Luận văn thạc sĩ: Landslide hazard and risk assessment for road network using RS and GIS: A case study of Xin Man district, Vietnam

LANDSLIDE HAZARD AND RISK ASSESSMENT FOR ROADNETWORK USING RS AND GIS: A CASE STUDY OF XIN MAN

DISTRICT, VIET NAM

Lai Tuan Anh

A thesis submitted in partial fulfillment of the requirements for the

degree of Master of Engineering

Examination Committee: Dr Kiyoshi Honda (Chairperson)Dr Marc Souris

Dr Ulrich Glawe

Nationality: Vietnamese

Previous Degree: Bachelor of Engineering in Geodesy

Hanoi University of Mining and Geology,Vietnam

Scholarship Donor: AIT Fellowship

Asian Institute of Technology

School of Engineering TechnologyThailand

May 2006

ACKNOW LEDGEMENT

Tis with delight that the author first of all extends his heardes gratitude to the Thesis researchcommittee Chairperson, DR Kiyoshi Honda for is professional guidance, advice,tencouragement throughout the study

The technical and conceptual support of Dr Mare Souris, thesis committee memiber, helped meto conduct the research for which I express my thanks to hìm,

Valuable suggestions support of Dr Ulrich Glawe and thesis committee member help me towork enthusiastically so T am grateful to him,

1 would like to express my sincere thanks to DANIDA for the scholarship and Star programforthe research grand, thereby making this study possible.

Special thanks go to RSL staff, Mr Do Minh Phuong for providing all the necessary on time.Tam grateful lo the local in Xin Man province who provided and g

landslide points to measurement GPS.

ide me go to all theMy vote of thanks goes to all my fiends, Mr Tran Trung Kien, Miss Dao Thi Chau Ha, fortheir helps, supports, sharing the difficulties to my life in AIT.

Most ofall, [want to express my deep appreciation to my family: Parents, my sister for theirendless love, constant support and encouragement for the graduate study.

Xin Man district in the South west Ha Giang has high landslide hazard However, theavailable information on landslide in Xin Man district is still limited We constructed theessential spatial database of landslides using GIS techniques The quantitative relationships

between landslides and factors affecting landslides are established by the Certainty Factor

(CF) The affecting factors such as slope, elevation, landcover, geology, road distance,lineament distance, drainage density are recognized By applying CF value integration andlandslide zonation, the most significant affecting Factors are selected.

By using RS&GIS technology landslide occurrences on all these factors have been analyzed,

The vector based GIS has been used for diitizing to produce thematic maps, as analysis for

study was based on the pixel based information therefore Raster based GIS has been used forthe analysis,

Pixel based calculation was made by using the CF value Model By using the CF model we‘obtain the CF value forall clases al all factor maps On the basis of these CF value all factormaps are recoded and matrix analysis was perform to produce a Landslide Hazard Zonation

The Landslide Hazard Zonation map has been applied to develop a methodology to producehazard maps considering the behavior of landslide and to evaluate potential damage toinfrastructure specific road system, Different factors have been cons dered for this study,

‘TABLE OF CONTENTS

CHAPTER TITLE PAGETitle page i

ACKNOWLEDGEMENT a‘ABSTRACT it

‘TABLE OF CONTENTS wLIST OF TABLES vi

LIST OF FIGURES vị

LIST OF MAPS ixLIST OF ABBREVIATIONS x

1 INTRODUCTION 1

1.1 Background 112 Statement of problem 113 Objectives 21.4 Scope and limitation 22 LITERARUTE REVIEW 32.1 Haơand, risk & vulnerability 32.2 Landslide Hazard mapping 42.3 Fundamental of Remote sensing 5

24 GIS overview 2

2.5 Global Positioning System (GPS) 22.6 Web Map Server 227 Landslide Studies 153 DESCRIPTION OF THE STUDY AREA 1831 Area and situation Is32 Chmate 18

33 Rainfall Is3.4 Population 18

35 Geology 203.6 Elevation 247 Slope ”38 Lineament 2539 Road system 23.10 Drainage density 293.11 Landcover 30

TABLE OF CONTENTS (CONT.)

