Landslide Vulnerability Assessment (LVAs): a Case Study From Kota Kinabalu, Sabah, Malaysia

DOI: 10.17014/ijog.4.1.49-59The topic on Landslide Vulnerability Assessment (LVAs) in Malaysia is relatively new and received little attention from geoscientists and engineers. This research paper tries to formulate the concept of LVAs by taking into account the science and socio-economic aspects. A new approach in vulnerability concept is also introduced herein. To achieve this goal, a framework was designed for assessing the LVAs. The framework was formulated semiquantitatively through the development of database for the risk elements (human and properties) based on information from secondary data (technical reports), extensive review of literature, and field observations. The vulnerability parameters included in assessing LVAs are 1) physical implication (building structures, internal materials, property damage, infrastructural facilities, and stabilization actions), 2) social status (injury, fatalities, safety, loss of accommodation, and public awareness), and 3) interference on environment (affected period, daily operation, and diversity). Each considered parameter in the vulnerability assessment is allocated with a certain index value ranges from 0 (0 % damage/victims/period), 0.25 (1 - 25% damage/victims/period), 0.50 (26 - 50% damage/victims/period), 0.75 (51 - 75% damage/victims/period), and 1.00 (75 - 100% damage/victims/period). All of these parameters are compiled and analyzed with “Landslide Distribution Map” (LDM) to generate a “Landslide Vulnerability Degree map (LVD)”. The LDM was produced based on field studies and satellite image interpretations in order to locate the landslide locations in the studied area. Finally, three types of physical, human, and environment vulnerabilities were then classified into five classes of vulnerabilities, namely: Class 1 (< 0.20): Very Low Vulnerability; Class 2 (0.21 - 0.40): Low Vulnerability; Class 3 (0.41 - 0.60): Medium Vulnerability; Class 4 (0.61 - 0.80): High Vulnerability); and Class 5 (> 0.81): Very High Vulnerability. Results from this study indicate that a further study is needed to the areas of high to very high vulnerability only. This LVAs approach is suitable as a guideline for preliminary development planning, controlling, and managing the landslide hazard /risk in the studied area and potentially to be extended with different background environments.


Introduction
Vulnerability is defined as the potential degree of loss (damage) to a given element or risk elements resulting from the occurrence of a natural phenomenon of a given magnitude. Vulnerability expressed on a numerical scale from 0 (no damage) to 1 (total damage) depends on the intensity of the landslide that occurred (ISSMGE 2007;UPC 2011).
Vulnerability concept was developed in the context of natural disasters research over the last 30 years; this means that the more days it is becoming increasingly diverse. The explanation for this wide diversity also being doubled and takes the relationship between the human and natures which was triggered by issues of dynamic, multi-dimensional and multi-scalar such as globalization and climate change or the global environment (Rodeano and Tajul, 2012). Due to the absence of border globalization, the various disciplines which differ in their background have defined some sense of vulnerability (Cutter, 1996;Aleotti and Chowdhur, 1999) and as a result there are many mixed methodology and conceptualization of vulnerability.
Since the early 1980s, Timmerman, (1981) in: Fuchs et al., (2007) have indicates that the term "vulnerability" which not only covers the areas of natural disasters, but also applied in other fields such as business, psychology or health society.
Landslide Vulnerability Assessment (LVAs) references can be found in very much quantities. LVAs concept is depends on (a) a runoff; (b) the volume and velocity of slides; (c) risk elements (properties) such as buildings and other infrastructure facilities (nature and proximity against slippage); and (d) elements at risk (life) as humans (vulnerability to disasters, the situation and their position in the building / road) (Finlay, 1996;in Dai et al., 2002) (Fig. 1). How people perceived LVAs approach is depends on many factors: social (Blaikie et al., 1994;Slovic et al., 2004;Jóhannesdóttir and Gisladottir, 2010), psychology (Pidgeon et al., 1992;Blaikie et al., 1994;Jóhannesdóttir and Gisladottir, 2010), economy (Jóhannesdóttir and Gisladottir, 2010), culture (Boholm, 1998;Sjöberg, 2000) and environment (Mileti, 1994;Haynes et al., 2008;Jóhannesdóttir and Gisladottir, 2010) or a combination from all of those factors (Alexander, 2000;Jóhannesdóttir and Gisladottir, 2010). Furthermore, residents or other risk element also has a threat of damage and losses due to factors mentioned above. In this connection, LVAs can raise awareness in determining the loss of an area to arrive at a higher level than the disaster itself.

