influence ratio is the ratio of the odds for variable xi = 1 (i = 1,2, ..., p) to the odds for xi = 0 (Hosmer and Lemeshow 2000). In slope failure analysis, the influence ratio approximates how much more likely it is for the slope failure to be present (y = 1)
among those events with variable xi = 1 than for those events with variable xi = 0. The influence ratio is computed by exponentiations the coefficient estimate for each dichotomous explanatory variable and it can be expressed: Influnce ratio = e b …………………………………………………….(3) 64
Where e = 2.718 and b is coefficient value. Using the logistic regression model, the spatial relationship between slope failure and the factors influence to slope failure was assessed. The spatial databases of each factor were converted use in the statistical package, and the correlations between slope failures were calculated. Though there were two cases, in the first case, only one factor was analysis. Besides, logistic regression mathematical equations were formulated for each case. Finally, the probability that predicts the possibility of slope failure was calculated using the spatial database, data from equation (1) and (2) with coefficient value of logistic regression of each factor. However, in the second case, all factors were used and logistic regression mathematical equations were formulated as shown in equation (2) and (4) for each factor. Mathematically, probabilities of the possibility of slope failure can be express: Zp = b0 + (Cl xLithology) + (Ci x Inclination angle)+(Cv x vegetation) + (Clt x Landscape topography) + (Cd x Direction) + (Ce x Elevation) ……………………………(4) Where Zp : probabilities of the possibility of slope failure Cl: coefficient of lithology, Ci: coefficient of slope inclination angle, Cv : coefficient of vegetation, Clt : coefficient of landscape topography, Cd : coefficient of direction, and Ce ; coefficient of elevation. The model of analyzing building involves five main steps: • • • • • Selection of variables based on a slope failure distribution analysis; Selection of statistically significant variables by a P-value significance test; Logistic regression analysis with those variables that passed the significance test; Logistic regression analysis with significant variables including the interaction terms; and Evaluation of the analysis results.
In the first step, a slope failure distribution analysis is used to pre-select the variables that are relevant for the regression. This analysis involves overlaying the variables of category of slope failure occurrences and the variables of category of a factor (such as sedimentary rocks), then calculating the percentage of coverage of the slope failure occurrence on each class for each input factor, such as slope inclination angle within elevation factor. By comparing the slope failure distributions, a preliminary ranking of the variables can be developed. Important variables will be considered in the following significance tests.
65
In the second step, the significance p-value of 0.05 is specified as the cut-off value to choose the variable for further analyses and > 0.05 is chosen as the value for elimination of insignificant variables. The variables that passed the significance test can be entered into the logistic regression analysis in the next step (SPSS 1999). After the steps of pre-selection and significance test, some independent variables will be out of the original independent variables were selected for the regression analysis. In the third step, the model is checked for its goodness of fit by entering a variable or removing a variable. Following the SPSS procedures, iteration of some variables are preferred to obtain optimal analysis. The final suitable logistic regression analysis is based on the variables presented in the final step of the statistical calculation in the SPSS program, and the regression coefficients are obtained. In the fourth step, the interaction terms representing the interactions among variables are entered into the logistic regression analysis. In particular, the interactions among variables from six factors affecting slope failure are selected to form the interaction terms for the regression. The interactions among two, three, and four variables at one time were tested. Only significant interaction terms are retained for analysis. When interaction terms are introduced into the model, the ranking of the significance of some of the variables will change. Some of the variables showing significance in the previous step may become insignificant, and some of the interaction terms showing significance are added into the model. After many tests with the interaction terms, the model that produces the best prediction result is adopted as the final optimal model. Despite the fact that all independent variables including different interaction terms were introduced in the regression analysis, logistic regression analysis will be showed significance in the final best model when interaction terms were added. In the fifth step, the models obtained from above and the factors influence to the slope failure occurrences generated from the analysis are evaluated. Slope failure probability values between 0 and 1 at each unique-condition unit are obtained from the final regression. A general description of the slope failure probability is adopted in this study, and the range of slope failure probability is grouped into five categories to create the final.
66
Table 4.1 Classification of predicted the probabilities of slope failure from the logistic regression analysis (Atkinson and Massara, 1998)
Estimated Probabilities of occurrence 0.75 ~ 1.0 0.55 ~ 0.75 0.30 ~ 0.55 0.10 ~ 0.30 0.00 ~ 0.10 Relative of probabilities class Very high High Moderate Low Very low
4.2 Independent Variables and Sampling
Several different geological and geographical parameters considered to be relevant to the occurrence of slope failure were selected as the independent variables. lithology, direction, vegetation and landscape topography were treated as categorical independent variables, whereas inclination angle of slope and elevation were continuous independent variables (table 14). For the purpose of the statistical analysis, sample data representing both failure and unfailure of slopes must be provided to fit the logistic regression analysis. The way in which these data are obtained will affect both the nature of the regression relation and the accuracy of the resulting estimates (Atkinson and Massari, 1998). In this study, the data set of slope failure inventory is an indispensable data source representative of samples of slope failure occurrences. All locations of the slope failure scars were thus used to extract the physical parameter (independent variables) automatically from the existing data layers. Altogether, 506 locations were chosen for the representing the unfailure area. These locations were obtained using a random sampling scheme. In the present situation, the dependent variable is a binary variable representing the failure and unfailure of slope. Where the dependent variable is binary, the logistic link function is applicable (Atkinson and Massari 1998). The dependent variable must be input as either 0 or 1, so the model analysis applies well to analysis for possibility of slope failure occurrence. Logistic regression coefficients can be used to estimate ratio for each of the independent variables in the model analysis. The training data were then used to input to the
67
logistic regression analysis within the Statistical Package for Social Science (SPSS), desk-top statistical software, to obtain the coefficient and odds ratio for the logistic regression analysis.
Table 4.2 Categories of the independent variables
Items
Lithology
Variable of categories
Sedimentary Rocks and mixtures with recent materials (i.e., heterogeneous soil and small rocks) Littoral deposit and mixtures with recent materials (i.e.,
Code
S_R
L_R
heterogeneous soil and small rocks Igneous rocks and mixtures with recent materials (i.e., I_R
heterogeneous soil and small rocks Metamorphic rocks and mixtures with recent materials (i.e., M_R
heterogeneous soil and small rocks Volcanic rocks and mixtures with recent materials (i.e., heterogeneous soil and small rocks Inclination Angle 60 – 120 12 – 18 18 – 24
0 0 0 0 0 0
V_R
Inc_6 Inc_12 Inc_18 Inc_24 Inc_30 Inc_36 Inc_42 HT LT G NV V R F N NE E SE S SW W NW
240 – 300 30 - 36 36 - 42
0 0 0 0
42 – 48 Vegetation
Woodland or High Tree Scrubland or Low Tree Grassland Bare land or n vegetation
R2,X2, test NO
R2= Hosmer and Lameshow Test X2= Fit goodness Test
P(event) test P(event) ≤ 0.05 NO Yes
Finish
Figure 4.1 Flow chart of logistic regression analysis
69
4.3 GIS Application for Slope Failure Mapping
In the last twenty years, Geographical Information Systems (GIS) and Remote Sensing have become integral tools for the evaluation of natural hazard phenomena (Nagarajan et al 1998; Liu et al 2004). Moreover, GIS is an excellent and useful tool for the spatial analysis of a multi-dimensional phenomenon such as landslides and for the landslide susceptibility mapping (Carrara et al 1999; van Westen et al 1999; Lan et al 2004). Slope failure events are associated with various physical factors and therefore almost all methods of slope failure i.e., landslide susceptibility mapping focus on: a) the determination of the physical factors which are directly or indirectly correlated with slope failure (slope failure factors); b) the selection of the rating-weighting system of all factors and of the classes of each one of them; c) the overall estimation of the relative role of causative factors in producingslope failure; and d) the final susceptibility zoning by classifying the land surface according to different hazard degrees (Anbalagan 1992; Guzzetti et al 1999; Dai et al 2002). The slope failure i.e., landslide and surface failure map is a practical tool in natural and urban planning; it can be applied for determining land use zones, in construction design and planning various future projects. In this study, GIS based on predicting probability of slope failure maps were generated; in the study site are western parts of East timor. This was accomplished by methods for correlating factors, which affect slope failure occurrences. The factors influence to slope failures which were taken into account was: lithology, slope inclination angle, slope gradients, vegetation, landscape topography, slope aspect (direction) and elevation. A frequency distribution of the number of the slope failure events of the study area in each items of the factor category was performed in order to rate the classes. The models used to combine the factors influence to slope failure and assess the overall probability of slope failure by statistics logistic regression analysis. The produced maps were classified into four zones: the Very Low, Low , Moderate, High and Very High probability zone and validated using the other number of the slope failure events of the study area. Evaluation of the results is optimized through a Landslide Models Indicator, the application of which demonstrated that the best desired outcome is provided by the model. Moreover it was estimated that this model is easier to set up and operate than the first model.
70
Regarding the identification of the factors influence to slope failure, the used data are in some cases either readily available or can be easily collected. In other cases statistical analysis was performed. As for the assigned rates and weights, the methodology used involves landslide inventory and frequency distribution, frequency ratio, density, multivariate statistical methods, trial and error method, local experience, field knowledge and literature (Gupta and Joshi 1990; Anbalagan 1992; Zêzere et al 1999; Temesgen et al 2001; Lee and Min 2001; Donati and Turrini 2002; Saha et al 2002; Gritzner et al 2001; Liu et al 2004, Lee and Sambath 2006). Most of the methods employed for the overall estimation of the relative contribution factors influence to slope failure are based on statistical mathematical operations, which combine the factors (Temesgen et al 2001; Saha et al 2002; Chau et al 2004, Ayalew et al 2005). Finally, the goals of this study are: a) the production of probabilities of slope failure susceptibility maps based on GIS techniques using the models of combining the instability factors and estimation of overall slope failure susceptibility and b) the evaluation of these models and produced maps. A GIS database has been developed using ArcGIS version 3.3 software. The slope failure occurrences in the study area and the influence factors have been recorded and saved as separate layers in the database. All the data layers were in vector format, transformed in grids with cell size 30x30 meters. Start
Slope Failure inventory map
Rating of each factors and category where influence to slope failure Apply Statistics Logistic regression Analysis Production of probabilities of slope failure maps based on GIS techniques Finish 71
Figure 4.2 Flow chart of Production of probabilities of slope failure maps based on GIS techniques
CHAPTER V ANALYSIS RESULT
5.1 Introduction
A logistic regression analysis was constructed initially based on the physical parameters or independent variables as defined above. Then, each step, independent variables are evaluated for removal one by one if they do not contribute sufficiently to the regression equation. In this analysis, the likelihood-ratio test is always used for determining whether variables should be added to the analysis. This involves estimating the model analysis with each independent variable eliminated in turn and looking at the change in to the logarithm of likelihood when each independent variable is deleted. If the result analysis observed significance level is greater than probability for stepwise (0.05 in this analysis) for remaining in the analysis, the variables is removed from the analysis recalculated to see if any and statistics analysis are
other independent variables are eligible for removal. The
independent variables in this analysis are: lithology, inclination angle, vegetation, landscape topography, direction and elevation. By studying and analysis the causal factors affecting for slope failure in regional mountainous of these study site, this study tries to contribute to the restricted knowledge on slope failure in East Timor. After a brief introduction of the study area and the spatial distribution and characteristics of its slope failure, the preconditions, preparatory and triggering causal factors will be discussed with attention to their spatial variation.
5.2 All Study site Analysis Result
Logistic regression analysis result of all study (i.e., Bobonaro, cailaco, Zumalai, Atsabe, Maliana, Ainaro, Hatolia and Hatobuilico) are shows in table 5.1 and 5.2. It can be seen that the model analysis produced a concordance rate of 90.3 % with the use of 0.50 as a classification cutoff value. This result is in agreement with the work in northern Italy by Carrara and others (1991).By examining this result to predict probabilities of slope failure affecting by the independent variables, we can see what a different classification rule should be adopted when applying the model analysis to each factor in the study area and obtained regression model composed of significant variables. 72
Table 5.1 Classification table of the cut value 0.50
Observed Status of Step slope Predicted Status of Slope Unfailure Failure Un-failure 450 56 Failure 50 456 Overall Percentage Percentage Correct 90.5 90.1 90.3
Table 5.2 Coefficient values and influence ratio of logistic regression of each item and category in all study site
Item Category Number
Lithology Sed. rocks Lit. dep. rocks Igneous rocks Meta. Rocks Vol. rocks Inclination Angle ( )
o
Table 5.3 Coefficient values and influence ratio of the logistic regression of interaction term when combined with other item and categories in all study site
Item Category Number
Lithology Sed. rocks Lit. dep. rocks Igneous rocks Meta. Rocks Vol. rocks Inclination Angle ( )
o
From the analysis result (Table 5.2), the regression coefficients of the lithology item and category of sedimentary rocks are 3.66, which is the highest among all items and category. This variable most contributes and affecting to slope failures; the influence ratio of slope failure against unfailure slope is 39 times when this variable is present and other items and category are controlled. The coefficient of the vegetation item and category of no vegetation or bare land is 3.49 which is the second highest, with an influence ratio of 33. And the next most categories is volcanic rocks, northeast, littoral deposit rocks, valley side, northwest, north, grassland, and ridge side (Figure 5.1).
75
45 40 35 Influence ratio 30 25 20 15 10 5 0
Sedimetary rocks Littoral deposit rocks Grassland Bare land Volcanics rocks Northwest Northeast Valley Ridge North
All site
Cate gory
Fig. 5.1 Ranking of the top ten significant item and category based on influence ratio
160 140 120 100 80 60 40 20 0
Bare land Northeast Sedimetary rocks Volcanics rocks
All site
No interaction Interaction with each item and category
Influence ratio
Littoral deposit rocks
Northwest
Valley
Category
Figure 5.2 The top ten ranking of interaction term when combined with other variables based on the influence ratio
76
Grassland
Ridge
North
The next is the interaction term when combined with other variables controlled by this analysis, the regression coefficient and influence ratio of most of the variables gradually increases 1 to 4.5 times an individual variable (Table 5.3 and Figure 5.2). Based on this analysis result and the actual condition and characteristics slope failure distribution in study area, it can be seen that geology features are most important variable in this study, distribution of lithology as well as bedrock of sedimentary rocks and littoral deposit rocks, surface materials, and the difference between surface aspect and dip direction of bedding are more important than elevation and difference between slope and inclination angle in controlling slope stability. Most slope failure occurred in study area where the factors representing the terrain aspect nearly parallel to the dip direction of the bedrock coexists with other influential conditions including the a few igneous rocks, metamorphic rocks and volcanic rocks thin till or other unconsolidated material, steep slope and elevation from 200 to 2100 m. Vegetation variables were used in this analysis and shown significance, as the vegetation used in this study might be different from that of the time the slope failure occurred, the interpretation of the importance of the vegetation cover may vary over time, but in actual condition in East Timor, most of study site where slope failure are covered by bare land and grassland. However, when heavy rainfall infiltrates in to the soil slope, it will clearly increase the moisture content of the soil above the phreatic surface, but as the water flows downward, it may also result in a rise in the position of the phreatic surface. Such a rise could be the caused of slope failure. Landscape topography is one of the important variables affecting to slope failure occurrences. Landscape of soil mantled ridge and valley topography, shallow landslides typically only involve the soil mantle and commonly occur at or near the soil-bedrock boundary. These landslides may mobilize and travel a short distance down slope before coming to rest either still on the hillside. The analysis result shows that emerges from this work on topography landslides shows that surface topography has a great bearing on the location and frequency of shallow landslide. Importantly, it is not just the local slope that matters, but also the curvature of the topography and how it focuses or spreads runoff down slope. A physically, that quantifies the influence of surface topography on pore pressure in a
77
shallow slope stability model may effectively capture the essential linkage between topography and slope failure. The direction of slope has the potential to influences its physical properties and it susceptibility to failure. The process that may be operating including to sunlight, drying winds and possibly rainfall (Evans and others 1999). The distribution of aspect among the mapped and the significance analysis shows that the frequency of slope failure was highest on northeast – northwest and north – facing slopes, indicating that natural terrain landslide is more common on these slopes. The frequency of slope failure was lowest on those slopes facing south and west, while the frequency of slope failure remained moderate on the East – southeast and southwest-facing slopes. From the air photograph interpretation shows that this may be attributed to fact that there is more vegetation cover on south and west slopes. Based on the logistic regression analysis result and slope failure distribution analysis in those areas, vegetation, lithology, landscape topography of slope and elevation are more important than elevation and inclination angle of slope failure. From the Figure 5.3 is the histogram to predict the probabilities of slope failures affected by independent variables are used in this analysis. Theoretically, if we have an analysis model that successfully distinguishes the two independent variables on a classification cutoff value of 0.5, the cases for which slope failure has occurred should be to the right of 0.5, whereas the cases for which slope failure has not occurred should be to the left of 0.5( Figure 5.3). A fivefold classification scheme, ranging from very high probabilities of slope failure, to very low, was employed for the predicted probabilities of occurred. It should be noted that the complexity of the failure processes means that any evaluation of stability contains a considerable amount of uncertainty. The use of predicted probability of slope failure in this study is limited and is not suitable for site specific evaluation. The reliability of the assessment result depends on a multitude of factors ranging from the quality of the data base, the introduction of potential errors associated with data entry to the limitations and assumptions inherent in the statistical techniques ( Rowbotham and Dudycha 1998).