CHAPTER TITLE PAGE45 Methodol xy and Analysis Data 346 Data Entry 4047 Analysis data dị4.8 Landslide Hazard Zonation Map 85 RESULTS AND DISCUSSIONS “45.1 Characterize several types of landslide in Xin Man district “45.2 Landslide Hazard Zonation map 46533 Landslide Hazard Zonation map 545.4 Accuracy Check for Landslide Hazard Zonation Map 605.5 Develop a methodology to produce hazard maps considering the behavior oflandslides ø536 Publish Landslide Hazard Zonation to Inernet using Web Map Server 69)6 CONCLUSIONS AND RECOMMENDATIONS 166.1 Conclusions 166.2 Recommendation n

APPENDICES 80

Landsat 7 ETM image characteristic

Geology, major litho-stratigraphic units with their corresponding classes‘Area under Geology

‘Area under Elevation‘Area under slope

‘Area under distance to lineamentArea under distance to road‘Area under drainage density“Area under Landeover

‘Analysis data from different sourcesHazard zones

“The hazard value ranges used for road bufferThe hazard value ranges used for whole area

area for landslide hazard zone for butfer area` area fo landslide hazard zone for whole area

Defining the risk(Classification risk level

Result of the risk class based onbuffer analysis

PAGE

WES Processing Reque:

Map of large Landslide areas of Vietnam“The yearly rainfall from 1961 to 2003

Location of Study area Xin Man district, Viet NamGeology chart

Elevation chart

Slope chart in Xin Man distsictDistance to the lineament chartRoad area under the bufferiinage density chan:Landeover chart

ow Diagram For Landeover MapFlow Diagram for Digitized Map

Flow Diagram for Landslide Map using GPS

Flow Diagram For TIN and maps extraction from TIN

Flow Diagram For Landcover Map extracted From Satellite DataFlow Diagram for Buffered Road and lincament Maps

‘Methodology of thematic data layer preparationShow the landslide attacked road.

‘Wedge slip oceur along the road.CF value of geological

CCF value of distance to lineament

Statistical map of slope angle distribution in Xin Man District

CCF value of slope angle layer 50CF value of elevation layer 5CF value of drainage density layer 32CE value of lindeover layer 33CCF value fr distance road, 54[Bar chart showing the distribution of various hazatd zones 37or buffer area in Xin Man district 37‘Bar chart showing the distribution of various hazard zones for whole area in XinMan district, 37Relative distributions of various hazard zones and landslide probability within

‘each zone in road buffer in Xin Man district, 60

FIGUREsass5163.17S18519sau

Tin in Xin Man district

Geological in Xin Man districtElevation in Xin Man district,

Slope in Xin Man district

Distance to Hineament ia Xin Man,Buffer road system in Xin ManDrainage density in Xin ManLandeover in Xin ManRoad system in Xin Man

Landslide distribution sin Xin Man district

[Landslide hazard zonation for bufer area in Xin ManLandslide hazard zonation for buffer area in Xin ManRisk of slope in Xin Man's Road Network

Risk of distance to Xin Man's oad networkRisk in Xin Man

Risk assessment for road network in Xin Man

‘Geographical Information SystemGlobal Position System

‘Geographical Mark Up Language

Join Photographic Experts Group

Uniform Resource Locator‘World Wide Web

Web Map Service

Feature Service

CHAPTER 1

INTRODUCTION1.1 Background

Landslide has become one of the world’s major natural disasters forthe few years in manyeounies Landslides are the most common natural hazard in mountainous terrain Landslidecan be a major threat to population in the mountainous area Even when they occur away fromthe inhabited areas, landslide can be a significant hazard and have a serious economic impactby blocking roads and river (Aniya, 1985, J Achache, B Fruneau and C Delacotrt 1995).

Landslides are widespread in many countries and cause great economic loses, especiallywhen engineering constructions are designed and erected without heeding the stabilityconditions ofthe slopes (Q Zaruba, V.Menel 1967),

Landslides become a problem when they interfere with human activity The frequency andthe magnitude of the slope failures can be increased due to human activities such asdeforestation or urban expansion,

Landslide hazard analysis is a difficult task, Tt requires large number of parameters andtechniques for analysis, Remote sensing and GIS are the powerful analysis tools to handle ths

type of problems A in the analysis of landslide spatial information e.8 topography, geology,

landeover, ete are involved, therefore application of Remote sensing and GIS will beeffective

1.2 Statement of problem

- Although landslide usually occur in Xin Man district, but people who live near or inthelandslide’s local do not ilusrate the different between them, But actualy, there are many’types of landslide which ean occur and each of them have separate characterize, We need to

ive some information to describe characterize some types of landslide in Xin Man district,

~ Landslide is a serious disaster in Viet Nam In recent 10 years, there are more than 10areas occurred violent landslide, causing above 300 human deaths and thousands of hectaresof solids was buried by stone, sand, pebble and hundreds of inhabitant settlements having toe ther living places and locations These are responsible for considerably greater socioeeonommie loss than is generally recognized There are some projects and research applying forlandslide but only for mid_center of Viet nam Up to now, there is not hazard map, risk mapabout landslide in Xin Man district, the leader of province only have measure to preventlandslide every year and they have not had any project to study about landslide in the Xin Man.district Hence, there is an urgent need to prepate landslide hazard zonation maps in the highlylandslide susceptible mountainous terrain special is Xin Man district.

