A GIS-Based Landslide Hazard Framework for Road Repair and Maintenance

 

Salah Sadek

Associate Professor, Department of Civil and Environmental Engineering
American University of Beirut
salah@aub.edu.lb

Ramzi Ramadan

Graduate Student Department of Civil and Environmental Engineering
American University of Beirut
rramadan@inco.net.lb

and

Hani Naghi

Graduate Student Department of Civil and Environmental Engineering
American University of Beirut
ha18@aub.edu.lb 

ABSTRACT

The task of effective and optimal resource allocation for road repair maintenance is typically a complex and at times subjective undertaking. In that context, the need for an objective decision support system relevant to the maintenance of a national road network is clear. The work presented in this paper discusses the inception and implementation of such a framework and tool, which integrate the strengths and capabilities of GIS. For the purpose of this research effort, a dynamic tool was developed within ArcGIS® that generates slope stability hazard maps and associated maintenance priority areas as a function of a number of specifically developed climatological, topographical, geotechnical and historical criteria. The proposed methodology was deployed and used in an actual real-world implementation. The study area lies in Lebanon, in the Mount-Lebanon province. Mount-Lebanon was specifically chosen as a case study, given the high degree of variability in topography, geology and rainfall precipitation regimes and the density and complexity of the road network. The global landslide risk zones were interactively generated by cross-referencing and combining various data coverages. Furthermore, a historic data record of actual slope instability occurrences was compiled over three years, and was incorporated in the framework. Based on the generated hazard maps, an optimized maintenance resource allocation scheme is proposed. The approach outlined in this paper could easily be extended to cover the whole Lebanese territory or any other geographical location for which the necessary data layers are developed.

Keywords: GIS; Slope Stability; Hazard Maps; Road Maintenance.

BACKGROUND

Country wide precipitation records in Lebanon for the year 2003 exceeded the average levels established over the past 50 years by approximately 33%. In the winter of that year alone, heavy rains resulted in more than three thousands recorded failures which were reported to the different public agencies and emergency repair and maintenance units at the municipal and state levels. The types of failures varied significantly and included:

 


Figure 1. Examples of failures recorded.

The degree and extent of the damage reported varied from location to location and spanned the spectrum from minor to severe. Different types of expertise were needed to handle such varied and challenging situations. Engineers from different agencies, public and private, were sent in teams to the field in order to achieve the following:

In the absence of any hazard management system or plan which is specifically geared to handle road-related failures and the lack of a database covering past occurrences, no clear criteria or decision strategies could be identified.

It is in this context that the idea of developing an automated decision-aid tool for resource allocation for road maintenance and repair evolved. The use of GIS as a framework for such an assessment and prioritization effort was an easy choice given the capabilities and flexibility inherent in such systems.

A number or researchers and practicing engineers have explored the use of GIS in applications involving the establishment of landslide hazard maps for both static and seismic loading conditions (Zaitchik and van Es, 2003; Malkawi et al., 2000; Khazai and Sitar, 2000). Sakellarious et al. (2001) presented a decision support system which integrates GIS in evaluating slope stability. Their work focused more on establishing characteristic parameters for the soil layers and analysis algorithms for stability calculations. Xie et al. (2003) take the work from the common two-dimensional analysis environment to the three dimensional, with all the associated complexities, which are managed through the use of a GIS backbone. In presenting an integrated GIS-based framework for the siting or evaluation of potential highway alignments, Sadek et al. (1999) included as part of the their assessment framework a slope stability component embedded within the automated GIS-based assessment. The work described above, does not incorporate a feedback scheme which involves actual data from the areas of concern describing location, type and extent of slope instability and more importantly their potential impact on the surroundings and likely repair costs. Such information would be very valuable in building a “memory” to the process, which could then be invested in producing better management decisions and resource allocation priorities. The work described in this paper is driven by those concerns and potential applications.

OBJECTIVES

The objective of the work presented in this paper was to establish a framework for developing a slope stability hazard assessment tool, using state of the art software and digital technologies. A proposed work methodology is presented including: data collection, failure/event classification and/or analysis, maintenance prioritization, and future preventive measures.

The final outcome of the research effort was the development of a decision-aid tool which allows the relevant parties to:

METHODOLOGY FRAMEWORK

Given the limited and fixed yearly resources allocated by local and central governments for road maintenance and repair, the task of prioritizing and optimizing the works is a real challenge. This is more so in regions or areas prone to damage induced by slope instability and road structure failures. This last element, which we will refer to as the Road Damage Hazard Factor is the focus of the work presented in this paper.