78
Table 5.4 Classification of predicted the probabilities of slope failure from the logistic regression analysis (Atkinson and Massara, 1998)
Estimated Probabilities of occurrence 0.75 ~ 1.0 0.55 ~ 0.75 0.30 ~ 0.55 0.10 ~ 0.30 0.00 ~ 0.10 Relative of probabilities class Very high High Moderate Low Very low
The ranges individual classes presented in Table 5.4 were derived based on the histogram of the estimated of probabilities of slope failure shown in Table 5.5 and Figure 5.3, and Figure 5.4. Zones classified for predicting of slope failure in this study site as being of “very high probabilities”, accounting for 65% of this study area and exhibit a strongly clustered pattern of spatial distribution and cover by grassland and bare land. This category is distinguished from the “high” category by relatively high elevations and steeper terrain. Most of the locations of identified slope failure actually occurred within this class. The” high probabilities class”, occupies 11% of the study area, is mainly distributed in the middle section of slopes and bears a high potential for slope failure occurrence. The zone of moderate class covers 11% of the study are, and are featured by lower sections of slopes and ridges. The zone of low probabilities of slope failure occurred, covering 9% of this study area, is relatively dispersed in its spatial distribution, and hence the chance for slope failure to develop within this class is small. And finally, zone of “very low” covering 4% of total study area are distributed on high mountains that are characterized by relatively gentle gradient of slope. All these sites are highly table and are not favorable to development of slope failure. In this study, a particular problem with uncertainty is that the 1:15,000-scale topographic condition cannot fully reflect the micro-topography conditions prerequisite for slope failure because slope failure in the all study area is characterized by small and bigger volumes, that a slight change in micro-scale landform may have a strong influence on the slope failure. This, however, has not been reflected in the topographic map. Another problem is the 1:350,000scale geological map used in this study cannot fully reflect the distribution of colluviums or residual soils that are of critical significance to the slope failure.
79
R E Q U E N C Y
240 ô ó ó ó1 160 ô0 ó0 ó0 ó0 80 ô0 ó0 ó0 ó0000 0010 010 10 10 10 10 110 01111
ô ó ó 0ó 1ô 1ó 1ó 1ó 1ô 1ó 1ó 111ó
Predicted òòòòòòòòòòòòòòôòòòòòòòòòòòòòòôòòòòòòòòòòòòòòôòòòòòòòòòòòòòòò Prob: 0 .25 .5 .75 1 Group: 000000000000000000000000000000111111111111111111111111111111 Unfailure Failure Predicted Probability is of Membership for 1.00 The Cut Value is .50 Symbols: 0 - .00 1 - 1.00 Each Symbol Represents 20 Cases.
Figure 5.3 Observed Groups and Predicted Probabilities (Logistic regression analysis) Table 5.5 Predicting for probability of slope failure
Failure Probability ranges 0 ~ 0.10 Number 20 20 20 20 20 20 20 40 80 200 460 Percentage 4.35 4.35 4.35 4.35 4.35 4.35 4.35 8.70 17.40 43.45 100 Number 220 60 40 20 20 20 20 20 20 20 460 Unfailure Percentage 47.80 13.05 8.70 4.35 4.35 4.35 4.35 4.35 4.35 4.35 100
All site
0.9~1 Probabilities of occurrence 0.8~0.9 0.7~0.8 0.6~0.7 0.5~0.6 0.4~0.5 0.3~0.4 0.2~0.3 0.1~0.2 0~0.1 0 10 20 30 40 50 60
Unfailred Failured
Percentage of occurrences
Figure 5.4 Histogram of predicted probabilities of slope failure
Figure 5.5 Map of relative slope failure susceptibility
81
5.3 Specific Site Analysis 5.3.1 Bobonaro site
Logistic regression analyses of Bobonaro site are shows in Table 5.6 and Table 5.7. It can be seen that the model analysis produce a concordance rate of 88.6% with the use of 0.5 as a classification cutoff value. By examining this result to predict probabilities of slope failure affecting by the independent variables, we can see what a different classification rule should be adopted when applying the model analysis to each factor in the study site and obtain the regression model composed of significant variables. Table 5.6 Classification table of the cut value 0.50
Observed Status of slope Predicted Status of Slope Unfailure Unfailure 153 Failure 24 Overall Percentage Percentage Correct Failure 14 143 91.6 85.6 88.6
Step
Table 5.7 Coefficient values and influence ratio of logistic regression of each item and category
Item Category Number
Lithology Sed. rocks Lit. dep. rocks Igneous rocks Meta. Rocks Vol. rocks Inclination Angle ( )
o
Table 5.8 Coefficient values and influence ratio of the logistic regression of interaction term when combined with other item and categories in Bobonaro site
Item Category Number
Lithology Sed. rocks Lit. dep. rocks Igneous rocks Meta. Rocks Vol. rocks Inclination Angle ( )
o
From the analysis result (Table 5.7), the regression coefficients of the vegetation item and category of no vegetation or bare land are 3.83, which is the highest among all items and category. This variable most contributes and affecting to slope failures; the influence ratio of slope failure against unfailure slope is 46 times when this category is present and other items and category are controlled. The coefficient of the elevation item and category of elevation 1100m ~ 1400m is 3.77 which is the second highest, with an influence ratio of 44. And the next most category is inclination angle 36 o ~ 42 o , North, ridge, grassland, northeast, sedimentary rocks, valley side and elevation 800m ~ 1100m ( Figure 5.6).
84
The next is the interaction term when combined with other variables controlled by this analysis, the regression coefficient and influence ratio of most of the item and category gradually increases (Table 5.8 and Figure 5.7).
50 45 40 35 30 25 20 15 10 5 0
Elev.1100~1400 Inclination angle_36~42 Bare land
Influence ratio
Bobonaro site
Category
Figure 5.6 Ranking of the top ten significant item and category based on influence ratio
85
Elev.800~1100
Sedimetary rocks
R idge
N orth
Northeast
Grassland
Valley
180 160 140 Influence ratio 120 100 80 60 40 20
Elev.1100~1400 Elev.800~1100 No interaction Interaction with each item and category
Bobonaro site
0
Bare land
North
Ridge
Inclination angle_36~42
Grassland
Northeast
Sedimetary rocks
Category
Figure 5.7 The top ten ranking of interaction term when combined with other variables based on the influence ratio Based on this analysis result, the actual condition and distribution of slope failure in Bobonaro site, it can be seen that vegetation is most important factor were used in this analysis and show significance. In this site, most of slope failure has occurred in the bare land and grassland area. However, when heavy rainfall occurred and infiltrates in to the soil, it will clearly increase the moisture content of the soil above phreatic surface and the water flows downward, it may also result in a rise the position of the phreatic surface and could be the caused of slope failure. While the variation in soil types and characteristics throughout the Bobonaro ridge is large, some of them have a distinct boundary between the soil and the underlying rock is common. During heavy rains, water stagnates on this continuity, creating positive pore water pressures on this shear plane on which the soil can easily slide down. Elevation is one of the important factors affecting for slope failure in this site. Most of slope failure occurred frequently do have ranging from 200m to 1700m. The distribution of elevation of slopes failure in this site shows that hill slopes between 200m to 500m in elevation had frequencies of slope failure that were greater than those on hill slopes that are
86
Valley
800m to 1100m and 1400m to 1700m and less than hill slopes that are 500m to 800m and 1100m to 1400m in elevation. At intermediate elevations there are mountain summit, which is more prone to landslide that are usually characterized by weather rocks, and the shear strength of these is much higher. At very low elevations, the frequency of slope failure is low because the terrain is gentle, and is covered by residual soil, and higher perched water table will required initiating slope failure. In this site, slope inclination angle is an important variable occurrence and show significance. Inclination angle is an essential component of slope stability analysis. As slope inclination angle increases, the level of gravitation-induced shear stress in the residual soils increases as well. It can be seen that examination of the distribution of number of slope failure with corresponding slope inclination angle in Bobonaro study site shows that most of slope failures has occurred with inclination angle ranges increase in the 120 – 360 and gradually decrease in the ranges 60 – 120 and 360 – 480. This is refection that steep natural slope with outcropping bedrock and hence much higher shear strength may be susceptible to shallow landslide. The direction of slope has the potential to influences its physical properties and it susceptibility to failure. The process that may be operating including to sunlight, drying winds and possibly rainfall (Evans and others 1999). The distribution of aspect among the mapped and the significance analysis shows that the frequency of slope failure was highest on north and northeast– facing slopes, indicating that natural terrain landslide is more common on these slopes. The frequency of slope failure was lowest on those slopes facing east - south – west and northwest, while the frequency of slope failure remained moderate on the southeast and southwest -facing slopes. From the air photograph interpretation shows that this may be attributed to fact that there is more vegetation cover on east, south, west and northwest slopes. Geology features are most important variable in this study, distribution of sedimentary rocks and a few of igneous rocks and metamorphic rocks, surface materials, and the difference between surface aspect and dip direction of bedding are more important than elevation and difference between slope and inclination angle in controlling slope stability. Most slope failure has occurred in study area where the factors representing the terrain aspect
87
nearly parallel to the dip direction of the bedrock coexists with other influential conditions including the littoral deposit bedrock thin till or other unconsolidated material, steep slope and elevation from 200m to 1700 m. It should be note that thin colluvium or residual soil in steep terrain, which is most susceptible to slope failure, is not fully reflected in the geological map by lithological characteristics of underlying bedrock. Structural information is also available from digital geological maps. However, qualitative examination of spatial distributions suggests that the correlation between slope failure and mapped linier structural feature at the 1:350,000- scale is not good, and the structural information is, thus, excluded in this study. Landscape topography is one of the important variables affecting to slope failure. In this study sites, landscape of soil mantled ridge and valley topography, shallow landslides typically only involve the soil mantle and commonly occur at or near the soil-bedrock boundary. These landslides may mobilize and travel a short distance down slope before coming to rest either still on the hillside. The analysis result shows that emerges from this work on topography landslides shows that surface topography has a great bearing on the location and frequency of shallow landslide. Importantly, it is not just the local slope that matters, but also the curvature of the topography and how it focuses or spreads runoff down slope. A physically, that quantifies the influence of surface topography on pore pressure in a shallow slope stability model may effectively capture the essential linkage between topography and slope failure. Compared to other study site, the critical slope for slope failure in Bobonaro site is rather higher, with slope failure occurring on slopes do have inclination angle from 6 o onward. Gentle slopes exhibiting slope failure are common in the Bobonaro zone, where soil stratification and human interference are also important. From the Figure 5.8 and Figure 5.9 are the histograms to predict the probabilities of slope failures affected by independent variables are used in this analysis. Theoretically, if we have an analysis model that successfully distinguishes the two independent variables on a classification cutoff value of 0.5, the cases for which slope failure has occurred should be to the right of 0.5, whereas the cases for which slope failure has not occurred should be to the left of 0.5(Figure 5.8). A fivefold classification scheme, ranging from very high probabilities
88
of slope failure, to very low, was employed for the predicted probabilities of occurrence. It should be noted that the complexity of the failure processes means that any evaluation of stability contains a considerable amount of uncertainty. The use of predicted probability of slope failure in this study is limited and is not suitable for site specific evaluation. The reliability of the assessment result depends on a multitude of factors ranging from the quality of the data base, the introduction of potential errors associated with data entry to the limitations and assumptions inherent in the statistical techniques ( Rowbotham and Dudycha 1998).
F R E Q U E N C Y
ó 60 ô ó ó ó 40 ô0 ó0 ó0 0 ó0 0 20 ô0 0 ó0 0 0 1 0 0 1 00 1 0 10 10 10 110 1 1 11110 ó0 010 0 ó0 000000
1ó 1ô 1ó 1ó 1ó 1ô 1ó 1ó 1ó 1ô 11ó 11ó 10111ó
Predicted òòòòòòòòòòòòòòôòòòòòòòòòòòòòòôòòòòòòòòòòòòòòôòòòòòòòòòòòòòòò Prob: 0 .25 .5 .75 1 Group: 000000000000000000000000000000111111111111111111111111111111 Unfailure Failure Predicted Probability is of Membership for 1.00 The Cut Value is .50 Symbols: 0 - .00 1 - 1.00 Each Symbol Represents 5 Cases.
Figure 5.8 Observed Groups and Predicted Probabilities (Logistic regression analysis)
89
Table 5.9 Predicting for probability of slope failure in Bobonaro Site
Failure Probability ranges
0 ~ 0.10 0.11 ~ 0.20 0.21 ~ 0.30 0.31 ~ 0.40 0.41 ~ 0.50 0.51 ~ 0.60 0.61 ~ 0.70 0.71 ~ 0.80 0.81 ~ 0.90 0.91 ~ 1.00 Total
Figure 5.9 Histogram of the predicted probabilities of slope failure
90
The ranges individual classes presented in Table 5.4 were derived based on the histogram of the estimated of probabilities of slope failure shown in Table 5.9 and Figure 5.8 and Figure 5.9. Zones classified for predicting of slope failure in Bobonaro site as being of “very high probabilities”, accounting for 76% of Bobonaro site and exhibit a strongly clustered pattern of spatial distribution and cover by grassland and bare land. This category is distinguished from the “high” category by relatively high elevations and steeper terrain. Most of the locations of identified slope failure actually occurred within this class. The” high probabilities class”, occupies 7.5% of the study area, is mainly distributed in the middle section of slopes and bears a high potential for slope failure. The zone of moderate class covers 7.5% of the study area, and is featured by lower sections of slopes and ridges. The zone of “low probabilities” of slope failure, covering 6% of this study area, is relatively dispersed in its spatial distribution, and hence the chance for slope failure to develop within this class is small. And finally, zone of “very low” covering 3% of total study area are distributed on high mountains that are characterized by relatively gentle gradient of slope. All these sites are highly table and are not favorable to development of slope failure. It should be noted that the complexity of the failure processes means that any evaluation of stability contains a considerable amount of uncertainty. The use of predicted probability of slope failure in this study is limited and is not suitable for site specific evaluation. The reliability of the assessment result depends on a multitude of factors ranging from the quality of the data base, the introduction of potential errors associated with data entry to the limitations and assumptions inherent in the statistical techniques ( Rowbotham and Dudycha 1998).
91
5.3.2 Cailaco site
Logistic regression analyses of Cailaco site are shows in Table 5.10 and Table 5.10. It can be seen that the model analysis produce a concordance rate of 94% with the use of 0.5 as a classification cutoff value. By examining this result to predict probabilities of slope failure affecting by the independent variables, we can see what a different classification rule should be adopted when applying the model analysis to each factor in the study site and obtain the regression model composed of significant variables.