= No other landslide investigation or risk assessment has been performed in Xin Man

district to date

= Understanding and prevent landslide hazard is very important for every people What can

people do when lack of information about natural hazard? Nowadays, internet is popular and

useful for every people People can update, download all information and all thing which theyneed to Know, In this regard, we need to publish and share information about landslide onIntemet by using Web Map Sever.

1.3 Objectives

The general objective of this study is using Remote sensing and GIS technique to makinglandslide hazard zonation mapping in Xin Man district.

The specific objectives of the study are

1 Characterize several types of landslide in Xin Man district.

2 Create zoning maps for landslide hazard that usually occurs in Xin Man district,3 Develop a methodology to produce hazard maps considering the behavior oflandslides

4 Publish and share landslide hazard zonation map's information on internet using‘Web Map Server.

14 Scope and li

Landslide hazard map zonation will be focuses on critical physical factors by ust

overlaying thematic maps.itation

cishave high risk of landslide ba

topology, geology, hydrology, and geomorphology’

The limitation is associated with the availability of reliable and adequate data sets fromsecondary soutees to support making landslide hazard zonation map,

c on investigation, study

- Risk assessment only for road networks, not consider about the others as population,L

= Data collection is not enough to be analysis,

~ Landsat TM images will be used for analysis of landcover of the study area.

= Apply existing program to publish landslide hazard zonation map on internet using WebMap Server

(CHAPTER 2

LITERARUTE REVIEW

Natural Hazard is extreme events in the earth's ecosystem The concepts of hazard, risk, andvullcrability are often confused with one another and with the extreme event ise Althoughthe extreme event is inherent in hazard, risk and vulnerability terminology, it is notsynonymous with the terminology Therefore itis necessary to distinguish between the termshazard, risk and vulnerability.

Hazard assessment determines the type of hazardous phenomenon, frequency, magnitude andthe extent of he area that may be affected Vulnerability indicates the degree of loss caused topeople, infrastucture, buildings, economics cte distinguishing physical (buildings,

nfrastructure), functional difelines, communication) and social aspects (health, populationdensity) Risk combines the knowledge about hazard and vulnerability to make a quantitative

prediction of the elements at risk, like numbers of lives to be possibly lost, people to be

Injured, cost of property being damaged and destroyed or economic activities a affected.3,1 Hazard, risk & vulnerability

In order to provide a systematic approach to study the landsvarious types of hazard, risk & vulnerability

ide, Varnes (1984) definedNatural hacard the probability of occurrence of a potentially damaging phenomenonwithin a speeifie period of time and within a given area,

Vulnerability the degree of loss to a given element or set of elements resulting from the‘occurrence of a natural phenomenon of a given magnitude,

Element at risk the population, propertis, economic activities ete at risk in a given area

Risk the expected degree of loss due to a particular natural phenomenon Hence it is a

product of hazard and vulnerability.

Criteria fo assessment is represented schematically as below (Figure2-1)

Although by the term landslide is used for mass movements occurring along z well defined

sliding surface, it has been used in literature as the most general term) forall kinds of massmovements, including some with litle oF no true sling, such as rock-falls, topples, and debrisflows (Varmes, 1984) In this context, mass movement is used subsequently as a synonymousterm for landslide, similar to slope movement

Zonation refers to the division of the land surface into areas and the ranking of these areasaccording to degrees of actual or potential hazard, Hence landslide hazard zonation showspotential hazard of landslides or other mass movements on a map, displaying the spatial

and present failures Thus, we have the possibility to estimate the features of potentialfuture failure The absence of past and present failures does not mean that Edlures will‘not oecur in the future

> The main conditions that cause landslides can be identified: The basic cause of slopefailures can be recognized, are fairly well known from several case studies and theeffects of them can e rated or weighed It is possible to map correlate the contributingfactors to each other.

> Degree of hazard can be estimated: If the condition and processes that promote

instability can be identified, itis often possible to estimate thei relative contributionand give them some qualitative or semi-quantitative measurement, Thus, a summery of

se of potential hazard in an arca ean be built up, whieh depends on the numberre including factors present, their severity, and their interaction,

2.2.2 Seale of mapping for landslide hazard zon:

There are several technique for landslide hazard zonation can be applied, making use of GIS.Therefore the appropriate scale on which the data is collected and the result presented variesconsiderably More detailed hazard maps require mote detailed input data Thus the objectiveof the analysis and the requires accuracy of the input data determine the scale.