In adopting a platform for the proposed decision-aid framework we opted for Geographic Information System (GIS) applications. GIS presents an excellent environment for storing, displaying and maintaining all relevant data. GIS provides the capability to perform the required spatial analyses and incorporate all factors that are mainly related to geography including soil type, geology, topography, climatology and demography. Based on all the above, we have developed a customized application for Road Damage Hazard Factor (RDHF) assessment, by creating user friendly interfaces within the ArcGIS platform (Figure 2).

 


Figure 2. User-friendly interface within ArcGIS for RDHF assessment

The custom-built application presents the user with two interfaces: The first interface is designed for information data entry regarding a particular site and prompts for and/or presents the following information:

The above interface is used for data entry, data retrieving and updating. The second application interface is designed and used for data analysis and processing, as well as thematic mapping generation.

The proposed methodology can be divided into three phases: Field data collection, office data processing, and finally generating the road failure hazard maps, maintenance prioritization, and future hazards assessment (Figure 3).


Figure 3. Schematic of the proposed methodology and framework

At the end of every significant or severe rainstorm, the reported locations of road failures are obtained from the concerned authorities and compiled. This stage is followed by a comprehensive field data collection program which is implemented through site visits to all relevant locations, where the following necessary information is gathered:

Table 1. Risk level assessment – Proposed characterization chart

Failure Consequence
Probability of hazard
Degree of instability: area affected and rate of movement.
Danger to life or disruption of traffic if road fails.No instability. Low, e.g. low probability of instability, small area affected and slow movement. High, e.g. high probability of instability, large area affected and rapid movement.
Low, e.g. agricultural land and low traffic.0 1 2
High, e.g. urban land and high traffic.0 2 3

 

Table 2. Damage level classifications and relevant work needed
Rating Classification Definition Probable works
0Excellent No slope defects.
1Good Some surface instability with minor falls. No ongoing movement. No maintenance needed.
2Fair Minor slips in face. Surface instability, with falls that encroach onto pavement or block drains.
Open cracking or and loosening of minor blocks.
Blocked surface or ground drains. Minor seepage. No ongoing movement.
Clear debris, minor maintenance or repairs.
3Poor Slip affecting part of face
Many falls onto pavement.
Cracking and loosening of blocks.
Water leakage or standing water.
Ongoing movement unlikely.
Some slope stabilisation necessary.
4Very Poor Face slip and threatening property.
Major blocks at risk of falling.
Drainage system not functioning, leakage down face.
Possible ongoing movement.
Major stabilisation or retaining wall needed.
5Critical Major landslide affecting pavement.
Probable ongoing movement.
Retaining wall or major slope stabilisation.

 

The customized GIS application and its database management options are then used to store and maintain all the collected data, and to locate all the visited sites on the base map. At this stage, the geo-processing capabilities of GIS are used to overlay all the mapped failure sites over the already compiled base map and are automatically assigned additional attributes from the following data layers:

Given all the compiled and generated data for the depicted sites, a failure inventory map can be generated and multi-criteria queries and analysis can be performed to produce the following:

CASE STUDY: MOUNT LEBANON

In the following sections, a case study encompassing all the described components of the structured methodology detailed earlier is presented. The Mount Lebanon area was adopted as the study zone. The “Mount Lebanon” administrative unit represents 19% of the total Lebanon area with 24.7% of the total national road network. The base ground elevations range from 0m to 2600m above sea level. The prevailing soil types vary widely: Arable soils, sandy clayey marls/limestones (C2), pale bedded limestone (C3-C5), basal sandstones (C1), grey limestone (J4), local reefal and sandy limestone (J5-J7) (Duberteret, 1960).

Hundred of cases of road failures are typically reported every year. They occur in the winter season (October to March) where total rainfall precipitation over the period ranges from 600 to 1400 mm (Lebanese weather observatory, 2004). In many if not all cases, repairs are implemented relatively quickly without any long term strategy or future preventive considerations.

A comprehensive data collection campaign covering the whole Mount Lebanon area was conducted during years 2003 and 2004 as per the methodology described earlier (Ramadan et al., 2004). The total number of mapped failures was 514 in 2003 and 164 in 2004. The recorded average precipitations were 1148 mm and 713 mm for the years 2003 and 2004 respectively. The correlation between precipitation levels and reported number of failures is clear. It demonstrates the impact of the precipitation levels on the road failure occurrences. Other factors play a significant role and may also impact the scale, location and number of slope instabilities. The comprehensive database of failures and failure types is presented in Figure 4, which indicates that most cases were due to soil and rock instability in addition to structural failures in the road appurtenance.