Table 5.10 Classification table of the cut value 0.50
Observed Status of slope Predicted Status of Slope Unfailure Failure Unfailure 125 8 Failure 8 125 Overall Percentage Percentage Correct 94.0 94.0 94.0
Step
Table 5.11 Coefficient values and influence ratio of logistic regression of each item and category
Item Category Number
Lithology Sed. rocks Lit. dep. rocks Igneous rocks Meta. Rocks Vol. rocks Inclination Angle ( )
o
Table 5.12 Coefficient values and influence ratio of the logistic regression of interaction term when combined with other item and categories
Item Category Number
Lithology Sed. rocks Lit. dep. rocks Igneous rocks Meta. Rocks Vol. rocks Inclination Angle ( )
o
From the analysis result (Table 5.11), the regression coefficients of the vegetation item and category of no vegetation or bare land are 3.83, which is the highest among all items and category. This variable most contributes and affecting to slope failures; the influence ratio of slope failure against unfailure slope is 46 times when this category is present and other items and category are controlled. The coefficient of the elevation item and category of elevation 1100m ~ 1400m is 3.77 which is the second highest, with an influence ratio of 44. And the next most category is inclination angle 36 o ~ 42 o , grassland, northwest, inclination angle of slope are 30 o ~ 36 o , sedimentary rocks, east –southeast - facing of slopes, elevation 800m ~ 1100m and valley side of slopes ( Figure 5.10).
94
The next is the interaction term when combined with other variables controlled by this analysis, the regression coefficient and influence ratio of most of the item and category gradually increases 1 to 4 times an individual variables (Table 5.12 and Figure 5.11). Based on the logistic regression analysis result and slope failure distribution analysis in Cailaco site, vegetation, direction, inclination angle of slope, lithology and landscape topography of slope are more important than slope elevation.
120 100 Influence ratio 80
Cailaco site
60 40 20 0 Northwest Inclination angle_30~36 Sedimetary rocks Northeast Bare land Grassland Elev.800~1100 East Southeast Valley
Cate gory
Fig. 5.10 Ranking of the top ten significant item and category based on influence ratio
No interaction Interaction with each item and category
Sedimetary rocks
Northeast
Southeast
East
Elev.800~1100
Bare land
Grassland
Inclination angle_30~36
Category
Figure 5.11 The top ten ranking of interaction term when combined with other variables based on the influence ratio
Vegetation variables were used in this analysis and shown significance, as the vegetation used in this study might be different from that of the time the slope failure has occurred, the interpretation of the importance of the vegetation cover may vary over time, but in actual condition in Cailaco site, most of slope failure occurred are covered by bare land and grassland. However, when heavy rainfall infiltrates in to the soil slope, it will clearly increase the moisture content of the soil above the phreatic surface, but as the water flows downward, it may also result in a rise in the position of the phreatic surface. Such a rise could be the caused of slope failure. The direction of slope has the potential to influences its physical properties and it susceptibility to failure. The process that may be operating including to sunlight, drying winds and possibly rainfall. The distribution of aspect among the mapped and the significance analysis shows that the frequency of slope failure was highest on northeast and
96
Valley
northwest facing slopes, indicating that natural terrain landslide is more common on these slopes. The frequency of slope failure was lowest on those slopes facing south, while the frequency of slope failure remained moderate on the East and east -facing slopes. From the air photograph interpretation shows that this may be attributed to fact that there is more vegetation cover on south – west and southwest facing slopes. In this site, slope inclination angle is an important variable of slope failure and show significance. Inclination angle is an essential component of slope stability analysis. As slope inclination angle increases, the level of gravitation-induced shear stress in the residual soils increases as well. It can be seen that examination of the distribution of number of slope failure with corresponding slope inclination angle in Cailaco site shows that most of slope failures occurred with inclination angle ranges increase in the 120 – 300 and gradually decrease in the ranges 60 – 120 and 300 – 420. Landscape topography is one of the important variables affecting to slope failure. Landscape of soil mantled ridge and valley topography, shallow landslides typically only involve the soil mantle and commonly occur at or near the soil-bedrock boundary. These landslides may mobilize and travel a short distance down slope before coming to rest either still on the hillside. The analysis result shows that emerges from this work on topography landslides shows that surface topography has a great bearing on the location and frequency of shallow landslide. Importantly, it is not just the local slope that matters, but also the curvature of the topography and how it focuses or spreads runoff down slope. A physically, that quantifies the influence of surface topography on pore pressure in a shallow slope stability model may effectively capture the essential linkage between topography and slope failure. Based on the logistic regression analysis result and slope failure distribution analysis in this area, vegetation, lithology, landscape topography of slope and elevation are more important than elevation and inclination angle of slope failure. From the Figure 5.4 is the histogram to predict the probabilities of slope failures affected by independent variables are used in this analysis. Theoretically, if we have an analysis model that successfully distinguishes the two independent variables on a classification cutoff value of 0.5, the cases for which slope failure has occurred should be to the right of 0.5, whereas the cases for which slope failure has not occurred should be to the left of 0.5
97
(Figure 5.12). A fivefold classification scheme, ranging from very high probabilities of slope failure, to very low, was employed for the predicted probabilities of occurrence. It should be noted that the complexity of the failure processes means that any evaluation of stability contains a considerable amount of uncertainty. The use of predicted probability of slope failure in this study is limited and is not suitable for site specific evaluation. The reliability of the assessment result depends on a multitude of factors ranging from the quality of the data base, the introduction of potential errors associated with data entry to the limitations and assumptions inherent in the statistical techniques ( Rowbotham and Dudycha 1998). The ranges individual classes presented in Table 5.4 were derived based on the histogram of the estimated of probabilities of slope failure shown in Table 5.13 and Figure 5.13. Zones classified for predicting of slope failure in Cailaco site as being of “very high probabilities”, accounting for 80% of this study area and exhibit a strongly clustered pattern of spatial distribution and cover by grassland and bare land. This category is distinguished from the “high” category by relatively high elevations and steeper terrain. Most of the locations of identified slope failure actually occurred within this class. The” high probabilities class”, occupies 10% of the study area, is mainly distributed in the middle section of slopes and bears a high potential for slope failure occurrence. The zone of moderate class covers 10% of the study are, and are featured by lower sections of slopes and ridges.
Figure 5.13 Histogram of the predicted probabilities of slope failure
5.3.3 Zumalai Site
Logistic regression analyses of Zumalai site are shows in Table 5.14 and Table 5.15. It can be seen that the model analysis produce a concordance rate of 84.7% with the use of 0.5 as a classification cutoff value. By examining this result to predict probabilities of slope failure affecting by the independent variables, we can see what a different classification rule should be adopted when applying the model analysis to each factor in the study site and obtain the regression model composed of significant variables. Table 5.14 Classification table of the cut value 0.50 in Zumalai site
Observed Status of slope Predicted Status of Slope Unfailure Unfailure 66 Failure 14 Overall Percentage Failure 9 61 Percentage Correct 88.0 81.3 84.7
Step
100
Table 5.15 Coefficient values and influence ratio of logistic regression of each item and category
Item Category Number
Lithology Sed. rocks Lit. dep. rocks Igneous rocks Meta. Rocks Vol. rocks Inclination Angle ( )
o
Table 5.16 Coefficient values and influence ratio of the logistic regression of interaction term when combined with other item and categories in Zumalai site
Item Category Number
Lithology Sed. rocks Lit. dep. rocks Igneous rocks Meta. Rocks Vol. rocks Inclination Angle ( )
o
From the analysis result (Table 5.15), the regression coefficients of the direction item and category of northeast are 3.92, this is the highest among all items and category. This variable most contributes and affecting to slope failures; the influence ratio of slope failure against unfailure slope is 50 times when this variable is present and other items and category are controlled. The coefficient of the direction item and category of no north is 3.58 which is the second highest, with an influence ratio of 36. And the next most important categories are inclination angle of slope are 30 o ~ 36 o , sedimentary rocks, south, valley, northwest, bare land, southwest, southeast and ridge side of slope (Figure 5.14). The next is the interaction term when combined with other variables controlled by this analysis, the regression coefficient and influence ratio of most of the item and category gradually increases 1 to 4 times an individual variables (Table 5.16 and Figure 5.15). Based on the logistic regression analysis result and slope failure distribution analysis in Zumalai area, direction, inclination angle of slope, lithology, vegetation and landscape topography of slope are more important than slope elevation.
60 50
Zumalai site
Influence ratio 40 30 20 10 0
angle_30~36 Bare land North Sedimetary Valley Inclination Southwest Southeast Northwest Northeast rocks Ridge
Cate gory
Fig. 5.14 Ranking of the top ten significant item and category based on influence
103
Zumalai site
No interaction Interaction with each item and category
90 80 70 influence ratio 60 50 40 30 20 10 0
Inclination angle_30~36 Northeast Bare land Southwest Northwest Sedimetary rocks Southeast Valley Ridge North
Category
Figure 5.15 The top ten ranking of interaction term when combined with other variables based on the influence ratio The direction of slope has the potential to influences its physical properties and it susceptibility to failure. The process that may be operating including to sunlight, drying winds and possibly rainfall. The distribution of aspect among the mapped and the significance analysis shows that the frequency of slope failure was highest on northeast and southwest facing slopes, indicating that natural terrain landslide is more common on these slopes. The frequency of slope failure was lowest on those slopes facing north-east-westnorthwest, while the frequency of slope failure remained moderate on the southeast -facing slopes. From the air photograph interpretation shows that this may be attributed to fact that there is more vegetation cover on north – east - west – south and west facing slopes. In this site, slope inclination angle is important variable slope failure and show significance. Inclination angle is an essential component of slope stability analysis. As slope inclination angle increases, the level of gravitation-induced shear stress in the residual soils increases as well. It can be seen that examination of the distribution of number of slope failure with corresponding slope inclination angle in Zumalai site shows that most of slope 104
failures occurred with inclination angle ranges increase in the 12 o – 24 o and gradually decrease in the ranges 6 o – 12 o and 24 o – 48 o . Landscape topography is one of the important variables affecting to slope failure. Landscape of soil mantled ridge and valley topography, shallow landslides typically only involve the soil mantle and commonly occur at or near the soil-bedrock boundary. These landslides may mobilize and travel a short distance down slope before coming to rest either still on the hillside. The analysis result shows that emerges from this work on topography landslides shows that surface topography has a great bearing on the location and frequency of shallow landslide. Importantly, it is not just the local slope that matters, but also the curvature of the topography and how it focuses or spreads runoff down slope. A physically, that quantifies the influence of surface topography on pore pressure in a shallow slope stability model may effectively capture the essential linkage between topography and slope failure. Geology features are most important variable in this study site, distribution of sedimentary rocks, surface materials, and the difference between surface aspect and dip direction of bedding are more important than elevation and difference between slope and inclination angle in controlling slope stability. Most slope failure occurred in study area where the factors representing the terrain aspect nearly parallel to the dip direction of the bedrock coexists with other influential conditions including the littoral deposit bedrock thin till or other unconsolidated material, steep slope and elevation from 200m to 800 m. It should be note that thin colluvium or residual soil in steep terrain, which is most susceptible to slope failure, is not fully reflected in the geological map by lithological characteristics of underlying bedrock. Structural information is also available from digital geological maps. However, qualitative examination of spatial distributions suggests that the correlation between slope failure and mapped linier structural feature at the 1:350,000- scale is not good, and the structural information is, thus, excluded in this study. Based on the logistic regression analysis result and slope failure distribution analysis in that area, vegetation, lithology, landscape topography of slope and elevation are more important than elevation and inclination angle of slope failure. From the Figure 5.4 is the histogram to predict the probabilities of slope failures affected by independent variables are used in this analysis. Theoretically, if we have an analysis
105
model that successfully distinguishes the two independent variables on a classification cutoff value of 0.5, the cases for which slope failure has occurred should be to the right of 0.5, whereas the cases for which slope failure has not occurred should be to the left of 0.5(Figure 5.16). A fivefold classification scheme, ranging from very high probabilities of slope failure, to very low, was employed for the predicted probabilities of occurrence. It should be noted that the complexity of the failure processes means that any evaluation of stability contains a considerable amount of uncertainty. The use of predicted probability of slope failure in this study is limited and is not suitable for site specific evaluation (Figure 5.3). The reliability of the assessment result depends on a multitude of factors ranging from the quality of the data base, the introduction of potential errors associated with data entry to the limitations and assumptions inherent in the statistical techniques ( Rowbotham and Dudycha 1998). The ranges individual classes presented in Table 5.4 were derived based on the histogram of the estimated of probabilities of slope failure occurrence shown in Table 5.17 and Figure 5.17. Zones classified for predicting of slope failure in this study site as being of “very high probabilities”, accounting for 75% of this study area and exhibit a strongly clustered pattern of spatial distribution and cover by grassland and bare land. This category is distinguished from the “high” category by relatively high elevations and steeper terrain. Most of the locations of identified slope failure actually occurred within this class. The” high probabilities class”, occupies 11.50% of the study area, is mainly distributed in the middle section of slopes and bears a high potential for slope failure. The zone of moderate class covers 7.5% of the study are, and are featured by lower sections of slopes and ridges. And finally, zone of “very low” covering 5% of total study area are distributed on high mountains that are characterized by relatively gentle gradient of slope. All these sites are highly table and are not favorable to development of slope failure.
106
F R E Q U E N C Y
ó 24 ô ó ó ó 16 ô ó ó ó 8 ô ó ó 00 00 00 00 00 00 00 00 00 001 00 1 1 11 1111
ó ô ó ó ó ô ó 11 ó 11 ó 11 ô 11 ó 1111 ó
ó 00 000 10 100 10 10 10 110 1111 1111 ó Predicted òòòòòòòòòòòòòòôòòòòòòòòòòòòòòôòòòòòòòòòòòòòòôòòòòòòòòòòòòòòò Prob: 0 .25 .5 .75 1 Group: 000000000000000000000000000000111111111111111111111111111111 Unfailure Failure Predicted Probability is of Membership for 1.00 The Cut Value is .50 Symbols: 0 - .00 1 - 1.00 Each Symbol Represents 2 Cases.
Figure 5.16 Observed Groups and Predicted Probabilities (Logistic regression analysis)
Table 5.17. Predicting for probability of slope failure in Zumalai Site
Failure Probability ranges 0 ~ 0.10 Number 0 2 2 2 2 2 4 6 20 32 72 Percentage 0 3 3 3 3 3 6 8 28 43 100 Number 36 14 6 4 2 2 2 2 0 0 60 Unfailure Percentage 52 21 9 6 3 3 3 3 0 0 100
Figure 5.17 Histogram of predicted probabilities of slope failure
108
CHAPTER VI CONCLUSIONS AND FUTURE SUBJECT
6.1 Conclusions
By studying and analysis on the causal factors affecting for slope failure in East Timor, this study is contributed to the restricted knowledge on slope failure in East Timor. Type of slope failure in East Timor dominantly by landslide and surface failure, and the actual condition and characteristics of slope failure were investigated based on analysis of aerial photograph and topography map. After a brief introduction of the study area and knowing the actual condition and characteristics of slope failure, and determine the factors influencing slope failure in East Timor by logistic regression analysis, the following conclusions can be obtained:
•
Types of slope failure occurred in East Timor dominantly by landslide 56% with density 0.24 Number/km2 ,surface failure are 37% with density 0.16 Number/km2 and mix of landslide and surface failure are 7% with density 0.06 Number/km2.
•
Distribution of slope failure in study area relatively highest density in sedimentary rocks and littoral deposit rocks and lowest in igneous rocks, metamorphic rocks and volcanic rocks.
•
Most of slope failure occurred on bare land and grassland in highest and lowest on woodland and scrubland.
•
The direction of slope failure was highest on northeast – northwest and north – facing slopes, the frequency of slope failure was lowest on those slopes facing south and west, while the frequency of slope failure remained moderate on the East – southeast and southwest-facing slopes. Most slope failures occurred with inclination angle ranges increase in the 12o – 36o and gradually decrease in the ranges 6o – 12o and 36o – 48o.
•
•
Based on the multivariate statistical analysis results and the observed distribution of slope failure in those study sites, vegetation, lithology, landscape topography, slope inclination angle, slope direction, and elevation were found to be the most important factors affecting to the of slope failure in mountainous study area.
109
•
By logistic regression analysis, the interaction term were introduce, the proportion of the observed all items and category predicted as high influence ratio increased by 1 to 4 times of individual category.
•
Zone classified for predicting probability of slope failure in East Timor as being of “very high probabilities” occupies 8.6% of study site, The “high probabilities” occupies 73.7 %, “moderate class “ occupies 12.2%, “low probabilities” occupies 4%, and very low probabilities occupies 1.5% of the study site.