The following scales of analysis have been differentiated for landslide hazard zonationaccording to the definition by the Intemational Association of Engineering geologists (1976)

"+ National seale(<1:1,000,000),

‘+ Regional seale(1:100,000 ~ 1: 1,000,000)+ Medium seale(1:25,000 ~ 1:100,000)

Electromagnetic radiation, whichis reflected or emitted from an object, is the usual source ofremote sensing data A device, to detect the electro-magnetic radiation, reflected or emitted,fom an object is called a “remote sensor” or “sensor” A vehicle to carry the sensor is called a“platform

Remote sensing is classified into three types with respect tothe wavelength regions.> Visible and reflective Infrared remote sensing,

> Thermal infrared remote sensing.> Microwave remote sensing

2.3.2 Spectral Reflectance of Landeovers

Spectral reflectance is assumed to be different with respect to the type of landcover.This isthe principle that in many cases it allows the identification of landeovers with remote

sensing by observing the spectral radiance from s distance far removed from the surface.

Fig.2.2 shows three curves of spectral reflectance for typical land covers; vegetation, soil and

water As seen inthe figure, vegetation has a very high reflectance in the near infrared region,

though there ate three low minima due to absorption,

Soil as rather higher values for almost all spectral regions Water has almost no reflectance inthe infrared region.

Figure 2.2:Spectral reflectance of vegetation, soil and water

Figure 22 shows two detailed curves of leaf reflectance and water absorption Chlorophyll,contained in Tea, has strong absorption at 0.45 vm and 0,67 Hm, and high reflectance at nearinfrared (0-09 0m) This results in @ small peak at 015-046 (green color band), which makesvegetation green tothe human observer.

Near infrared is very useful for vegetation surveys and mapping because such a steep gradientat 0.7-0.9 0m is produced only by vegetation,

Because of the water content in a Tea, thee ate two absorption bands at about 15 em and 1.9im This i also used for surveying vegetation vigor.

Percent Reflectance Wavelength (zm)

Figure 2.3: Spectral reflectance of a green left

Description of the data setLandslide image

Database of remote sensing is used inthis thesis is Landsat 7 ETM The application of satelitedata has increased enormously in the past decade After the initial low-spatial resolutionimages of the LANDSAT MSS ( which were about 6D by 80 m), LANDSAT now hác aSignificant improve in its characteristics with thematic mapper (TM) images Tt has a spatial

resolution image of the 30 m and excellent spectal resolution Landsat TM provides sevens

bands to cover the entire visible, near infrared and middle infrared portions of the spectrum,

with one additional band providing a lower resolution of the thermal infrared (table 21)

Landsat satellite orbits are arranged to provide good coverage of a large portion of the eath’s

surface, The satellite passed over each location every 18 days, offering a theoretical temporal

resolution of 18 days,

Table 2.1: Landsat 7 ETM image characteristic

Band | Spectral range(um) | Spatial resolution(am)1 045.052 30

2 052-060 303 068.069 30a 026-090 3o

5 155116 30

6 104125 607 208235 30

Pan 050-090 5

2.3.4 Image interpretation

Image interpretation is defined as the extraction of qualitative and quantitative information inthe form of a map, about the shape, location, structure, function, quality, condition,

relationship of and between objects, ct by using human knowledge or experience

Information extraction in remote sensing can be categorized into four types which are asfollows:

Classification is a type of categorization of image data using spectral, spatial and temporalinformation

Change detection i the extraction of chang

Extraction of physical quantities comresponds to the measurement of temperature, atmosphericconstituents, and elevation and so on from spectral

enfieation of specifi features is the identiother fearure ete

between multi-date images

ation, for example, of disaster, lineament andFigure 24 show a typical flow of the image interpretation process:

> Reconstruction/ correction> Transformation,

> Classification,> Output

Figure 25 shows the dataflow in remote sensing

‘AD conversion using a | | Primary Processing D/Dfilm scanner ete ‘conversion.

2.3.6 Geometric Correction

Geometric correction is undertaken to avoid geometric distortions from a distorted image, andis achieved by establishing the relationship between the image coordinate system and the

geographic coordinate system using calibration data of the sensor, measured data of position

and attitude, ground conol points, atmospheric condition eteThe steps to follow for geometric correction areas follows:

2.3.7 Registration and Rectification

Refael C Gonzalez Rechard E Woods (1993) explained that the another important applicationis the image registration or finding correspondence between two images The procedure forimage registration is the same as the method just illustrated for geometnc correc ton,However, the emphasis is on transforming an image so that it will correspond with anotherimage of te same seience but viewed pethaps from other prospective,

2.3.8 Classification

+ ope con nle

4, Multi-level die cleierDecision wee essiierli Minimum dsc classifier[castewion [=] Naku kta asi

ier clas such Sẽ uøn theory and expert

vericaon of rma ‘Vesfyng of he acury or rib ofthe rsa

Figure 2.7-Procedures of Classification> Supervised Classification,

It is done in order to determine rule of classification, It is necessary to know the spectralcharacteristics or features with respect to the population of each class The spectral featurescan be measured using ground based spectrometers However due to atmospheric effects,direct use of spectral features measured on the ground are not always available, for tis reason,sampling of taining data from clearly identified training areas, corresponding to defined

classes is usually made for estimating the population statistics Statistically unbiased sampling

of training data should be made in order to represent the population correctly.

