4th International Conference on Integrating GIS and Environmental Modeling (GIS/EM4):
Problems, Prospects and Research Needs. Banff, Alberta, Canada, September 2 - 8, 2000.


Geotechnical Hazard Zonation of Lorestan Province by GIS

GIS/EM4 No. 239

M. R. Mahdavifar
Masoomeh Rakhshandeh
Piran Veyseh

Abstract
Lorestan is one of the most geotechnically problematic provinces in Iran. Landslide is the most dangerous geotechnic hazard that affects the area .In the study area, 250 landslides and landslide zones were recognized which have caused many damages including destruction of villages, farmlands, roads and erosion of land surface. In this study, for assessment of landslides and other geotechnical hazards such as liquefaction, subsidence and problematic soils, GIS softwares including Arc/Info, Ilwis and Arcvieware used to simplify some stages of the work like digitization, AutoCAD was also used. The site characteristics as related to geotechnical hazards were evaluated using univariable statistical method. Six hazard maps (including 3 hazard maps for problematic soils) were produced from the results of the analyses. This paper describes one of the geotechnical hazard assessments, landslide, in the area.

Keywords
Landslide, hazard zonation , GIS, Lorestan province , Iran.


Introduction

Iran is a developing country with a complex geology, seismicity and seasonal rainfall. These factors have influence on the development of landslides and other types of geotechnical hazards, which may result in loss of life and damage to economy. The preparation of geotechnical hazard zonation maps is the first step in assessing the degree of hazard evaluating its potential. Lorestan province, an area of some 2806 km2, is situated in the southwest Iran (fig.1). The area is one of the most sensitive parts of Iran to geotechnical hazards.
 



Figure 1. The location of the study area in Iran
 

  Using GIS in geotechnical hazard zonation

Data handling for assessment of geotechnical hazards is often difficult, time consuming and costly (Dhakal, et al, 2000). Geographic information system has overcome the many working difficulties associated with data handling (Carrara; 1983). The relative contribution of the landslide causing factors are assessed on the assumption that the landslide hazards will be more likely to occur under conditions similar to those of previous landslides hazards. Hazard maps can be produced from the results of the analyses. One of the most important features of this study is making use of GIS in most stages of the work, including data analyses.
 


Acquisition of databases and GIS data layers

The stereo pairs of black and white vertical aerial photographs (scale 1:50000) were interpreted using the stereo zoom - transferscope to plot landslides, subsidence and other basic information on a topographic map at a scale of 1:250000 which was digitized after field verifications.
From digitized contours (contour interval 100 m), a digital elevation model (DEM) was generated from a Triangulated Irregular Network (TIN) model. Slope gradient and slope aspect layers were derived from the DEM.
 


Landslide hazard zonation

Landslide is defined as " the perceptible downward sliding or falling of a relatively dry mass of earth, rock or a mixture of two" (Sharp, 1938). In this study, using existing data, interpretation of aerial photos and field working, 274 landslides including the largest known subaerial landslide in the world, SEIDMAREH , were located on topography maps .For preparation of landslide hazard map (scale: 1/250,000) the following procedures was used:
  1. Factor maps like geology, topography, land use, landslide inventory, earthquake hazard, rainfall, roads and rivers maps were digitized;
  2. The intermediate maps like slope map, slope aspect map and engineering geology map were prepared from base maps;
  3. Every factor map is crossed by the landslide inventory map and the total area of polygons having landslide is distinguished;
  4. To evaluate the rate of landslide distribution with regard to each factor, Surface Percentage Index (SPI) was used. This can be defined as:
      5.  
      (Eq.1)           SPI = [ Surface area affected by landslide for a defined factor group / Total area of that factor
                                     group ] *100
Using this formula, the SPI can be investigated and calculated for each individual factor and the susceptibility of the factor can be defined .The main factors influencing instability and other SPIs are indicated in table 1.
  1. In order to evaluate the potential for slope failure, the weight of each contributory factor was considered with reference to its surface distribution (SPI).
  2. From dispersivity of standard deviation of SP numbers, factor maps were weighted; (Table.1)
  3. The factor maps overlaying on each other by GIS and 60,000 polygons are built;
  4. A value of hazard potential index (HPI) for each of the defined factors was calculated based on the equation (2) .
           (Eq.2)
Where: HPI = the index that represents landslide hazard potential, Ri = the number that represents rating of the alternative based on a particular factor, Wi = the number that represents the weight of the particular factor.
  1. Based on HPI range, three hazard groups are defined (low, medium, and high);
  2. Based on HPI and defined ranges, landslide hazard zonation map of the province was provided by GIS .the landslide hazard map is shown in fig.2.