6.2 Future Subject
To predicted probability of slope failure in East Timor is limited and is not suitable for site specific evaluation. The reliability of the analysis result depends on a multitude of factors ranging from the quality of the data base, the introduction of potential errors associated with data entry to the limitations and assumptions inherent in the statistical techniques. In this study, a particular problem with uncertainty is that the 1:15,000-scale topographic condition cannot fully reflect the micro-topography conditions prerequisite for the slope failure because slope failure in the study area is characterized by small and bigger volumes that a slight change in micro-scale landform may have a strong influence on the slope failure. Another problem is the 1:350,000-scale geological map used in this study cannot fully reflect the distribution of residual soils that are of critical significance to the slope failure. Intensity of rainfall with a failure time are most important data to analyst of slope failure hazard but unfortunately at this time it is difficult to find out in East Timor, However, it difficult to assess whether climate is changing related to the environmental hazard like slope failure in East Timor. Therefore, for further study on investigation and analyzing of slope failure in East Timor, those data has to be considered.
110
REFERENCES
Anbalagan D. (1992): Landslide hazard evaluation and zonation mapping in mountainous terrain. Eng. Geol 32 : 269–277 Atkinson P.M., Massari R., (1998): Generalized linear modeling of landslide susceptibility in the Central Apennines, Italy. Comput Geosci 24(4):373–385 Baeza C., Corominas J., (2001): Assessment of shallow landslide susceptibility by means of multivariate statistical techniques. Earth Surf Process Landf 26:1251–1263 Bouma G., and Kobryn, H.,(2004):Vegetation covers change in East Timor, 1989-1999, Natural Resources Forum, in press.
Brahmana, Radja Karina and Emanuel, Ulu., 1996: The book of 20 years of Development of East Timor,Corps of the Indonesian Civil Service East Timor Province, Dili, East Timor
Carrara A., Cardinali M., Detti R., Guzzetti F., Pasqui V.,Reichenbach P., (1991): GIS Techniques and statistical models in evaluating landslide hazard. Earth Surf Proc Landforms 16: 427–445 Carrara A., Cardinali, M., Guzzetti, F., Reichenbach, P., (1995):GIS technology in mapping landslide hazard. In: Carrara, A., Guzzetti, F. (Eds.), Geographical Information Systems in Assessing Natural Hazards. Kluwer Academic Publishers, The Netherlands, pp. 135–175. Carrara A., Guzzetti F., Cardinali M., and Reichenbach P., (1999): Use of GIS technology in the prediction and monitoring of landslide hazard; Natural Hazards 20 117 – 135. Chung C.F., Fabbri A.G (1995) :Multivariate regression analysis for landslide hazard zonation. In: Carrara A, Guzzetti F (eds) Geographical information systems in assessing natural hazards. Kluwer, Dordrecht, pp 107–133 Chung CF, Fabbri AG (2003) :Validation of spatial prediction models for landslide hazard mapping. Nat Hazards 30:451–472 Clerici A, Perego S, Tellini C, Vescovi P (2002) :A procedure for landslide susceptibility zonation: by the conditional analysis method. Geomorphology 48:349–364 Cruden D.M., (1984),: A simple definition of a landslide. Bulletin international association for engineering geology, 43,27-29. Dai FC., Lee CF., Li J., Xu ZW., (2001) :Assessment of landslide susceptibility on the natural terrain of Lantau Island, Hong Kong. Environmental Geology 40:381–391
111
Dai, F.c., Lee, C.F., (2002), Landslide characteristics and slope instability modeling use GIS Lantau Island, Hong Kong, Geomorphology 42, 213 - 238 DeGraff, J., Romesburg, H., (1980). Regional landslide-susceptibility assessment for wildland management: a matrix approach. In: Coates, D., Vitek, J. (Eds.), Thresholds in Geomorphology. George Allen and Unwin, London, pp. 401–414. Dolcemascolo, G., (2003). ‘Climate Risk and Agriculture in Timor Loro’Sae’, paper presented at Seminar on Climate Change and Severe Weather Events in Asia and the Caribbean, Barbados, July 24-25. Donati L., Turrini M.C., (2002): An objective method to rank the importance of the factors predisposing to landslides with the GIS methodology, application to an area of the Apennines (Valnerina; Perugia, Italy). Eng Geol 63:277–289 Ercanoglu M, Gokceoglu C (2002):Assessment of landslide susceptibility for a landslideprone Area (north of Yenice, NW Turkey) by fuzzy approach. Environ Geol 41: 720–730 Faculdade de Arquitectura (Universidade Técnica de Lisboa), and GERTIL (Grupo de Estudos de Reconstrução de Timor Leste). (2002). Atlas de Timor Leste. LIDEL, Lisboa. Greenbaum, D., Tutton, M., Bowker, M., Browne, T., Buleka, J., Greally, K., Kuna, G., McDonald, A., Marsh, S., O’Connor, E., Tragheim, D., (1995): Rapid Methods of landslide Hazard Mapping: Papua New Guinea Case Study. British Geological Survey. Technical Report WC/95/27. Gokceoglu C, Sonmez H, Ercanoglu M (2000): Discontinuity controlled probabilistic slope failure risk maps of the Altindag (settlement) region in Turkey. Eng. Geol. 55:277–296 Guzzetti F, Carrarra A, Cardinali M, Reichenbach P (1999): Landslide hazard evaluation: a review of current techniques and their application in a multi-scale study, Central Italy. Geomorphology 31:181–216 Hosmer DW, Lemeshow S (2000) :Applied logistic regression. John Wiley & Sons, Inc., 375 pp Jibson WR, Edwin LH, John AM (2000): A method for producing digital probabilistic seismic landslide hazard maps. Eng Geol 58:271–289 Jordan, C., O’Connor, E., Marchant, A., Northmore, A., Greenbaum, D., McDonald, A., Kovacik, M., Ahmed, R., (2000):Rapid landslide susceptibility mapping using remote sensing and GIS modelling. Proc. 14th International Conference on Applied Geologic Remote Sensing, Las Vegas, pp. 113–120.
112
Joseph, F., et all, (2006):Multivariate data analysis, Prentice hall, New Jersey, USA. Kantor Statistik(1989): Timor Timur dalam angka. The LandSat-based study is RePPProt: Review of Phase 1 results: Maluku and Nusa Tenggara, Volume 1: Main report (Jakarta: Ministry of Transmigration, Directorate General of Settlement Preparation and Land Resources Department, ODNRI and ODA. Keefer, G. (2000): Report on the Restoration of Meteorological Network – Timor Loro’Sae. Report II,pp 110-111, United Nations Transitional Administration in East Timor, Dili Kleinbaum DG (1994) :Logistic regression-a self-learning text. Springer, Berlin Heidelberg New York. Lamb, P.C., (1996): Residual soils. In: Turner, A.K., Schuster, R.L. (Eds.), Landslides: Investigation and Mitigation, Special Report, vol. 247. National Academic Press, Washington, DC, pp. 507– 524. Lee, S., Min, K., (2001): Statistical analysis of landslide susceptibility at Yongin, Korea. Environ. Geol. 40:1095–1113 Lee S, Chwae U, Min K (2002a): Landslide susceptibility mapping by correlation between topography and geological structure: the Janghung area, Korea. Geomorphology 46:149– 162 Lee S, Choi J, Min K (2002b) :Landslide susceptibility analysis and verification using the Bayesian probability model. Environ Geol 43:120–131 Lee S, Choi U (2003c): Development of GIS-based geological hazard information system and its application for landslide analysis in Korea. Geosci J 7:243–252 Lee S, Ryu JH, Won JS, Park HJ (2004a) :Determination and application of the weights for landslide susceptibility mapping: using an artificial neural network. Eng Geol 71: 289–302 Lee S, Choi J, Min K (2004b): Probabilistic landslide hazard mapping using GIS and remote sensing data at Boun, Korea. Intl J Remote Sens 25:2037–2052 Lineback, M., Andrew,W., Aspinall, R., Custer, S., (2001): Assessing landslide potential using GIS, soil wetness modelling and topographic attributes, Payette River, Idaho. Geomorphology 37, 149– 165. Liu J G, Mason P J, Clerici N, Chen S A, Davis A, Miao F, Deng H and Liang L (2004): Landslide hazard assessment in the Three Gorges area of the Yangtze River using ASTER imagery: Zigui–Badong; Geomorphology 61: 171 – 187.
113
Lulseged, A., Yamagishi, H., (2005): The application of GIS – based on logistic regression for landslide susceptibility mapping in Kakuda-Yahiko Mountains, Central Japan, Geomorphology vol. 65, pp. 15 – 31 Luzi L, Pergalani F, Terlien MTJ (2000): Slope vulnerability to earthquakes at sub regional scale: using probabilistic techniques and geographic information systems. Eng Geol 58:313– 336 Mark RK, Ellen SD (1995) :Statistical and simulation models for mapping debris flow hazard. In: Carrara A, Guzzetti F (eds) Geographical information systems in assessing natural hazards. Kluwer, Dordrecht, pp 93–106 Marston, R., Miller, M., Devkota, L., 1998: Geoecology and mass movements in the Manaslu Ganesh and Langtang-Jural Himals, Nepal. Geomorphology 26, 139– 150
Metzner, Joachim K. 1977. Man and Environment in Eastern Timor. 1977. The Australian National University, Canberra, Australia Monk, K., & Yance de Fretes, G.R.-L. (1997). The Ecology of Nusa Tengara and Maluku., Oxford University Press, Oxford, United Kingdom.
Nagarajan R, Mukherjee A, Roy A and Khire M V (1998): Temporal remote sensing data and GIS application in landslide hazard zonation of part of Western Ghat, India; Intern. J. Remote Sensing 19(4) 573 – 585. Niemann KO, Howes DE (1991) :Applicability of digital terrain models for slope stability assessment. Int Inst Aerospace Survey Earth Sci J 1991–3 : 127–137 Ochoa, G., (1978):La influencia de la altitud sobre algunasn propiedades fı´sico-quı´micas de los suelos de los Andes venezolanos. Revista Geogra´fica XVI–XIX, 56–72. Ohlmacher, C.G., Davis, C.J., (2003):Using multiple regression and GIS technology to predict landslide hazard in northeast Kansas, USA, Engineering Geology 69, 331 - 343 Pachauri AK, Pant M (1992) Landslide hazard mapping based on geological attributes. Eng Geol 32: 81–100 Parise M, Jibson WR (2000): A seismic landslide susceptibility rating of geologic units based on analysis of characteristics of landslides triggered by the 17 January, 1994 Northridge, California Earthquake. Eng Geol 58:251–270 Rautelal P, Lakhera RC (2000): Landslide risk analysis between Giri and Tons rivers in Himachal Himalaya (India). International Journal of Applied Earth Observation and Geoinformation 2:153–160
114
Rece A, Capolongo D (2002): Probabilistic modeling of uncertainties in earthquakeinduced landslide hazard assessment. Comput Geosci 28:735–749 Remondo J, Gonzalez A, Diaz de Teran JR, Cendrero A, Fabbri AG, Chung CJF (2003): Validation of landslide susceptibility maps: examples and applications from a case study in Northern Spain. Nat Hazards 30:437–449 Richard Dikau, Denys runsden, Lothar Schrott and Maia-Laura Ibsen,(1996): Landslide recognition,” identification, movement and Courses, John Wiley & Sons Ltd, West Sussex, England. Romeo R., (2000): Seismically induced landslide displacements: a predictive model. Eng Geol 58:337–351 Rowbotham D, Dudycha DN (1998) GIS Modeling of slope stability in Phewa Tal watershed, Nepal. Geomorphology 26:151–170
Saldanha, E. (1999): Developing natural resources in East Timor. (un pub.) Strategic Development Planning for East Timor Conference, Melbourne 5-9 April 1999.
Sanyu Consultants.,(2001): The study on integrated agricultural development of East Timor: Progress report (Dili: Japan International Cooperation Agency [JICA], East Timor Transitional Administration [ETTA], UNTAET), pp. 3-35. Shou KJ, Wang CF (2003): Analysis of the Chiufengershan landslide triggered by the 1999 Chi-Chi earthquake in Taiwan. Eng Geol 68:237–250 Temesgen B, Mohammed MU, Korme T (2001): Natural hazard assessment using GIS and remote sensing methods, with particular reference to the landslides in the Wondogenet area, Ethiopia. Phys Chem Earth 26:665–675 Van Westen C J, Seijmonsbergen A C and Mantovani F.,(1999): Comparing landslide Hazard maps; Natural Hazards 20 137 – 158. Varnes DJ(1984):”landslde hazard zonation”: a review of pricimples and practice, Natural hazard no 3. UNESCO, Paris Vivas, L., (1992) :Los Andes Venezolanos. Academia Nacional de la Historia, Caracas. Wischmeier,W., Smith, D., (1978):Predicting Rainfall Erosion Losses—A Guide to Conservation Planning. U.S. Department of Agriculture Science and Education Administration, Washington, DC. Wu W, Sidle RC (1995) :A distributed slope stability model for steep forested basins. Water Resources 31(8):2097–211
115
Zhou CH, Lee CF, Li J, Xu ZW (2002): On the spatial relationship between landslides and causative factors on Lantau Island, Hong Kong. Geomorphology 43:197–207 Zhou G, Esaki T, Mitani Y, Xie M, Mori J (2003): Spatial probabilistic modeling of slope failure using an integrated GIS Monte Carlo simulation approach. Eng Geol 68: 373–386 Zêzere J L, De Brum Ferreira A and Rodrigues M L (1999): The role of conditioning and triggering factors in the occurrence of landslides: a case study in the area north of Lisbon Portugal; Geomorphology 30 133 – 146.