> The minimum distance classifier is used to classify unknown image data to classes whieh‘minimize the distance between the image data and the class in multi-feature space Thedistance is defined as an index of similarity so thatthe minimum distance is identical to thế‘maximum similarity.

> The maximum likelihood classifier is one of the most popular methods of classification inremote sensing, in which a pixel with the maximum likelihood is classified into thecorresponding class The likelihood is defined as the posterior probability of a pixel

belonging to class k

2.3.9 Spatial Filtering

Spatial filtering is used to obtain enhanced images or improved images by applying, filter

function or filter operators in the domain of the image space (x.y) or spatial frequency (x).Spatial filtering in the domain of image space aims at image enhancement with so-calledenhancement filters, while in the domain of spatial frequency it sims at reconstruction With socalled reconstuction filters, which is in the domain of spatial frequency it aims at

reconstruction with sơ called reconstruction filters An output image from filtering of spatial

pass filters, band pass filter etc are typical filters with frequency control Low pass filters

‘which out puts only lower frequency, noise, while igh pass are used for example stripe noiseof low frequency.

2.4 GIS overview

2.4.1 Geographic Information System

According to Burrough (1986), Geographic Information system (GIS) is a powerful setof tools for collecting retrieving, transforming, and displaying spatial data from the real worldfora particular set of perposes

Aronoff (1998) states that GIS is designed for the collection, storage and analysis offbjects and phenomena where geographic loction is eriical to analysis For example, thelocation of fire station or the locations where soil erosion is most severe are keyconsiderations in using information In each case, what itis and where i is must be taken into

24.2 Terrain Modeling for Mountain

Mountain may be defined as dynamic system in which both the extent and variabilityin relief are key controlling elements Altitude, aspect and slope strongly both the human andthe physical characteristics of mountain ecosystems, such as the distribution of agriculture, thetype of forestry, micro and local climates and the extent of the mass movement, A model ofrelief is therefore an essential component of a mountain GIS At present the most powerfulmethod of representing relief is to construct a mathematical model of the earth's surface: 8digital terrain model( DTM) or digital elevation model(DEM) This mathematical model canbe used to drive information on height, aspect, slope, angle, watersheds, bill shadows and eutand fill estimates which may be essential components of management plan or inputs to a

process model

‘Any digital representation of continuous variation of relief over space is known as aDEM, DEMs were originally developed for modeling relief; they can of course be used tomodel the continuous variation of any other attribute Z over two dimensional surfaces(Burrough, 1986)

2.5 Global Positioning System (GPS)

The US Deparment of defense developed a navigation system called GlobalPositioning System “GPS” Ics based on the 24 satellites which orbit around the earth at analtitude of 20,200 km the satellites are high enough to avoid land based system problemsWith this technology one can find the location of an object any where in the World 24hrs day.The accuracy for measurement with GPS is from 5 to 10 meters range With differential post

processing the accuracy can be few milimeter.

GPS is digitalelectronie equipment based on satelite ranging; it means we can figure out ourposition on earth by measuring our distance from a group of satellites in space The satellitesact as a reference point

There are atom clocks on the satellites which are show accurate time, due to which we caneliminate any error caused by the watch of the GPS receiver

2.6 Web Map Server

extension which makes the PostereSQL connected to the Minnesota MapServer in order todisplay the database in various formats Minnesota MapServer is used to explore GIS datacover World Wide Web (WWW),

(Open Source technology is base on the premise that the programming source code is freelyavailable to anyone who wishes to read, add to, or even modify and redistribute the computersoftware code Open source technology offers a level of stability and flexibility that is not

typically available with “out-of-the-box” software.

Benefit of Open Source Technologies

> Five to use- there are no licensing fees,

> The software can be duplicated and installed on as many machines in as manyenvironments, with no restrictions.

> Full access to source code

> Highly responsive to end user requirements

(Open source does not just mean aceess to the source code, The distribution terms of source software must comply with the following criteria:

open-a Free redistributionb Source Code.«, Derived Works,

dd Integrity of the Author's Source Code

«, No Discrimination against Petsons or Groups.

£ License Must Not Be Specific to a Product (Open GIS).

Minnesota Map Server provides Open GIS Consortium’ (OGC) Web map ServiceCWMS) andWeb Feature Service These two specifications will use operation on the client's request to

produce maps of georeferenced data in various formats such as JPEG, PNG, GIF and GML.