Figure 4. Failure occurrences categorized by failure type

The distribution of failures overlain on the precipitation contours is illustrated in Figure 5. The correlation between precipitation levels and number of failures is evident. It is important to note here that the observed correlation is also attributed to ground slopes which tend in the Mount Lebanon Zone to be higher in the higher precipitation areas.

The characteristic tools possible in a GIS environment allow the compilation of the slope failure results in a number of other interesting contexts. For instance, if the slope/road failures are cross-referenced using the soil/geology type data layers, slope instabilities appear to occur mostly in sandy clayey marls/limestones (Figure 6).

Furthermore, failure densities showed identifiable patterns (Figure 7), where failure clusters appear to occur within locations that are characterized by the following soil/land covers geological features:


Figure 5. Distribution of failure sites overlain on a map of precipitation levels

 


Figure 6. Failure distribution by soil type

 


Figure 7. Slope failure clusters neighboring fault lines

The effect of slope levels and/or sudden variation in topography is illustrated in Figure 8. The areas where slopes are steepest (closer topographic contour lines) and where abrupt changes (jagged topographic contour lines), appear to correlate with high numbers of reported road failures.


Figure 8. Map delineating the zones of high failure numbers (darkest zones) overlain with the contours coverage.

Maps showing high risk zones can be generated by intersecting all data layers with the factors influencing road/slope stability and comparing those with actual failures reported. Such a map based on the computed RDHF factors was produced for Mount-Lebanon post the 2004 season and is presented in Figure 9.


Figure 9. Hazard Map generated for the ML area highlighting high risk zones

Furthermore, high risk road sections and their neighboring towns/villages can be identified as shown in Figure 10. Such information is very important in establishing emergency alternative routings and prioritizing resources in order to assure that access to remote villages is not compromised and that efficient rescue planning schemes in case of emergencies are developed.


Figure 10. Hazard Maps showing high risk roads and the villages they serve

CONCLUSION

A methodology framework was developed for road failure risk assessment and management using a GIS base reference and the associated GIS capabilities. The tool is designed for road/highway administrators and should allow for more efficient road rehabilitation plans by prioritizing and classifying failure locations and identifying possible correlations with causative factors. The proposed methodology and tool should also help to optimize preventive maintenance by identifying critical zones and producing updated landslide and road damage hazard maps. The proposed tool is embedded fully within ArcGIS and presents user friendly interfaces, which allow ease of access in updating and querying all the collected data. The case study of the Mount-Lebanon region was presented to illustrate some of the capabilities of the tool and the associated possibilities.

REFERENCES

  1. Malkawi A., B. Saleh, M.S. Al-Sheriadeh, and M.S. Hamza (2000) “Mapping of Landslide Hazard Zones in Jordan Using Remote Sensing and GIS,” Journal of Urban Planning and Development, Vol. 126, No. 1, pp. -17.
  2. Dubertret, L. (1960) Carte Geologique au 50000 Feuilles de Hermon. Institut Geographique National, Beirut, Lebanon.
  3. Lebanese Weather Observatory Department, General Directorate of Civil Aviation, Ministry of Public Works & Transport. Precipitation Data (1950-2004).
  4. Mowen X., T. Esaki, G. Zhou, and Y. Mitani (2003) “Geographical Information Systems-Based Three Dimensional Critical Slope Stability Analysis and Landslide Hazard Assessment,” Journal of Geotechnical and Geoenvironmental Engineering, Vol. 129, No. 12, pp. 109-1116.
  5. Ramadan, R., S. Birdand, M. Woodbridge (2004) “Maintaining Lebanon’s Roads Against the Elements,” Proceedings, International Conference on Geotechncial Engineering, Geo-Beirut 2004, pp. 05-110.
  6. Sadek S., M. Bedran, and I. Kaysi (1999) “A GIS Platform for Multi-Criteria Evaluation of Route Alignments,” Journal of Transportation Engineering. V. 125 No.2, pp. 44-152.
  7. Khazai, B., and N. Sitar (2000) “Assessment of Seismic Slope Stability Using GIS Modeling,” Geographic Information Sciences, Vol.5, No. 2, pp. 121-128.
  8. Sakellariou, M.G., and M. Ferentinou (2001) “GIS-Based Estimation of Slope Stability,” Natural Hazard Review, Vol. 2, No. 1, pp. 12-21.
  9. Zaitchik, B.F., and van Es. H.M. (2003) “Applying a GIS slope-stability model to site-specific landslide prevention in Honduras,” Journal of Soil and Water Conservation, Vol.58, No.1, pp. 45-53.

 

© 2005 ejge