116
Appendix A: Physical data of slope failure and unfailure slope A.1 Physical data of slope failure
A.1.1 Bobonaro site
Number Min (m) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 107 46 77 107 92 107 168 92 92 77 122 153 46 31 46 46 61 61 46 61 46 46 61 46 46 92 47 78 233 153 Width Max (m) 230 77 77 199 107 168 168 168 153 153 230 153 77 46 92 46 61 61 46 107 122 46 77 62 46 92 109 124 233 184 Mean (m) 168 61 77 153 99 138 168 130 122 115 176 153 61 38 69 46 61 61 46 84 84 46 69 54 46 92 78 101 233 168 Length Horizontal (m) 214 230 92 490 153 306 306 168 138 122 61 153 77 77 138 92 77 92 92 61 153 138 138 77 77 61 155 78 31 77 inclined (m) 231 239 94 507 155 315 340 171 146 130 62 158 84 78 141 93 85 99 95 66 165 143 146 78 82 62 165 84 33 82 Min (m) 1138 510 856 750 838 800 850 1000 975 475 500 340 550 433 313 563 538 475 388 388 406 438 363 335 290 263 563 663 400 675 Elevation Max (m) 1225 578 875 880 860 875 998 1030 1025 518 513 378 585 448 344 575 575 513 413 413 469 475 413 350 319 275 620 695 413 705 Mean (m) 1181 544 866 815 849 838 924 1015 1000 497 506 359 568 440 328 569 556 494 400 400 438 456 388 343 304 269 591 679 406 690 Height difference (m) 88 68 19 130 23 75 148 30 50 43 13 38 35 15 31 13 38 38 25 25 63 38 50 15 29 13 58 33 13 30 Inclination angle Direction Type of Slope Failure *) NE NW E N SE SE N NW NE SE SE SE N SE SE N NE NE N N N N SE SE SE SE NE W W SW LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS Lithology **) Vegetation Landscape Cover ***) Topography
SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR IR IR IR IR IR IR SR SR SR SR SR SR SR SR
HT NV 1 NV G NV NV NV G G G G NV NV G G NV G G NV NV NV G G G G NV NV G LT
RIDGE RIDGE VALLEY RIDGE RIDGE VALLEY RIDGE RIDGE RIDGE VALLEY VALLEY RIDGE RIDGE RIDGE RIDGE VALLEY FLAT FLAT RIDGE RIDGE RIDGE RIDGE RIDGE RIDGE RIDGE FLAT RIDGE VALLEY VALLEY VALLEY
117
Slope failure data of Bobonaro study site (continued) 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 31 31 46 31 46 31 62 93 31 31 47 47 46 46 109 47 78 93 46 46 61 31 47 47 78 47 62 47 62 78 155 47 31 31 171 77 61 46 46 31 46 61 62 124 31 61 62 62 46 46 140 47 155 124 46 46 61 47 62 78 109 109 109 78 109 109 155 47 31 62 233 77 46 38 46 31 46 46 62 109 31 46 54 54 46 46 124 47 116 109 46 46 61 39 54 62 93 78 85 62 85 93 155 47 31 47 202 77 184 92 61 61 61 92 31 93 47 92 31 47 122 61 93 124 78 78 122 61 46 124 62 109 62 132 155 47 124 54 62 93 93 47 93 31 192 97 63 62 63 96 33 97 48 96 33 52 134 66 99 134 79 84 134 66 47 128 63 111 65 134 158 48 128 56 63 96 99 52 100 33 663 580 1219 1213 1125 530 428 413 438 530 405 415 395 438 588 713 723 525 395 438 635 495 475 450 500 567 550 525 535 485 375 470 463 438 563 450 720 613 1235 1225 1140 558 438 440 450 558 415 438 450 463 620 763 740 558 450 463 644 525 488 475 518 590 582 538 565 500 388 495 495 460 600 463 691 596 1227 1219 1133 544 433 426 444 544 410 426 423 450 604 738 731 542 423 450 639 510 481 463 509 579 566 531 550 493 381 483 479 449 581 456 58 33 16 13 15 28 10 28 13 28 10 23 55 25 33 50 18 33 55 25 9 30 13 25 18 23 32 13 30 15 13 25 33 23 38 13 17 19 15 12 14 17 18 16 15 17 18 26 24 22 19 22 13 23 24 22 11 14 11 13 16 10 12 15 14 15 11 15 19 26 22 22 N S SE SE SE NW W W W NW W W NE NE SE SE SE E NE NE S SW SW SW SW SW SW SW SW SW SW NE NE E E S LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS SR SR SR SR SR SR SR SR SR SR SR VR VR VR VR VR VR IR IR IR IR SR SR SR SR SR SR SR SR SR SR IR IR IR SR SR LT G LT LT HT LT G G G LT G G NV NV LT NV G NV NV NV G G LT LT G NV NV NV NV LT NV NV G G NV G VALLEY RIDGE RIDGE RIDGE RIDGE RIDGE VALLEY RIDGE RIDGE RIDGE VALLEY VALLEY RIDGE RIDGE RIDGE VALLEY RIDGE VALLEY RIDGE RIDGE RIDGE RIDGE RIDGE RIDGE VALLEY VALLEY VALLEY VALLEY VALLEY RIDGE RIDGE RIDGE VALLEY RIDGE RIDGE RIDGE
118
Slope failure data of Bobonaro study site (continued) 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 124 124 124 78 78 93 31 47 109 78 248 109 61 77 77 77 31 46 77 31 93 264 92 31 46 61 77 77 92 31 31 47 93 78 155 46 124 155 155 124 124 186 31 47 171 140 279 155 153 77 77 92 46 46 77 31 124 310 92 46 77 61 77 122 92 47 62 62 140 124 186 61 124 140 140 101 101 140 31 47 140 109 264 132 107 77 77 84 38 46 77 31 109 287 92 38 61 61 77 99 92 39 47 54 116 101 171 54 155 62 62 93 140 93 109 124 279 279 248 93 306 46 61 46 92 92 138 78 54 279 122 92 153 61 46 61 61 47 47 62 202 186 62 77 165 67 67 96 148 98 111 128 289 283 259 100 314 48 62 48 95 95 149 80 56 286 150 105 165 72 54 72 72 53 53 72 231 224 80 85 705 550 738 725 600 700 515 500 625 550 475 500 750 1025 575 550 500 488 444 582 813 688 1525 550 1288 1150 698 575 875 400 400 575 500 550 500 513 763 575 763 750 650 730 538 530 700 600 550 538 819 1038 588 563 525 513 500 600 825 750 1613 600 1350 1188 725 613 913 425 425 613 613 675 550 550 734 563 750 738 625 715 526 515 663 575 513 519 784 1031 581 556 513 500 472 591 819 719 1569 575 1319 1169 711 594 894 413 413 594 556 613 525 531 58 25 25 25 50 30 23 30 75 50 75 38 69 13 13 13 25 25 56 18 13 63 88 50 63 38 28 38 38 25 25 38 113 125 50 38 20 22 22 15 20 18 12 14 15 10 17 22 13 15 12 15 15 15 22 13 13 13 36 29 22 31 31 31 31 28 28 31 29 34 39 26 E NE NE N N N NE NE NE NE NE NE NE N S S N N N W NW E NE SE NE SE NE N N W W E E E E SW LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS SF SF SF SF SF SF SF SF SF SF SF SF SF SF SR SR SR SR SR SR SR SR IR IR IR SR SR SR SR SR VR VR VR VR VR VR VR VR IR IR IR IR SR SR SR SR SR SR SR SR G G G NV G NV G NV NV NV NV NV NV G G G G G G LT LT NV HT G LT HT G LT G G G LT NV NV G LT RIDGE VALLEY RIDGE VALLEY RIDGE RIDGE RIDGE VALLEY VALLEY VALLEY VALLEY RIDGE RIDGE RIDGE RIDGE RIDGE RIDGE VALLEY FLAT RIDGE RIDGE RIDGE FLAT RIDGE RIDGE VALLEY RIDGE RIDGE RIDGE VALLEY RIDGE RIDGE RIDGE RIDGE RIDGE FLAT
119
Slope failure data of Bobonaro study site (continued) 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 109 77 93 31 46 124 77 186 47 46 31 124 31 31 78 62 78 45 77 31 168 77 92 47 155 124 47 155 62 76 31 92 46 536 138 138 186 153 140 62 61 264 153 326 109 61 31 202 31 31 109 78 78 45 107 107 383 122 122 47 217 186 62 217 109 106 31 245 92 138 138 168 147 115 116 47 54 194 115 256 78 54 31 163 31 31 93 70 78 45 92 69 275 99 107 47 186 155 54 186 85 91 31 168 69 337 138 153 124 77 93 93 77 140 77 124 78 31 39 93 31 93 186 62 39 61 77 122 153 122 107 62 47 78 78 124 47 61 31 77 61 61 61 122 145 85 107 112 85 172 85 149 98 36 54 112 34 102 206 77 46 68 88 149 172 137 134 69 61 100 87 166 63 79 40 91 72 72 72 141 600 388 563 700 513 600 388 563 530 475 500 388 425 483 425 830 575 860 838 790 938 863 920 475 650 588 600 590 538 850 1025 563 413 550 513 398 675 425 615 763 550 700 425 645 590 494 538 450 440 525 513 875 600 892 880 875 1015 925 1000 505 690 650 640 700 580 900 1050 613 450 588 550 468 638 406 589 731 531 650 406 604 560 484 519 419 433 504 469 853 588 876 859 833 976 894 960 490 670 619 620 645 559 875 1038 588 431 569 531 433 75 38 53 63 38 100 38 83 60 19 38 63 15 43 88 45 25 32 43 85 78 63 80 30 40 63 40 110 43 50 25 50 38 38 38 70 31 26 29 34 26 36 26 34 38 31 44 34 26 25 25 36 33 28 29 35 27 27 37 26 41 39 27 42 42 40 39 33 31 31 31 30 NE NE NE NE SW NW NE NE NE N NE NE N NE NE NW NE NE N SE N N N W N NE NW SW W SW N N N NE N N SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR IR IR IR IR IR SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR NV NV LT NV LT NV NV NV NV G G NV G G NV G G HT HT NV G NV NV G NV NV G NV G HT LT G G LT G NV RIDGE RIDGE VALLEY VALLEY FLAT VALLEY RIDGE RIDGE RIDGE RIDGE VALLEY VALLEY FLAT RIDGE VALLEY RIDGE RIDGE VALLEY RIDGE VALLEY RIDGE RIDGE RIDGE RIDGE VALLEY VALLEY RIDGE VALLEY RIDGE VALLEY RIDGE RIDGE VALLEY RIDGE RIDGE RIDGE
120
Slope failure data of Bobonaro study site (continued) 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 31 92 46 77 61 31 202 47 30 61 76 61 31 62 46 233 124 109 92 153 46 54 61 61 61 61 46 47 31 61 138 46 77 77 31 248 62 30 91 91 153 31 78 46 264 155 109 122 46 61 54 77 61 61 61 46 78 31 46 115 46 77 69 31 225 54 30 76 83 107 31 70 46 248 140 109 107 99 54 54 69 61 61 61 46 62 31 61 61 92 38 107 47 62 62 121 242 182 184 23 155 23 186 47 23 61 61 46 77 46 77 77 46 77 93 78 71 77 105 44 127 60 72 80 149 273 232 197 26 172 25 211 53 26 66 67 52 85 51 83 83 50 85 101 81 403 438 488 388 400 613 875 750 888 838 775 931 1013 425 1050 750 738 588 1100 518 425 475 463 494 388 406 363 465 650 438 485 538 410 469 650 913 800 975 963 920 1003 1025 500 1060 850 763 600 1125 544 450 513 485 525 419 425 400 505 675 420 461 513 399 434 631 894 775 931 900 848 967 1019 463 1055 800 750 594 1113 531 438 494 474 509 403 416 381 485 663 35 48 50 23 69 38 38 50 88 125 145 71 13 75 10 100 25 13 25 26 25 38 23 31 31 19 38 40 25 30 38 29 30 33 39 31 39 36 27 39 21 29 26 24 28 28 28 22 23 29 26 26 22 22 22 26 23 18 N N N N SE SW NE NE NE NE NE NE N NE N NE NE NE NE SE N N N N N N SE SE SE SF SF SF SF SF SF SF SF SF SF SF MIX MIX MIX MIX MIX MIX MIX MIX MIX MIX MIX MIX MIX MIX MIX MIX MIX MIX SR SR SR VR VR VR VR VR VR VR VR SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR G G NV G G G G NV HT NV NV G G NV LT NV NV NV G G G G NV G G G G G G RIDGE RIDGE VALLEY FLAT RIDGE RIDGE RIDGE RIDGE VALLEY VALLEY VALLEY RIDGE RIDGE VALLEY RIDGE RIDGE RIDGE VALLEY RIDGE VALLEY RIDGE RIDGE VALLEY RIDGE VALLEY FLAT RIDGE RIDGE RIDGE
*)Types of Slope Failure : - SF = Surface Failure; LS = Landslide, MIX = Landslide and Surface Failure **) Lithology : - SR = Sedimentary rocks; LR = Littoral deposit rocks; IR = Igneous rocks, MR = Metamorphic rocks; VR = Volcanics rocks ***)Vegetation Cover : - HT =High tree; LT= Low Tree; G= Grass, NV=No Vegetation
121
A.1.2 Cailaco site
Number Min (m) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 62 78 109 47 109 78 62 78 78 62 31 47 62 124 109 155 47 47 47 31 248 109 140 31 31 93 47 31 47 47 47 47 Width Max (m) 233 155 155 78 171 78 62 78 186 62 31 78 93 140 140 202 47 78 62 62 279 124 388 47 62 171 47 62 93 47 62 78 Mean (m) 147 116 132 62 140 78 62 78 132 62 31 62 78 132 124 178 47 62 54 47 264 116 264 39 47 132 47 47 70 47 54 62 Length Horizontal (m) 310 248 279 62 124 47 39 47 372 47 31 140 124 62 78 93 171 155 310 186 140 202 140 62 155 434 217 217 341 124 101 217 inclined (m) 334 267 283 67 133 48 44 48 392 48 33 144 130 67 80 96 182 167 338 201 148 211 148 65 159 441 227 233 350 128 112 226 Min (m) 425 450 600 625 663 650 655 638 363 400 363 325 350 250 305 310 363 438 515 575 260 275 300 305 315 345 300 315 325 500 750 738 Elevation Max (m) 550 550 650 650 710 663 675 650 488 413 375 363 388 275 325 335 425 500 650 650 310 338 350 325 350 425 365 400 405 530 800 800 Mean (m) 488 500 625 638 686 656 665 644 425 406 369 344 369 263 315 323 394 469 583 613 285 306 325 315 333 385 333 358 365 515 775 769 Height difference (m) 125 100 50 25 48 13 20 13 125 13 13 38 38 25 20 25 63 63 135 75 50 63 50 20 35 80 65 85 80 30 50 63 Inclination angle 22 22 10 22 21 15 27 15 19 15 22 15 17 22 14 15 20 22 24 22 20 17 20 18 13 10 17 21 13 14 26 16 SE NE N SE SE NE NE NE NE NE NE NE E E E E NE NE NE NE N NE NW NW N N NW NW NE NE E N Direction Type of Slope Failure*) LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS Lithology **) LR LR LR LR LR LR SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR LR LR LR LR Vegetation Cover ***) NV NV NV NV NV NV G G NV G G NV G LT NV NV G G NV NV G NV NV G G NV G LT G G G NV Landscape of Slope RIDGE RIDGE RIDGE VALLEY VALLEY RIDGE RIDGE VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY RIDGE RIDGE VALLEY VALLEY VALLEY VALLEY RIDGE VALLEY RIDGE RIDGE RIDGE RIDGE
122
Slope failure data of Cailaco study site (continued) 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 47 124 47 109 78 47 140 47 31 31 47 31 62 62 78 31 62 186 155 47 140 31 31 78 47 93 31 47 93 47 93 62 264 93 78 93 78 140 93 140 124 62 155 109 47 62 62 62 78 202 171 109 93 248 233 47 310 47 31 186 109 171 47 124 186 93 186 78 295 155 202 155 62 132 70 124 101 54 147 78 39 47 54 47 70 132 124 70 78 217 194 47 225 39 31 132 78 132 39 85 140 70 140 70 279 124 140 124 78 78 93 326 217 124 31 93 93 124 124 93 171 93 186 124 109 93 93 124 155 140 93 341 310 341 93 217 434 124 155 140 264 465 202 341 80 83 95 331 221 128 33 96 99 134 126 99 188 101 191 128 110 101 99 130 161 143 96 345 316 355 95 223 452 129 159 142 271 492 212 363 810 870 838 910 963 1005 920 1045 925 830 975 945 850 785 245 235 300 315 360 338 380 375 390 350 400 480 450 500 600 525 600 638 650 750 700 840 890 895 950 995 1015 945 1080 975 853 1010 1025 890 830 275 250 340 350 400 380 413 400 440 413 500 500 500 625 635 560 625 700 810 815 825 825 880 866 930 979 1010 933 1063 950 841 993 985 870 808 260 243 320 333 380 359 396 388 415 381 450 490 475 563 618 543 613 669 730 783 763 20 30 20 58 40 33 10 25 35 50 23 35 80 40 45 30 15 40 35 40 43 33 25 50 63 100 20 50 125 35 35 25 63 160 65 125 14 21 12 10 10 15 18 15 21 22 10 21 25 23 14 14 8 23 21 18 15 13 15 8 11 16 12 13 16 16 13 10 13 19 18 20 N NE E E NW N N E NE NE NE E NE NE NE NE NE NW NW NW NW NW NW N NW NW N N N N N NE NE N NW NE LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LR LR LR LR LR LR LR LR LR LR LR LR LR LR LR LR LR LR LR LR LR SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR NV NV NV NV NV NV NV NV NV G NV NV NV NV NV NV NV G G G G NV NV NV NV NV G NV NV NV NV NV NV G G G VALLEY VALLEY VALLEY VALLEY VALLEY RIDGE RIDGE RIDGE VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY
123