2.6.2 Open GIS Standard

‘The OpenGIS Standard specifies the behavior ofa service that produces georeferencedmaps, This standard specifies operations to retrieve a description of the map offered by aservice instance, to rerieve a map, and to query a server about features displayed on a map.OpenGIS Standard is applicable to pictorial renderings of maps i a graphical format Thisstandard isnot applicable to retrieval of actual feature data or coverage data values

1 Web Map Service (WMS) Implementation Specification

A Web Map Service produces maps of georeferenced data, We defined a 'Map” as avisual representation of geodata; a map is not the data itself, These maps are generallyrendered in a pictorial format such as PNG, GIF or JPEG, or occasionally as vector-basedgraphical elements in Scalable Vector Graphics (SVG) or Web Computer Graphics Metafile(WebCGM) formats This specification standardizes the way in which maps are requested byclient and the way that servers deseribe their data holdings The three operations are asfollows:

> GeiCapadiltes (Roguired): Obtain service-level metadata, which is a machine readable

(and human-readable) description of the WMS’s information content and acceptable

queried and which location on the map is of interest.

2, Web Feature Service (WES) Implementation Specification

‘The OGC Web Map Service allows a client to overlay map images display served from‘multiple Web Map Services on the Intemet In a similar fashion, the OGC Web FeatureService allows a client to retrieve geospatial data encoded in Geography Markup Language(GML) from multiple Web Feature Services.

The WES operations support INSERT, UPDATE, DELETE, QUERY and DISCOVERY‘operations on geographic features using TTP as the distributed computing platform.

WPS Request WFS ResponseWeb Feature Servies

Figure 2.8: WFS Processing Request

2.7 Landslide Studies

-Kwang-Hoon Chi*, Kiwon Lee**, and No-Wook Park*(2001), studied LandslideStability Analysis and Prediction Modeling with Landslide Occurrences on KOMPSAT EOCImagery In this study, Slope-Area plot methodology followed by stability index mapping withthese hydrological variables is firstly performed for stability analysis with actual landdide‘ecurrences at Bocun area, Korea, and then landslide prediction modeling based on ikelihoodratio, model for landslide potential mapping is cartied out; in addition, KOMPSAT EOCimagery is used to detect the locations and searped scale of landslide occurrences These twotasks are independently processed for preparation of unbiased criteria, and then results of thoseare qualitatively compared, As results of this case study, land stability analysis based onDEM-based hydrological variables directly reflects terrain characteristics; however, the results,in the form of land stability map by landslide prediction model are not fully matehed withthose of hydrologic landslide analysis due to the heurkti scheme based on location of existedlandslide occurrences within prediction approach, especially zones of notinvestigated‘occurrences, Therefore, it is expected that the results on the space robustness of landslide

prediction models in conjunction with DEM-based landslide stability analysis cạn be

effectively utilized to search out unrevealed or hidden landslide occurrences.

-D-Z.Seker, M.O.Altan(2001) used various types of data to extract relevant information

This study is to determine a suitable methodology for predicting possible landslide areas andproducing landslide risk map inthe study area of Sebinkkarabisar Township, which is locatedatthe northeastern part of Turkey These include the satellite sensor data taken in the year of1987 and 2000, which are use for the extraction of land surface temperature and landusenformation 1:25000 scale standard topographic map has been digitized and the obtainedcontours were used for the derivation of DEM and slope map of the study site Sateliteimages, DEM and slope map of the study atea were used to investigate the possible landsliderisk areas and reason ofthis natural hazard which treat the study area frequently

Nguyen Quoe Phi, Bui Hoang Bac (2000) gives a view of landslide characteristics on

natural terrain of YangSan area, Korea and developing a GIS approach to modeling slopenstability, The relations between landslide distributions withthe physical parameters such aslithology, elevation, slope gradient, slope aspect, lineament, drainage, vegetation, and land suewere analyzed hy Bayesian statistical model A susceptibility map 1s modeled by incorporatingthese in weight of evidence model using Bayesian approach,

Atsushi Kajiyama, Takamoto, Truong Xuan Luan and Nhu Viet Ha (2002) used

stereo-photogrammetry fechnique (using Photomodeler software) application for monitering alandslide twice in Monsetatea, northwestern part of Viet Nam in 2002 to 2003 this techniquehas allowed us to deriver surface deformation maps of landslide with a high spatial resolutionand accuracy Photomodeler software can treat it easily using reference point and thephotograph, Which they have been taken photograph with common digital camera Using this,Software, we estimated the amount of movements of whole landslide for one year The resultBạc been validated by comparing independent measurements carried out by laser telemter

“Mike Doratti, Chris MeColl and Claire Tweeddale (2002) applied GIS to predictlandslide hazard areas following clearcut logging events The landslide prediction study

Pr6Eet sponsor is Tom Millard of BC Ministry of forest The objective of this project is to

produce a report and maps identifying potential landslide hazard areas within the Cascade

1s

Mountain rogion, Brith Columbia, Historically, use of GIS technologies in this area of ForestResource Management have been limited, therefore application of this software could

potentially improve current practices Slope stability factors were derived from TRIM digital

clevation models Algorithms were developed to create a bndslide hazard model This model‘was compared to existing landslide data, and to the current terain stability mapping standardto assess model accuracy.