    3. For other geotechnic problems such as  liquefaction, subsidence and problematic soils hazard, these 10 steps as used for landslides, were applied with minor differences and their hazard maps were produced.


 
Factor Wht
G
 Total area Affected  SPI Rate Factor wht G Total area Affected  SPI Rate
  (w)   (Km2) area(Km2)   (R)   (w)   (Km2) area(Km2)   (R)
Rainfall (mm) 2 <400 108 0 0 0 5 3 871 64 8 10
400-700 22424 851 4 3 4 248 24 10 10
>700 5726 517 9 10 5 232 3 1 10
Distance from fault (Km) 1 <2 5290 353 7 10 6 627 2 0 10
2-5 6315 267 4 3 7 1540 19 1 10
>5 16654 751 5 5 8 3140 171 6 6
Aspect  1 N 1943 137 7 10 9 1201 183 16 10
NE 5766 329 6 7 10 338 20 6 7
E 1936 102 5 6 11 1106 88 8 8
SE 1735 76 4 5 12 1518 47 3 4
S 4571 239 5 4 13 1876 32 2 2
SW 6948 257 4 5 14 1078 49 5 5
W 2841 102 4 4 15 984 27 3 3
NW 2130 117 5 7 16 112 6 6 8
Flat 10 0 0 0 17 660 43 7 10
Slop  4 0-5 5802 83 1 2 18 111 4 4 4
5-25 13345 713 5 7 19 210 11 5 6
25-60 7930 557 7 10 20 459 52 12 10
60-100 1043 13 1 2 21 648 20 3 4
>100 21 0 0 0 22 284 5 2 2
Road 2 Main road 2000 132 7 9 23 290 24 9 9
Minor road 5143 256 5 7 24 235 58 26 10
4wl drive 4893 311 6 9 25 914 97 11 10
Railway 359 16 4 5 26 184 2 1 1
Ect 15890 653 4 0 27 142 16 12 10
Earthquake 3 Very low 1775 118 7 2 28 336 10 3 3
Low 6126 292 5 5 29 114 2 2 2
Medium 10034 419 4 5 30 783 0 0 2
High 7996 254 3 8 31 264 1 0 1
Very high 2205 285 13 10 32 112 0 0 0
Extra high 123 0 0 10 33 270 0 0 0
Lithology 5 1 396 38 10 10 34 358 1 0 1
2 6462 180 3 10 35 133 0 0 0

Table1. Distribution of landslide in relation to its factor  [SPI : Surface Percentage Index:

                                         W = Weight , R = Rate , G: Group]
 
 
Rate of Weighting
Hazard Potential
HPI value
No. of Unit
No. of Affected unit
SPI
A
High
6 - 10
14106
3615
25.63
B
Medium
4 - 6
7642
166
2.17
C
Low
0 - 4
6468
68
1.05

Table 2. Hazard Potential Index (HPI) values and their surface

                                                              percentages index (SPI)
 


Figure 2. Landslide hazard zonation of Lorestan province
 

Acknowledgments

The Natural Disaster Reduction Committee of Lorestan province provided funding for this research. The paper benefited from helpful assistance of geotechnical department staffs of  IIEES.
The authors are grateful to Mr. Montazer-al-ghaem for his help in preparing of the paper.
 

References used

Carrara A. 1983. Multivariate models for landslide hazard evaluation, Mathematical Geology, 15(3).

Dhakal  S, Amada  T, Aniya M. 2000. Database and geographic information systems for medium scale landslide hazard
       evaluation: An example from typical mountain watershed in Nepal. In: Bromhead E, Dixon N, Ibsen ML, editors.
       Landslides in research, theory and practice, V.1. Proceedings of  an International Conference held at the Cardiff
       Univesity; 2000 June 26-30, UK: Thomas Telford. V.1, p 457-462.

Jafari M, Mahdavifar MR, Keshavarz M. 2000. Geotechnical earthquake  hazards of Lorestan province , International
        Institute of  Engineering Earthquake and Seismology, Report No.2, p 104-169.

Sharp CFS. 1938. Landslide and related phenomena.Columbia university press, NewYork.

Authors
M.R. Mahdavifar, Research Associate
Geotechnical Department, International Institute of Earthquake Engineering and Seismology,P O Box 19395-3913, Tehran, Iran,
Tel: +98-21-283-1116, Fax: +98-21-229-9479
Email: mahdavif@dena.iiees.ac.ir

Masoomeh Rakhshandeh and Piran Veyseh, Research Assistant,
Geotechnical Department, International Institute of Earthquake Engineering and Seismology, P O Box 19395-3913, Tehran, Iran,
Tel: +98-21-283-1116, Fax: +98-21-229-9479
Email: Lily@dena.iiees.ac.ir
Email: piran@dena.iiees.ac.ir