Slope failure data of Cailaco study site (continued) 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 93 109 62 62 62 47 62 155 62 62 47 124 47 93 78 78 47 47 47 62 47 31 47 47 47 93 31 47 47 62 31 109 78 109 93 93 171 140 109 78 109 93 78 202 78 62 47 124 47 140 140 78 47 62 47 62 47 31 93 78 47 171 47 140 47 93 47 171 124 171 124 155 132 124 85 70 85 70 70 178 70 62 47 124 47 116 109 78 47 54 47 62 47 31 70 62 47 132 39 93 47 78 39 140 101 140 109 124 186 186 124 124 186 140 186 140 109 124 62 85 62 171 248 109 109 186 78 124 93 78 186 62 93 62 54 62 93 47 62 78 78 109 47 310 188 188 126 126 190 142 189 144 111 126 64 92 65 181 255 111 111 197 81 132 99 81 195 64 99 77 67 72 106 55 67 87 97 130 55 349 415 375 345 300 250 240 265 325 350 275 350 355 1180 790 675 600 425 775 755 565 500 650 513 338 605 335 638 613 290 325 310 980 953 1140 1105 440 400 365 338 275 275 300 350 375 290 385 375 1240 850 700 625 490 800 800 600 525 710 530 370 650 375 675 663 320 350 350 1038 1025 1170 1265 428 388 355 319 263 258 283 338 363 283 368 365 1210 820 688 613 458 788 778 583 513 680 521 354 628 355 656 638 305 338 330 1009 989 1155 1185 30 25 25 20 38 25 35 35 25 25 15 35 20 60 60 25 25 65 25 45 35 25 60 18 33 45 40 38 50 30 25 40 58 73 30 160 9 8 11 9 11 10 11 14 13 11 14 22 18 19 14 13 13 19 18 20 21 18 18 16 19 36 36 31 28 33 22 27 37 34 33 27 E NE NE NE NE NE N N N NW E NE NE NE NW NE SE NE NE N NW NE NE NE NE NW NE NE E E E NW E E NE NE LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS SF SF SF SF SF SF SF SF SF SF SF SR SR SR SR SR SR SR SR SR SR SR LR LR LR LR LR LR LR LR LR LR LR LR LR LR LR SR SR SR SR SR SR SR SR SR SR NV NV NV NV NV G G NV NV LT G G G NV NV G NV G G G G NV G G G G G NV G LT G G NV NV LT NV FLAT FLAT FLAT FLAT FLAT FLAT FLAT FLAT FLAT VALLEY VALLEY RIDGE RIDGE VALLEY RIDGE VALLEY RIDGE VALLEY RIDGE RIDGE VALLEY VALLEY VALLEY FLAT FLAT RIDGE VALLEY VALLEY RIDGE VALLEY VALLEY RIDGE VALLEY VALLEY VALLEY VALLEY
124
Slope failure data of Cailaco study site (continued) 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 78 47 31 47 47 16 47 47 47 78 78 93 47 31 78 62 62 78 47 62 31 31 78 78 124 93 47 140 78 155 62 47 47 47 47 109 78 109 186 140 155 47 62 124 62 62 78 47 93 78 47 140 78 186 202 62 140 124 116 54 39 47 47 31 78 62 78 132 109 124 47 47 101 62 62 78 47 78 54 39 109 78 155 147 54 140 101 54 78 78 78 62 140 171 248 155 140 109 54 93 109 54 62 62 54 62 47 93 109 124 47 93 93 62 70 47 66 92 92 88 72 165 214 290 177 163 135 62 115 124 66 71 80 67 71 53 101 123 139 51 106 106 67 79 51 268 250 250 940 950 315 410 450 540 500 870 250 338 1000 475 475 870 435 450 650 400 338 550 905 375 500 425 713 330 305 300 300 982 988 404 540 600 625 585 950 280 405 1060 513 510 920 475 485 675 440 395 613 925 425 550 450 750 350 287 275 275 961 969 360 475 525 583 543 910 265 371 1030 494 493 895 455 468 663 420 367 581 915 400 525 438 731 340 37 50 50 42 38 89 130 150 85 85 80 30 68 60 38 35 50 40 35 25 40 57 63 20 50 50 25 38 20 34 33 33 28 31 33 37 31 29 31 36 29 36 29 35 29 39 36 29 28 23 28 27 23 28 28 22 28 23 NW NW NW E NE NW NW NW NW E NE NE NE E SE NE NE SE NE SE NW NW E SE SE NW SE E SE SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF MIX MIX MIX MIX MIX MIX MIX MIX MIX MIX SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR LT LT LT G NV G G G NV NV NV G NV NV LT NV G NV G NV NV NV NV G NV G NV NV LT VALLEY RIDGE RIDGE VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY RIDGE VALLEY VALLEY VALLEY FLAT RIDGE VALLEY VALLEY VALLEY VALLEY RIDGE RIDGE FLAT VALLEY VALLEY
*)Types of Slope Failure : - SF = Surface Failure; LS = Landslide, MIX = Landslide and Surface Failure **) Lithology : - SR = Sedimentary rocks; LR = Littoral deposit rocks; IR = Igneous rocks, MR = Metamorphic rocks; VR = Volcanics rocks ***)Vegetation Cover : - HT =High tree; LT= Low Tree; G= Grass, NV=No Vegetation
125
A.1.3 Zumalai site
Number Min (m) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 30 30 61 76 61 61 45 30 45 45 30 45 30 30 30 45 45 15 30 15 30 30 45 30 15 45 30 45 61 45 45 136 Width Max (m) 45 61 121 136 106 91 91 30 45 45 61 45 45 45 45 61 61 30 45 45 45 76 61 30 30 76 45 61 76 76 91 167 Mean (m) 38 45 91 106 83 76 68 30 45 45 45 45 38 38 38 53 53 23 38 30 38 53 53 30 23 61 38 53 68 61 68 151 Length Horizontal (m) 61 106 106 151 151 182 212 121 167 61 151 106 45 30 167 212 242 45 45 136 136 182 38 45 38 61 76 38 121 197 121 30 inclined (m) 64 109 113 160 155 186 218 124 170 62 154 109 48 33 174 227 258 47 47 151 145 189 40 47 40 62 77 40 124 206 127 32 Min (m) 500 525 450 463 485 500 525 563 590 410 490 615 600 650 513 520 550 313 338 550 500 538 450 413 288 315 388 425 350 468 463 450 Elevation Max (m) 520 550 490 515 520 540 575 590 625 425 520 640 615 663 563 600 640 325 350 615 550 590 463 425 300 330 403 438 375 530 500 460 Mean (m) 510 538 470 489 503 520 550 576 608 418 505 628 608 656 538 560 595 319 344 583 525 564 456 419 294 323 395 432 363 499 481 455 Height difference (m) 20 25 40 53 35 40 50 28 35 15 30 25 15 13 50 80 90 13 13 65 50 53 13 13 13 15 15 13 25 62 38 10 Inclination angle 18 13 21 19 13 12 13 13 12 14 11 13 18 22 17 21 20 15 15 26 20 16 18 15 18 14 11 19 12 17 17 18 E SE W SW SW SW SW SE SW NE SW SE SE SW E NE NE NE NE N N NE W E E SW SW NW SE W E NW Direction Type of Slope Failure*) LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS Lithology **) SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR LR LR LR LR LR LR LR LR LR LR Vegetation Landscape Cover ***) Topography LT LT NV NV G G G G G LT G NV NV NV HT NV NV LT G LT LT NV HT HT HT LT LT G LT LT LT G
Valley Valley Valley Valley Valley Valley Valley Valley Valley Valley Flat Valley Valley Valley Valley Valley Valley Valley Valley Valley Ridge Ridge Ridge Ridge Ridge Ridge Flat Valley Valley Valley Valley Valley
126
Slope failure data of Zumalai study site (continued) 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 45 30 30 30 45 15 76 136 61 45 76 15 30 30 76 45 45 45 45 61 76 151 30 30 61 30 30 45 15 30 30 30 30 30 45 76 30 30 61 61 30 91 167 76 121 212 30 30 30 106 61 76 61 61 61 106 293 61 30 76 61 30 45 30 45 30 30 30 45 76 61 30 30 45 53 23 83 151 68 83 144 23 30 30 91 53 61 53 53 61 91 222 45 30 68 45 30 45 23 38 30 30 30 38 61 91 76 45 91 151 61 45 121 45 182 30 76 45 76 30 38 45 30 30 30 45 61 76 30 45 151 30 30 45 38 45 30 76 76 53 94 81 47 94 160 64 51 125 48 194 38 91 54 94 36 48 52 39 38 33 53 79 81 33 59 171 33 33 48 45 56 39 81 98 73 475 413 525 400 338 425 563 360 385 400 635 600 620 325 310 590 525 490 388 438 388 475 525 588 500 550 525 538 435 525 438 450 485 475 550 500 440 538 425 390 445 585 390 400 468 658 650 650 380 330 620 550 515 410 450 415 525 555 600 538 630 538 550 450 550 470 475 515 538 600 488 426 531 413 364 435 574 375 393 434 647 625 635 353 320 605 538 503 399 444 401 500 540 594 519 590 531 544 443 538 454 463 500 507 575 25 28 13 25 53 20 23 30 15 68 23 50 30 55 20 30 25 25 23 13 28 50 30 13 38 80 13 13 15 25 33 25 30 63 50 15 20 15 15 19 18 26 14 18 21 37 33 33 36 33 38 29 40 37 22 31 40 22 22 40 28 22 22 18 33 36 40 22 40 43 NE NE NE NE NE NE SE SW SW NE SW NW NW SE NW NE SW SE NE NW SE SW SW NE NE NE NE NE NE NE NE NE NE N N LS LS LS LS LS LS LS LS LS LS SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF LR LR SR SR SR SR SR SR SR SR SR SR SR SR LR LR LR LR LR LR SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR LT LT G G NV NV LT G G G LT NV NV HT LT NV G LT HT G LT LT LT LT NV LT HT LT LT LT LT LT G LT G
Flat Ridge Ridge Ridge Ridge Ridge Ridge Ridge Ridge Ridge Ridge Ridge Valley Valley Valley Valley Valley Valley Valley Valley Valley Valley Valley Valley Valley Valley Valley Valley Valley Valley Valley Valley Valley Valley Valley
127
Slope failure data of Zumalai study site (continued) 68 69 70 71 72 73 74 75 61 45 76 61 45 121 45 45 61 45 91 76 61 151 45 106 61 45 83 68 53 136 45 76 30 23 38 61 151 38 30 30 39 27 44 69 165 45 36 36 600 410 388 475 430 475 770 300 625 425 410 508 495 500 790 320 613 418 399 492 463 488 780 310 25 15 23 33 65 25 20 20 40 33 31 29 23 33 33 33 SE SW SW NE NE NW SE NE SF SF SF SF SF SF SF SF SR SR SR SR SR SR SR SR LT LT G LT LT LT NV G
Valley Valley Valley Valley Ridge Ridge Ridge Valley
*)Types of Slope Failure : - SF = Surface Failure; LS = Landslide, MIX = Landslide and Surface Failure **) Lithology : - SR = Sedimentary rocks; LR = Littoral deposit rocks; IR = Igneous rocks, MR = Metamorphic rocks; VR = Volcanics rocks ***)Vegetation Cover : - HT =High tree; LT= Low Tree; G= Grass, NV=No Vegetation
A.1.4 Atsabe Site
Number Min (m) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 47 62 78 78 62 93 62 31 47 124 78 78 47 78 78 31 Width Max (m) 155 62 109 109 78 155 93 31 62 186 93 93 62 140 171 31 Mean (m) 101 62 93 93 70 124 78 31 54 155 85 85 54 109 124 31 Length Horizontal (m) 403 109 155 78 341 186 155 62 93 434 140 124 124 93 155 47 inclined (m) 410 112 163 81 349 189 159 63 100 438 142 126 126 100 167 48 Min (m) 500 485 388 388 375 350 365 388 438 413 460 500 550 550 575 313 Elevation Max (m) 575 513 438 413 450 385 400 400 475 475 488 525 575 588 638 325 Mean (m) 538 499 413 400 413 368 383 394 456 444 474 513 563 569 606 319 Height difference (m) 75 28 50 25 75 35 35 13 38 63 28 25 25 38 63 13 Inclination angle 11 14 18 18 12 11 13 11 22 8 11 11 11 22 22 15 NW NW NW NW NW NW NW NW NW NW NW NW NW NW NE SW Direction Type of Slope Failure*) LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS Lithology **) SR SR SR SR SR SR SR SR SR SR SR SR LR LR LR LR Vegetation Landscape Cover ***) Topography G LT G G G NV G G NV NV LT LT G G NV G VALLEY RIDGE RIDGE VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY RIDGE RIDGE RIDGE RIDGE RIDGE RIDGE RIDGE
128
Slope failure data of Atsabe study site (continued) 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 62 47 62 62 62 124 78 31 47 62 31 78 109 31 47 62 31 93 62 78 62 62 78 78 78 93 78 62 124 109 62 78 93 47 78 186 31 62 78 47 140 78 93 62 93 109 70 62 78 70 62 124 93 47 62 78 39 78 147 31 54 70 39 116 70 85 62 78 93 93 39 124 62 93 39 47 39 62 62 23 39 54 23 23 31 31 78 47 47 93 31 54 94 42 131 64 99 46 60 48 69 77 26 46 66 29 30 37 43 86 53 51 101 34 60 515 500 572 535 330 388 375 385 375 350 438 363 538 383 382 380 495 363 350 300 460 475 400 530 515 615 550 363 413 413 413 405 395 450 388 575 400 400 400 525 400 375 320 500 490 425 523 508 594 543 346 400 394 399 390 373 444 375 556 391 391 390 510 381 363 310 480 483 413 15 15 43 15 33 25 38 28 30 45 13 25 38 18 18 20 30 38 25 20 40 15 25 9 21 19 14 19 33 39 35 26 36 28 33 35 37 38 33 44 26 28 23 23 26 25 NE NW NW NW SW NW NW NW N N NW NW N SW SW SW NW NE SW NW NE NW NW LS LS LS LS LS SF SF SF SF SF SF SF SF SF SF SF SF MIX MIX MIX MIX MIX MIX LR LR LR LR LR LR LR LR LR LR SR SR SR SR SR SR SR SR SR LR LR LR LR G LT NV LT G G G G G G LT G G G G G G G G NV G LT LT RIDGE RIDGE VALLEY RIDGE RIDGE RIDGE RIDGE RIDGE VALLEY RIDGE RIDGE RIDGE RIDGE RIDGE RIDGE RIDGE RIDGE VALLEY RIDGE RIDGE RIDGE RIDGE RIDGE
*)Types of Slope Failure : - SF = Surface Failure; LS = Landslide, MIX = Landslide and Surface Failure **) Lithology : - SR = Sedimentary rocks; LR = Littoral deposit rocks; IR = Igneous rocks, MR = Metamorphic rocks; VR = Volcanics rocks ***)Vegetation Cover : - HT =High tree; LT= Low Tree; G= Grass, NV=No Vegetation
129
A.1.5 Maliana Site
Number Min (m) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 109 78 124 39 47 47 47 78 54 78 93 62 78 47 31 109 168 168 46 62 47 124 31 78 47 124 155 171 202 168 62 Width Max (m) 124 109 124 54 124 47 124 78 54 93 124 124 93 47 31 109 260 225 46 93 78 124 31 78 62 140 233 171 202 260 62 Mean (m) 116 93 124 47 85 47 85 78 54 85 109 93 85 47 31 109 214 197 46 78 62 124 31 78 54 132 194 171 202 214 62 Length Horizontal (m) 78 109 47 93 140 62 124 171 109 93 78 93 140 93 78 31 153 77 46 62 47 39 62 47 109 47 62 62 78 107 23 inclined (m) 81 114 48 96 142 63 133 186 113 100 81 100 147 95 81 33 190 91 58 72 60 43 68 60 125 58 80 72 90 129 28 Min (m) 475 475 450 825 675 650 863 575 894 875 695 650 600 794 463 463 1188 1275 1663 1125 1175 1275 435 413 450 463 575 475 700 1165 950 Elevation Max (m) 500 510 463 850 700 663 910 650 925 913 720 688 645 813 488 475 1300 1325 1698 1163 1213 1294 463 450 513 498 625 513 745 1238 965 Mean (m) 488 493 456 838 688 656 886 613 909 894 708 669 623 803 475 469 1244 1300 1680 1144 1194 1284 449 431 481 480 600 494 723 1201 958 Height difference (m) 25 35 13 25 25 13 48 75 31 38 25 38 45 19 25 13 113 50 35 38 38 19 28 38 63 35 50 38 45 73 15 Inclination angle 18 18 15 15 10 11 21 24 16 22 18 22 18 11 18 22 36 33 37 31 39 26 24 39 30 37 39 31 30 34 33 SW SW SW SE SE SE E SW SE S S SE SE SE E E NW NW NW SW SW SW E E E SW SW SW E E E Direction Type of Slope Failure*) LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS LS SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF Lithology **) SR SR SR SR SR SR SR SR SR SR LR LR LR LR LR LR LR LR LR LR LR SR SR SR SR SR SR SR SR SR SR Vegetation Landscape Cover ***) Topography G G G NV NV NV NV NV G G NV NV NV NV LT G LT LT HT LT LT LT NV NV NV G G LT G LT G RIDGE RIDGE RIDGE VALLEY RIDGE RIDGE VALLEY RIDGE RIDGE RIDGE RIDGE RIDGE RIDGE RIDGE VALLEY VALLEY VALLEY VALLEY VALLEY RIDGE VALLEY RIDGE VALLEY VALLEY VALLEY RIDGE VALLEY RIDGE RIDGE RIDGE VALLEY
130
A.1.6 Ainaro Site