-Amod Sagar, Takaaki Amoda and Masamu Aniya (1997) This paper presents the

evelopment of landslide susceptibility mapping using GIS The study area is KulekhanWatershed lying in the lesser Himalayan region of Nepal A landslide distibution map

produced from interpretation and field work was used to analyzed the important factors tô

landslides, empbying Quantification scaling type 2(Q-S2) method The six factor used wereslope gradient, slope aspect, elevation, geology, landuse/cover, and drainage basin ordOverlaying the factors with scores for their classes computed by the analysis, landslidesusceptibility maps were produced with classes of high, moderate, less and least susceptible,

‘Richard Kho Shu Yuan and Mohd Ibrahim (1997) In this study, land surface

temperate and landuse information have been derived from Landsat thematic’ Mapper dataThe elevation and slope inclination have been determined from DEMs generated from aerialphotographs Underground water level information has been obtain from the combination ofabove data, From these data, simple algorithms were used to classify the area nto differentFisk zones By combining all the risk maps using GIS techniques, final nsk maps were

produced which take into account all above factors,

-Purna, Dr kaew, Dr Jean Delsol, P Gupta, Prinya (1998)-The methodology is used

for landslide haZarH zonation, Landslide distbution is overlaid with other landslideinfluencing spatial parameters slope, aspect, and land system, landuse, bedrock, isohyetal andseismic zonation, amd first order buffered stream map The slope and aspect maps were

generated from the DTM The land use map and landslide map were produced from aerialphoto interpretation

Figure 29: Map of large Landslide areas of Vietnam

”

CHAPTER 3

DESCRIPTION OF THE STUDY AREA3.1 Areaand situation

Xin Man district is situated in the South West of Ha Giang province, The study area

lies geographically between 22°10" N to 23°30'N and 104020) to 108°34'E.

Xin Man and Hoang Su Phi districts lic in the high land and accounting for 18.3% of thetotal province's area and 17.2% of the population The potential for the area is to develop

plants for derived resin,

The mean annual rainfall, according to the records of Metrological station, for the 20 years

(1985-2005) is about 1,695 mm Most of the rain comes in the months of August and

September In that time, the intensity rain fall is 200012500mm in some high mountains

(>1500m) and causing fash flood and landslide

Figure 3.2: Location of Study area Xin Man district, Viet Nam.‘Map 3.1: Tin in Xin Man distriet

Elevation and slope maps are extracted from TIN which has been obtained from the Contourmap in Xin Man district The TIN is show in the map 3.1

= The gravely sol is not combining together and having different component, size live

in valu of river and sream, and the slope surface This gravelly soil group is verydevelopment in the watershed, including is sediment slope, sediment flood and smallerthan rising than level of sediment

© Rock weathes

very strong including metamorphism two mica schist, granite is

= All granite isnot weathering

Al of group gravely are also moved by the earth's crust making many local are compcatalectic, breaking, and ereate advantage condition for landslide and debris flow.

Inthe study area, the various litho-stratigraphic units were prouped into four categories tisshown in table 3.12 & figure 3.12

‘Table 3.1: Geology, major litho-stratigraphic units with their corresponding classes

Geology ‘Major litho-stratigraphic unit

Canh Quartabiotite schist, serieie-cMlodte schist, shunggite,green schist, marbleized limestone bearing oncolite,phyllite

Marble, motley limestone, clayish limestone, claysericiteChg? schist

Fine-to medium-grained, porphyritc biotite granite

Dies! ‘Coarse grained gneissoid biotite granite

Dise2 Graphite-bearing marble, wo-mica schist, quarta-mica

PR3 schist, quartzite, epidote-biotte schist, epidote-botte

hornblende schist, thin beds of marble.

‘Table 3.2: Area under Geology

Chese| ‘Total aretkm2) | Areain%

PRS 3539 539

Dise2 94237 tánCah! 18973 2099

Disel 359352 59.46

Among these categories, $9.46% of total area is presented by group of D1scl and 21% area is

ecupied by C2hgl and C2gh2 Disc? is oceupied 14.16% in total area Only small (5.39%)

of area belong to PR3.

Figure 3.3: Geology chart

2

‘Map 3.2: Geological map in Xin Man district

GEOLOGICAL MAP IN XIN MAN

| y2ng!