Number Min (m) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 62 62 45 15 30 30 30 30 61 30 61 30 45 45 30 45 61 76 30 77 62 123 61 Width Max (m) 108 93 91 30 45 61 45 76 61 61 76 45 76 61 61 45 136 76 30 123 123 139 167 Mean (m) 85 77 68 23 38 45 38 53 61 45 68 38 61 53 45 45 98 76 30 100 93 131 114 Length Horizontal (m) 170 139 76 61 182 23 23 121 45 45 30 23 45 45 91 23 23 23 30 154 108 39 45 inclined (m) 189 149 81 66 188 27 30 165 59 61 39 26 59 56 118 26 30 26 39 187 149 46 61 Min (m) 1063 935 235 300 775 510 480 788 1163 1325 1538 1538 1350 813 900 1075 1000 950 900 695 713 800 200 Elevation Max (m) 1145 990 263 325 825 525 500 900 1200 1365 1563 1550 1388 845 975 1088 1020 963 925 800 815 825 240 Mean (m) 1104 963 249 313 800 518 490 844 1181 1345 1550 1544 1369 829 938 1081 1010 956 913 748 764 813 220 Height difference (m) 83 55 28 25 50 15 20 113 38 40 25 13 38 33 75 13 20 13 25 105 103 25 40 Inclination angle 26 22 20 22 15 33 41 43 40 41 40 29 40 36 40 29 41 29 40 34 44 33 41 SE NW NW NW NW W NW NW NW NW NW NW NW NW NW NW NW NW NW NE NE W W Direction Type of Slope Failure*) LS LS LS LS LS SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF SF Lithology **) LR LR LR LR LR LR LR LR LR SR SR SR SR SR SR SR SR SR SR SR SR SR SR Vegetation Landscape Cover ***) Topography G G G HT HT LT HT LT HT G G G G LT LT NV G G LT NV NV G NV VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY
*)Types of Slope Failure : - SF = Surface Failure; LS = Landslide, MIX = Landslide and Surface Failure **) Lithology : - SR = Sedimentary rocks; LR = Littoral deposit rocks; IR = Igneous rocks, MR = Metamorphic rocks; VR = Volcanics rocks ***)Vegetation Cover : - HT =High tree; LT= Low Tree; G= Grass, NV=No Vegetation
131
A.1.7 Hatolia Site
Number Min (m) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 47 93 62 78 78 109 47 78 78 47 62 62 62 140 140 47 62 140 62 31 Width Max (m) 47 140 78 109 186 264 109 109 124 78 78 78 93 279 171 47 93 140 109 31 Mean (m) 47 116 70 93 132 186 78 93 101 62 70 70 78 209 155 47 78 140 85 31 Length Horizontal (m) 54 124 124 54 93 62 62 186 39 62 93 109 310 62 62 62 47 54 109 62 inclined (m) 56 132 126 59 100 71 64 191 43 64 106 119 313 80 77 71 53 65 132 80 Elevation Min (m) 575 505 650 660 688 665 675 275 388 410 650 475 185 685 680 585 600 650 400 210 Max (m) 590 550 675 682 725 700 690 320 405 427 700 525 230 735 725 620 625 685 475 260 Mean (m) 583 528 663 671 706 683 683 298 396 419 675 500 208 710 703 603 613 668 438 235 Height difference (m) 15 45 25 22 38 35 15 45 18 17 50 50 45 50 45 35 25 35 75 50 Inclination angle 15 20 11 22 22 29 14 14 24 15 28 25 8 39 36 29 28 33 35 39 SW SW SE SE S S S SE W W S W W SE SE S S SW N S Direction Type of Slope Failure*) LS LS LS LS LS LS LS LS LS LS LS LS LS SF SF SF SF SF SF SF Lithology **) VR VR VR VR SR SR SR SR SR SR SR SR SR SR SR SR MR MR MR MR Vegetation Landscape Cover ***) Topography G G G G NV NV NV G LT LT G LT LT LT G G G G G LT RIDGE RIDGE VALLEY VALLEY VALLEY VALLEY VALLEY RIDGE RIDGE RIDGE RIDGE VALLEY RIDGE VALLEY VALLEY VALLEY VALLEY RIDGE VALLEY RIDGE
*)Types of Slope Failure : - SF = Surface Failure; LS = Landslide, MIX = Landslide and Surface Failure **) Lithology : - SR = Sedimentary rocks; LR = Littoral deposit rocks; IR = Igneous rocks, MR = Metamorphic rocks; VR = Volcanics rocks ***)Vegetation Cover : - HT =High tree; LT= Low Tree; G= Grass, NV=No Vegetation
132
A.1.8 Hatobuilico Site
Number Min (m) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 62 77 46 77 46 31 31 93 123 31 31 46 31 62 77 31 139 185 Width Max (m) 93 93 77 108 62 46 77 139 216 62 31 46 62 77 108 62 201 231 Mean (m) 77 85 62 93 54 39 54 116 170 46 31 46 46 69 93 46 170 208 Length Horizontal (m) 216 139 170 123 140 123 62 231 154 123 93 170 93 123 62 154 77 77 inclined (m) 225 148 180 126 144 144 79 263 179 151 119 227 112 153 74 184 87 92 Elevation Min (m) 800 800 1030 863 1290 1288 1250 1550 1650 1988 2025 2000 2000 1625 1585 1500 1000 1200 Max (m) 863 850 1090 890 1325 1363 1300 1675 1740 2075 2100 2150 2063 1715 1625 1600 1040 1250 831 825 1060 876 1308 1325 1275 1613 1695 2031 2063 2075 2031 1670 1605 1550 1020 1225 Height difference (m) 63 50 60 28 35 75 50 125 90 88 75 150 63 90 40 100 40 50 Inclination angle Direction Type of Slope Failure*) SE SE SE SE SE NE N SE SE N N N N NE NE NW SE SE LS LS LS LS LS SF SF SF SF SF SF SF SF SF SF SF SF SF Lithology **) Vegetation Landscape Cover ***) Topography
LR LR LR LR LR LR LR LR LR LR LR LR LR SR SR SR SR SR SR SR SR SR SR SR SR SR SR
LT LT G G G G LT LT LT HT G G NV NV NV G LT HT HT HT HT LT LT LT LT LT NV
RIDGE RIDGE FLAT FLAT RIDGE RIDGE FLAT FLAT FLAT RIDGE VALLEY FLAT FLAT VALLEY FLAT VALLEY RIDGE RIDGE VALLEY FLAT FLAT FLAT FLAT VALLEY VALLEY VALLEY VALLEY
SR SR SR SR SR SR SR SR SR SR SR SR IR IR IR IR IR IR IR IR IR IR IR IR IR IR IR IR IR IR IR IR IR IR IR SR
NV NV LT LT LT G G HT HT HT HT HT LT LT LT LT LT LT LT G G G LT LT G LT HT HT G G LT LT LT G G G
RIDGE RIDGE RIDGE RIDGE VALLEY VALLEY RIDGE RIDGE RIDGE RIDGE RIDGE VALLEY VALLEY FLAT FLAT RIDGE RIDGE RIDGE RIDGE RIDGE RIDGE VALLEY VALLEY VALLEY VALLEY RIDGE RIDGE VALLEY FLAT RIDGE RIDGE FLAT FLAT FLAT FLAT FLAT
SR SR SR SR SR SR SR SR SR SR IR IR IR IR IR IR IR IR IR IR IR IR IR IR IR IR IR IR IR LR LR LR LR LR LR LR
LT LT LT LT LT HT HT HT HT G G G LT HT HT G LT LT NV NV LT LT LT LT LT HT HT HT HT LT LT LT LT G G G
RIDGE VALLEY VALLEY VALLEY VALLEY VALLEY VALLEY RIDGE RIDGE RIDGE RIDGE RIDGE RIDGE VALLEY VALLEY VALLEY VALLEY FLAT FLAT RIDGE RIDGE RIDGE VALLEY RIDGE FLAT VALLEY FLAT RIDGE VALLEY FLAT RIDGE VALLEY VALLEY VALLEY VALLEY RIDGE
LR LR LR LR LR LR LR LR LR LR LR
SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR SR
LT LT LT NV NV NV LT LT LT G HT
LT LT G G NV NV NV LT LT LT LT G G LT LT LT LT LT
RIDGE RIDGE FLAT FLAT FLAT RIDGE RIDGE VALLEY VALLEY VALLEY VALLEY
VALLEY VALLEY VALLEY VALLEY VALLEY FLAT FLAT VALLEY RIDGE RIDGE RIDGE VALLEY VALLEY VALLEY RIDGE RIDGE RIDGE RIDGE
A.2.3 ZUMALAI SITE Number Min (m) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 61 45 30 45 76 76 91 76 76 91 61 61 91 91 61 76 76 61 106 76 76 106 45 106 106 61 45 61 61 76 76 91 Width Max (m) 91 76 76 61 106 91 121 91 91 121 76 76 121 106 106 121 106 91 136 106 91 121 76 136 185 76 91 91 91 106 121 106 Mean (m) 76 61 53 53 91 83 106 83 83 106 68 68 106 98 83 98 91 76 121 91 83 114 61 121 146 68 68 76 76 91 98 98 Length Horizontal inclined (m) (m) 61 61 45 76 45 61 106 61 61 61 76 45 91 61 121 106 61 121 45 61 61 76 106 61 76 151 121 61 76 76 136 76 62 62 48 95 51 63 111 63 63 66 81 53 93 70 140 126 63 125 53 66 62 77 109 68 92 165 128 65 81 87 147 82 Min (m) 465 428 303 340 325 330 375 400 425 403 540 505 403 490 415 445 425 505 490 403 453 440 365 403 490 486 480 518 543 468 480 543 Elevation Max (m) 480 443 318 398 348 348 408 418 443 428 568 533 420 526 486 513 443 538 518 428 468 456 393 433 543 550 520 540 570 510 535 575 Avr 473 435 310 369 336 339 391 409 434 415 554 519 411 508 450 479 434 521 504 415 460 448 379 418 516 518 500 529 556 489 508 559 Height difference (m) 15 15 15 58 23 18 33 18 18 25 28 28 18 36 71 68 18 33 28 25 15 16 28 30 53 65 40 23 28 43 55 33 Inclination Direction angle 14 14 18 37 26 16 17 16 16 22 20 31 11 30 30 32 16 15 31 22 14 12 15 26 35 23 18 20 20 29 22 23 SE W W SW SW SW SW SW SW NE NE NE NE NE SW SW SW SW E E E E E E E E E E SW SW SW NW Type of Slope Failure Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured LR LR LR LR LR LR LR LR LR LR LR LR LR LR LR LR LR LR LR LR LR LR LR LR LR LR LR LR LR LR LR LR Lithology *) Vegetation Landscape Cover **) HT HT HT HT LT LT LT LT LT LT LT LT LT LT HT HT HT HT HT HT LT LT LT LT LT LT LT LT LT HT HT HT
Topography
FLAT FLAT FLAT VALLEY VALLEY RIDGE RIDGE RIDGE RIDGE RIDGE VALLEY VALLEY VALLEY RIDGE RIDGE RIDGE RIDGE RIDGE RIDGE RIDGE FLAT FLAT FLAT VALLEY VALLEY RIDGE RIDGE RIDGE RIDGE VALLEY VALLEY RIDGE
143
Un-failure slopes data of Zumalai study site (continued) 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 106 76 61 76 61 61 121 61 45 61 61 45 61 45 61 76 76 61 45 61 61 76 61 45 76 76 91 61 45 45 61 45 76 61 76 45 182 106 106 106 91 91 151 76 91 76 91 76 76 76 76 121 106 106 61 91 136 106 106 121 136 106 121 76 76 91 76 76 91 76 91 61 144 91 83 91 76 76 136 68 68 68 76 61 68 61 68 98 91 83 53 76 98 91 83 83 106 91 106 68 61 68 68 61 83 68 83 53 76 136 121 151 121 167 61 91 76 76 106 61 76 136 121 151 76 61 121 121 61 151 121 61 76 76 61 76 61 76 76 76 91 106 61 61 80 141 128 160 125 171 62 95 92 82 109 63 77 146 147 177 82 62 139 132 73 161 147 62 77 78 67 81 70 81 82 77 95 119 65 65 653 503 518 543 580 608 468 493 618 638 633 618 668 545 553 583 623 620 583 533 533 570 583 558 570 468 558 445 470 483 518 558 433 370 458 333 678 540 560 595 610 645 480 520 670 670 660 635 683 598 635 675 655 635 650 585 573 625 665 573 585 485 585 475 505 510 550 573 460 425 480 355 665 521 539 569 595 626 474 506 644 654 646 626 675 571 594 629 639 628 616 559 553 598 624 565 578 476 571 460 488 496 534 565 446 398 469 344 26 38 43 53 30 38 13 28 53 33 28 18 15 53 83 93 33 15 68 53 40 55 83 15 15 18 28 30 35 28 33 15 28 55 23 23 19 15 19 19 14 13 12 17 35 23 15 16 11 21 34 31 23 14 29 23 33 20 34 14 11 13 24 22 30 20 23 11 17 27 20 20 NW NW NW E E E W SW SW SW S S S S S S S S SE SE SE SE SE SE S S S S S SE SE SE S S SE SW Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured LR SR SR SR SR SR SR SR SR SR SR SR SR SR IR IR IR IR IR IR IR IR IR IR IR IR IR IR IR IR IR IR IR IR IR IR HT HT HT G G G G G LT LT LT LT HT HT LT LT LT LT LT LT HT HT HT HT LT LT LT LT LT G G LT LT LT HT HT RIDGE RIDGE RIDGE RIDGE FLAT FLAT FLAT FLAT RIDGE RIDGE FLAT FLAT FLAT RIDGE VALLEY VALLEY RIDGE FLAT VALLEY VALLEY VALLEY VALLEY VALLEY FLAT FLAT FLAT RIDGE RIDGE RIDGE RIDGE RIDGE FLAT RIDGE VALLEY RIDGE RIDGE
144
Un-failure slopes data of Zumalai study site (continued) 69 70 71 72 73 74 75 45 61 76 61 45 61 76 76 76 106 121 76 76 106 61 68 91 91 61 68 91 106 76 76 61 61 45 61 107 77 102 82 66 54 67 345 370 510 585 805 635 598 360 385 578 640 830 665 625 353 378 544 613 818 650 611 15 15 68 55 25 30 28 8 11 42 42 22 33 24 SW SW NW NW S S S Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured Unfailured IR IR IR IR SR SR SR HT G HT LT LT LT LT FLAT FLAT VALLEY VALLEY RIDGE VALLEY VALLEY
A.2.5 MALIANA SITE
Number Min (m) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 107 118 122 61 124 62 78 93 109 109 93 93 62 93 62 62 93 155 78 62 62 78 78 93 93 109 78 78 93 47 78 Width Max (m) 122 138 153 92 155 93 62 62 124 62 109 62 140 155 47 47 62 140 93 93 93 124 116 109 78 155 62 47 109 109 93 Mean (m) 115 128 138 77 140 78 70 78 116 85 101 78 101 124 54 54 78 147 85 78 78 101 97 101 85 132 70 62 101 78 85 Length Horizontal inclined (m) (m) 92 107 69 77 93 78 78 78 70 109 62 78 78 93 124 109 78 109 109 47 47 155 171 47 47 62 93 93 62 78 62 114 131 82 82 98 87 79 83 84 115 72 84 86 95 127 111 86 113 117 52 55 157 172 55 51 70 98 99 69 95 63 Min (m) 1150 1200 1260 1648 473 473 448 460 573 473 573 1123 1173 1273 823 891 873 693 648 598 673 791 648 860 948 698 460 433 410 450 463 Elevation Max (m) 1218 1275 1305 1678 503 513 465 490 620 510 610 1155 1210 1290 850 915 910 723 690 620 703 815 670 890 968 730 490 465 440 505 475 Avr 1184 1238 1283 1663 488 493 456 475 596 491 591 1139 1191 1281 836 903 891 708 669 609 688 803 659 875 958 714 475 449 425 478 469 Height difference (m) 68 75 45 30 30 40 18 30 48 38 38 33 38 18 28 24 38 30 43 23 30 24 23 30 20 33 30 33 30 55 13 Inclination Direction angle 36 35 33 21 18 27 13 21 34 19 31 23 26 11 13 12 26 15 21 26 33 9 8 33 23 28 18 19 26 35 11 SE W W W W W W W W W W W W W E E E E N N N N N SE SE SE SE SE SE SW SW Type of Slope Failure unfailured unfailured unfailured unfailured Unfailured Unfailure Unfailure Unfailure Unfailure Unfailure Unfailure Unfailure Unfailure Unfailure Unfailure Unfailure Unfailure Unfailure Unfailure Unfailure Unfailure Unfailure Unfailure Unfailure Unfailure Unfailure Unfailure Unfailure Unfailure Unfailured Unfailure LR LR LR LR LR LR LR LR LR LR LR SR SR SR SR SR SR SR SR SR LR LR LR LR LR LR LR LR LR LR LR Lithology *) Vegetation Landscape Cover **) LT LT LT HT LT LT LT LT HT HT HT HT LT LT NV NV NV G G G G G LT LT LT LT HT HT HT LT LT
Topography
RIDGE VALLEY VALLEY VALLEY FLAT RIDGE RIDGE FLAT VALLEY VALLEY VALLEY VALLEY VALLEY FLAT FLAT FLAT RIDGE FLAT RIDGE VALLEY VALLEY FLAT FLAT VALLEY RIDGE VALLEY FLAT FLAT VALLEY VALLEY FLAT
Topography
VALLEY VALLEY VALLEY RIDGE VALLEY RIDGE RIDGE RIDGE RIDGE FLAT RIDGE VALLEY VALLEY VALLEY RIDGE VALLEY RIDGE VALLEY RIDGE RIDGE RIDGE RIDGE VALLEY
Appendix B : Logistic Regression Analysis B.1 All study site
Case Processing Summary Unweighted Cases(a) Selected Cases Included in Analysis Missing Cases Total Unselected Cases Total N 1012 0 1012 0 Percent 100.0 .0 100.0 .0
1012 100.0 a If weight is in effect, see classification table for the total number of cases. Dependent Variable Encoding Original Value .00 1.00 Classification Table(a) Observed Predicted status_slope .00 Step 1 status_slope Overall Percentage a The cut value is .500 Variables in the Equation B Step 1(a) elev_200_500 elev_500.1_800 elev_800.1_1100 elev_1100.1_1400 elev_1400.1_1700 Constant -1.903 -1.499 -1.004 -.756 -.742 S.E. 1.128 1.126 1.132 1.144 1.164 Wald 2.846 1.771 .787 .437 .406 df 1 1 1 1 1 Sig. .042 .013 .035 .008 .024 Exp(B) .149 .223 .366 .469 .476 4.461 .00 1.00 450 50 1.00 56 456 890.5 390.1 90.3 Percentage Correct Internal Value 0 1
1.495 1.123 1.772 1 .183 a Variable(s) entered on step 1: elev_200_500, elev_500.1_800, elev_800.1_1100, elev_1100.1_1400, elev_1400.1_1700.
151
Variables in the Equation B -1.111 -.708 -.218 .036 .021 S.E. .314 .316 .352 .367 .302 Wald 12.514 5.033 .385 .010 4.835 df 1 1 1 1 1 Sig. .000 .025 .035 .022 Exp(B) .329 .492 .804 1.036 1.021 2.020
Step number: 1 Observed Groups and Predicted Probabilities 320 ô ó F R E Q U E N C Y ó ó 240 ô ó ó ó1 160 ô0 ó0 ó0 ó0 80 ô0 ó0 ó0 ó0000 0010 010 10 10 10 10 110 01111 ô ó ó ó ô ó ó 0ó 1ô 1ó 1ó 1ó 1ô 1ó 1ó 111ó
Predicted òòòòòòòòòòòòòòôòòòòòòòòòòòòòòôòòòòòòòòòòòòòòôòòòòòòòòòòòòòòò Prob: 0 .25 .5 .75 1 Group: 000000000000000000000000000000111111111111111111111111111111 Unfailure Failure Predicted Probability is of Membership for 1.00 The Cut Value is .50 Symbols: 0 - .00 1 - 1.00 Each Symbol Represents 20 Cases.
164
B.2 Specific site B.2.1 Bobonaro site
Omnibus Tests of Model Coefficients Chi-square 280.423 280.423 280.423 Model Summary -2 Log likelihood 182.600(a) Cox & Snell R Square .568 Nagelkerke R Square .757 df 10 10 10 Sig. .000 .000 .000
Step 1
Step Block Model
Step 1
a Estimation terminated at iteration number 7 because parameter estimates changed by less than .001. Classification Table(a) Observed Predicted status_slope .00 Step 1 status_slope Overall Percentage a The cut value is .500 Variables in the Equation B -1.792 -1.588 2.485 S.E. .375 .342 1.057 Wald 22.831 21.524 5.530 df 1 1 1 Sig. .000 .000 .019 .000 Exp(B) .167 .204 12.000 3.000 .00 1.00 153 24 1.00 14 143 Percentage Correct
a Variable(s) entered on step 1: elev_500.1_800, elev_800.1_1100, elev_1100.1_1400, inc_ang6.0_12, inc_ang12.1_18, inc_ang18.1_24, inc_ang24.1_30, inc_ang30.1_36, inc_ang36.1_42, low_tree, grass, no_veg, S_R, L_R, I_R, north, northeast, east, southeast, southwest, west, Valley, Ridge.
176
Step number: 1 Observed Groups and Predicted Probabilities 80 ô ó ó F R E Q U E N C Y ó 60 ô ó ó ó 40 ô0 ó0 ó0 0 ó0 0 20 ô0 0 ó0 0 0 1 0 1 1 1 1 ô ó ó 1ó 1ô 1ó 1ó 1ó 1ô 1ó 1ó 1ó 1ô 11ó 11ó
ó0 010 0
ó0 000000 0 00 0 10 10 10 110 11110 10111ó Predicted òòòòòòòòòòòòòòôòòòòòòòòòòòòòòôòòòòòòòòòòòòòòôòòòòòòòòòòòòòòò Prob: 0 .25 .5 .75 1 Group: 000000000000000000000000000000111111111111111111111111111111 Unfailure Failure Predicted Probability is of Membership for 1.00 The Cut Value is .50 Symbols: 0 - .00 1 - 1.00 Each Symbol Represents 5 Cases.
177
B.2.2 Cailaco Site
Omnibus Tests of Model Coefficients Step 1 Step Block Model Chi-square 274.676 274.676 274.676 Model Summary -2 Log likelihood 94.079(a) Cox & Snell R Square .644 Nagelkerke R Square .859 df 7 7 7 Sig. .000 .000 .000
Step 1
a Estimation terminated at iteration number 8 because parameter estimates changed by less than .001. Hosmer and Lemeshow Test Step 1 Chi-square 8.090 df 8 Sig. .425
1.526 .493 9.565 1 a Variable(s) entered on step 1: elev_200_500, elev_500.1_800, elev_1100.1_1400.
Variables in the Equation B 3.486 5.088 -2.407 S.E. .433 .568 .330 Wald 64.688 80.219 53.146 df 1 1 1 Sig. .000 .000 .000 Exp(B) 32.647 162.060 .090
Step 1(a)
grass no_veg Constant
a Variable(s) entered on step 1: grass, no_veg. Variables in the Equation(96.2,54.9) B -1.766 2.847 S.E. .389 .501 Wald 20.594 32.327 .741 df 1 1 1 Sig. .000 .000 .389 Exp(B) .171 17.228 .847
Step 1(a)
low_tree no_veg Constant
-.166 .192 a Variable(s) entered on step 1: low_tree, no_veg.
Variables in the Equation B 2.256 S.E. .296 .175 Wald 58.023 25.236 df 1 1 Sig. .000 .000 Exp(B) 9.542 .414
-.881 a Variable(s) entered on step 1: S_R.
Step 1(a)
S_R Constant
179
Variables in the Equation B 1.154 S.E. .308 .142 Wald 14.084 3.605 df 1 1 Sig. .000 .058 Exp(B) 3.172 .763
Constant -.270 a Variable(s) entered on step 1: L_R.
Step 1(a)
L_R
Variables in the Equation B 2.376 4.021 2.097 2.799 S.E. .483 .515 .451 .463 Wald 24.250 61.016 21.601 36.494 df 1 1 1 1 1 Sig. .000 .000 .000 .000 .000 Exp(B) 10.762 55.753 8.139 16.427 .136
Step 1(a)
north northeast east northwest Constant
-1.997 .321 38.606 a Variable(s) entered on step 1: north, northeast, east, northwest. Variables in the Equation B 3.511 1.586 2.289 1.582 S.E. .476 .406 .419 .505 Wald 54.474 15.273 29.773 9.794 df
Step 1(a)
northeast east northwest southeast Constant
1 1 1 1 1
Sig. .000 .000 .000 .002 .000
Exp(B) 33.474 4.886 9.862 4.863 .226
-1.486 .254 34.234 a Variable(s) entered on step 1: northeast, east, northwest, southeast. Variables in the Equation B 1.509 .351 -.901 S.E. .374 .401 .329 Wald 16.296 .765 7.501 df
Step 1(a)
Valley Ridge Constant
1 1 1
Sig. .000 .032 .006
Exp(B) 4.521 1.420 .406
a Variable(s) entered on step 1: Valley, Ridge. Variables in the Equation B 1.158 -.351 S.E. .290 .401 Wald 15.949 .765 5.756 df 1 1 1 Sig. .000 .048 .016 Exp(B) 3.184 .704 .577
Step 1(a)
Valley Flat Constant
-.550 .229 a Variable(s) entered on step 1: Valley, Flat.
elev_200_500 elev_500.1_800 elev_1100.1_1400 inc_ang6.0_12 inc_ang12.1_18 inc_ang18.1_24 inc_ang30.1_36 low_tree L_R north northeast east northwest Flat Valley grass Constant
.200 1.100 .033 1 .856 1.222 a Variable(s) entered on step 1: elev_200_500, elev_500.1_800, elev_1100.1_1400, inc_ang6.0_12, inc_ang12.1_18, inc_ang18.1_24, inc_ang30.1_36, low_tree, L_R, north, northeast, east, northwest, Flat, Valley, grass.
187
Step number: 1 Observed Groups and Predicted Probabilities 80 ô ó F R E Q U E N C Y ó ó 60 ô0 ó0 ó0 ó0 40 ô0 ó0 ó0 ó0 20 ô0 ó0 ó0 ó0 0 0 0 0 0 0 10 10 10 10 101 11 1011 ô ó ó ó 1ô 1ó 1ó 1ó 1ô 1ó 1ó 1ó 1ô 1ó 1ó 111ó
Predicted òòòòòòòòòòòòòòôòòòòòòòòòòòòòòôòòòòòòòòòòòòòòôòòòòòòòòòòòòòòò Prob: 0 .25 .5 .75 1 Group: 000000000000000000000000000000111111111111111111111111111111 Unfailure Failure Predicted Probability is of Membership for 1.00 The Cut Value is .50 Symbols: 0 - .00 1 - 1.00 Each Symbol Represents 5 Cases.
188
B.2.3 Zumalai Site
Omnibus Tests of Model Coefficients Chi-square 71.559 71.559 71.559 Model Summary -2 Log likelihood Cox & Snell R Square Nagelkerke R Square df 7 7 7 Sig .000 .000 .000
Step 1
Step Block Model
136.385(a) .379 .506 a Estimation terminated at iteration number 6 because parameter estimates changed by less than .001. Hosmer and Lemeshow Test Step 1 Chi-square 3.819 df 8 Sig. .873
-1.224 .509 5.786 1 .016 a Variable(s) entered on step 1: inc_ang12.1_18, inc_ang18.1_24, inc_ang24.1_30, inc_ang30.1_36, inc_ang36.1_42, inc_ang42.1_48. Variables in the Equation B 1.145 2.228 3.327 S.E. .487 .597 .865 Wald 5.527 13.937 14.799 df 1 1 1 1 Sig. .019 .000 .000 .002 Exp(B) 3.143 9.286 27.857 .269
Step 1(a)
low_tree grass no_veg Constant
-1.312 .426 9.496 a Variable(s) entered on step 1: low_tree, grass, no_veg.
190
Variables in the Equation B -1.145 1.083 2.182 -.167 S.E. .487 .481 .789 .237 Wald 5.527 5.082 7.647 .499 df 1 1 1 1 Sig. .019 .024 .006 .480 Exp(B) .318 2.955 8.864 .846
Step 1(a)
high_tree grass no_veg Constant
a Variable(s) entered on step 1: high_tree, grass, no_veg. Variables in the Equation B 3.145 S.E. .446 .266 Wald 49.617 23.875 df 1 1 Sig. .000 .000 Exp(B) 23.222 .273
-1.299 a Variable(s) entered on step 1: S_R.
Step 1(a)
S_R Constant
Variables in the Equation B -.068 a Variable(s) entered on step 1: L_R. Step 1(a) L_R Constant .319 S.E. .401 .184 Wald .633 .136 df 1 1 Sig. .026 .713 Exp(B) 1.376 .934
f. direction
Variables in the Equation B 3.584 3.920 1.322 1.997 2.086 2.197 2.351 S.E. 1.344 .890 .916 .870 .817 1.106 .930 Wald 7.112 19.420 2.083 5.262 6.524 3.950 6.391 df 1 1 1 1 1 1 1 Sig. .008 .000 .049 .022 .011 .047 .011 Exp(B) 36.000 50.400 3.750 7.364 8.053 9.000 10.500 .111
Step 1(a)
north northeast east southeast southwest west northwest Constant
-2.197 .745 8.690 1 .003 a Variable(s) entered on step 1: north, northeast, east, southeast, southwest, west, northwest. Variables in the Equation B Step 1(a) north northeast east southeast south southwest northwest Constant 1.386 1.723 -.875 -.201 -2.197 -.111 .154 .000 S.E. 1.384 .950 .975 .932 1.106 .882 .988 .816 Wald 1.003 3.289 .807 .046 3.950 .016 .024 .000 df 1 1 1 1 1 1 1 1 Sig. .017 .040 .039 .030 .047 .021 .016 1.000
-13.239 3.506 14.262 1 .000 .000 a Variable(s) entered on step 1: elev_200_500, inc_ang6.0_12, inc_ang12.1_18, inc_ang18.1_24, inc_ang24.1_30, inc_ang30.1_36, inc_ang36.1_42, S_R, north, northeast, east, southeast, southwest, northwest, Valley, Ridge, west.
200
Step number: 1 Observed Groups and Predicted Probabilities 32 ô ó ó F R E Q U E N C Y ó 24 ô ó ó ó 16 ô ó ó ó 8 ô ó ó 00 00 00 00 00 00 00 00 00 001 00 1 1 11 1111 ô ó ó ó ô ó ó ó ô ó 11 ó 11 ó 11 ô 11 ó 1111 ó
ó 00 000 10 100 10 10 10 110 1111 1111 ó Predicted òòòòòòòòòòòòòòôòòòòòòòòòòòòòòôòòòòòòòòòòòòòòôòòòòòòòòòòòòòòò Prob: 0 .25 .5 .75 1 Group: 000000000000000000000000000000111111111111111111111111111111 Unfailure Failure Predicted Probability is of Membership for 1.00 The Cut Value is .50 Symbols: 0 - .00 1 - 1.00 Each Symbol Represents 2 Cases.