(DEN) And its shown in the Map 3.1

Table 33: Area under Elevation

Eletatonelasem) oval area (ưn3) ‘Aecain0-300 33.082 798

300-500 947 1428

500-1000 246.366 Em

1000.1500 202.98 3050

1500.2000 65.127 009

% of Areaeos SBR ERSElevation class

Figure 3.4: Elevation chart

‘Map 3.3: Elevation map in Xin Man districtELEVATION MAP IN XIN MAN DISTRICT

3.7 Slope

Due to presence of rugged terrain, occurrence of landslide and field survey in the Xin Man theslope is divided into following six classes,

‘Table 34: Area under slope

Slope clases(degree) | Total area(kn) Area in %

Source: Contour Map of scale 1:50 000 and Digital terrain model

From the digital information of slope it has been obverted that 76.3% of area has slope rangeof 15-45 degree and only 5.176 has slope range more than 45 degree The percent of area

under different slope range are given in table 3.6 and slope variations are shown graphically in

figure 3.5 The slope map hat been extracted from the TIN and slope classes are shown in themap 3.3

245 848 0:5 3846 | 25-95 1526

Map 3.4: Slope map in Xin Man district

h ‘SLOPE MAP IN XIN MAN DISTRICT

3⁄8 Lineament

“The lineaments are the linear morpho-techtonic feature of the femsin which include faults,fractures, ridges, major discontinuities ete (S.Sarkar, 2003) In this work, lineament were

found depending on geology map This area are developed fault system of NW-SE, ME-SW

and sublaitudinal faults,

-NW-SE trending faults system is developed contrentratesly in the northeast, Together withNW-SE ending fault system they form faults of feather form in the southeast of the Chay

river Granite Massif

- Sublaitudimal faults from the boundary between the Chay river Granite massif and‘Cambrian Devonian sediments,

‘The lineament are buffered in different distance The distances to lineament factor was

‘computed using analyst extension of areview and were classified as tale 3.5 below.

25

‘Table 3.5: Area under distance to lineament

Distance classesim) | ‘Totalarea | Areain %

0-500 mm" mm

500-1000 90.88 1365

1000-2000 mm In2000-3000 79.48 11983000-4000 sua 1046

4000 204.49 3072

Total 665.61 100

Tt has been observed thatthe $8.82% of the area lie under the distance range 0-3000m, One

third ofthe total area belongs to class more than 4000m Only 10.46% of total area ie in range3000-40000.

Map 3.5 shows the distance to lineament classes in different color.

# 10

Distance classes

Figure 3.6:Distance to the lineament chart

ince to lineament map in Xin Man

(AP OF DISTANCE TO LINEAMENT

3.9 Road system

“The study area has one major road having width of 8 m A large number of houses are

‘connected to the road So we must to create bufler to the road with different distance, Theresults ate given in the table 36,

‘Table 3.6: Area under distance to road

Distance classesin) | Total areatkm2) | ‘Area

0-10 16698 251

10-35 33166 348

25-50 31213 +6)30 sim 43815-150 T196 tới>150 4040 1323Total 665.61 100

mm

thas been observed thatthe 73.23% of road lies under distance range more than 150 m, only2.51% of area belong to range 0- 10m, Total range from 10-150 m occupied nearly 25% of thearea Road network in Xin Man district i shown in the Map 3.6 The distance range variationfor the buffer road is shown in the figure 37,

3,10 Drainage density

Drainage density is defined as the ratio of sum the drainage length in the cell and the area ofthe corresponding cell (S Sarkar, 2003) The under cutting action of the river may include theeffect of this causative factor and converted into raster format The drainage density (figure3.10) was computed considering a 20 by 20 m cell and classified with intervals as show in

table 3.6

Table 3.7: Area under drainage density

denna Total area(km2) | % of Area

260.115 390

500-1500 302.594 4546

1500-2000 59.352 89222000-2500 25.856 88

2500:3500 l6 247

3500-4500 1245 019)

Tt has been observed that 84.545: of area has a density range from 0-1500 and only 0.195: ofarea has the drainage density range 3500-4500, Drainage density led range 1500-3500‘occupied approximately 15% of atea The drainage density variations are shown graphically in

figure 3.10 and the map of drainage density is shown inthe map 3.8

Figure 3.8:Drainage density chart

”

Map Drainage density map in Xin ManDRAINAGE DENSITY AP IN XIN MAN

TM 25003600 sp

3.11 Landeover

Landcover in Xin Man is clasifed into five categories, out of which dense forest bears thehighest percentage of the area (36.57%) Therefore, hareland occupied one of third ofthe area302% Landcover classes are given in the table 3.10 & figure 3.10 The landeoverinformation, extracted from the satellite data, is shown in the Map 3.9.

Taible 3.8: Area under Landcover

Landeover classes | Total areatkm2) | Scof AreaDense forest 23438 3657

Bush and thin forest 13.348 1703

Bareland 204353 307‘Cativated and 59278 891

Wetland 68