Study Area
Kota Kinabalu, Sabah, Malaysia (Fig. 2) emerging and growing population. Kota Kinabalu area proposed to be used as a pilot study for Landslide Vulnerability Assessment (LVAs). The impact of rapid development in the study area had led of slope cutting activities and increasingly spread to the hilly terrain. Therefore, LVAs research should be developed and implemented for the preliminary development planning, control and manage the landslide hazard / risk in the study area. The increment in the population growth rate in the Kota Kinabalu is estimated to be increase around 25.6% to 36.6% per 10 years (Jabatan Perangkaan Malaysia, 2008;2009). This information indicates that the element of risk (population, vehicles, infrastructure and property) exposed to vulnerability of landslide were also increased.
The rapid development in the study area is expected to be continued. Development of the area began to grow southwest of Kota Kinabalu City in 1978, right up to the north in the era of 2010 (Fig. 3). Most of the forest area has been explored for the purpose of agricultural development activities. This phenomenan can be proved by the change of forest or farm land area

Material and Methods
Vulnerability concept is often associated with the magnitude of the landslide depends on their propagation distance, volume and velocity of slides, and the risk elements (property and life) are involved. Loss of property is evaluated based on the relative damage to the property value involved. Human vulnerability refers to the probability that the number of victims whether alive or dead.
Landslide Vulnerability Assessment (LVAs) involves observation information about the types of landslide and how its impact can cause damage at different levels. In most literature, LVAs is often associated with expert judgment. This is because most of the existing information is usually incomplete due to lack of data or constraints of data access.
Based on a literature review was conducted, there has not been any consensus approach that can be used as appropriate standards and applied effectively for LVAs in Malaysia. Therefore, a reasonable workflow and more practical have been designed to suit the local conditions (Fig.  6).
LVAs taking into account several parameters such Physical Vulnerability (Vp), Social Vulnerability (Vs) and Environment Vulnerability (Ve). These data are collected through the field observations, compiling a landslide occurrences record and the secondary data. Based on Fig. 4, the first step begins with a literature review and gathering landslide hazard information based on the secondary data. The combination of literature information and the secondary data is to produce the Landslide Risk Element Identification (LREI).
Identification of causal factors of landslide was done in areas were identified as having high of Landslide Hazard Degree (LHD). Based on a combination of LREI (property and life) with vulnerability parameters, a database created and LVAs parameters were listed (Fig. 6) The next step is to observe the data fields for each vulnerability parameter sets. These data are then reanalyzed based on Standardization Method introduced by Voogd (1983). This method aims to rescale the field data by the Linear Transform Numerical (LTN) approach which starting from the 0.00 to 1.00 by applying the following equation: The generation of all the maps mentioned above is conducted by the Geostatistical-Kriging Interpolation Technique (GEOSTAINT-K) introduced by Rodeano et al. (2012). GEOSTAINT-K is the characteristic points in the geostatistical calculations. The goal of this model is to determine the probability variables for each location may not be identifiable or no data. The approach used in GEOSTAINT-K using the geostatistical interpolation.
Finally, Landslide Vulnerability Map (LVM) for Kota Kinabalu, Sabah were generated by the combination of the Physical, Social and Environment Vulnerabilities Map.

Results and Discussion
Physical Vulnerability (Vp) in this study involves the assessment of damage or destruction of the building structure, internal equipment, damage to property, infrastructure and stabilization measures. The proportion of Vp depending on the nature of the risk element is exposed, the mechanism of landslide and the level of danger, building structure, building materials used, the basic structure of the system, the size and shape of the elements of risk and long-life used. Similar damage assessment Vp can be estimated using vulnerability coefficient varying between 0 (no damage) to 1 (total destruction). The results of the Kota Kinabalu Vp indicates that 8.49% of the total area classified as Very Low, 10.28% as Low, 50.18% as Moderate, 29.07% as High and 1.98% as Very High (Fig. 7).
Social Vulnerability (Vs) in this study involves the assessment of the level of injury, death, salvation, homeless and public awareness vulnerability population exposed to landslide.
The proportion Vs involves consideration of the potential or actual victims as a unit. The results of the Kota Kinabalu Vs indicates that 10.39% of the total area classified as Very Low,17.43% as Low,25.47% as Moderate,8.41% as High,and 38.29% as Very High (Fig. 8). Environment Vulnerability (Ve) in this study involves the assessment of the duration of the repair, diversity and daily operations. Environment Vulnerability (Ve) is basically very difficult to implement. For example, the destruction of part or the whole of the agricultural and forestry sectors cannot be measured only in terms of the value of the lost timber, but should be evaluated and analyzed in the context of the increasing potential damage generated. The destruction of the natural environment caused by landslide also involves impairment of plant or animal species habitat. Long-term damage in the agricultural and forestry sector due to fire could also be contributing to the decline in the productivity. The results of the Kota Kinabalu Ve indicates that 6.36% of the total area classified as Very Low, 12.10% as Low, 57.60% as Moderate, 23.6% as High, and 2.86% as Very High (Fig.  9). Landslide Vulnerability Assessment (LVAs) for Kota Kinabalu, Sabah produced by combining or overlaid of all Vp, Vs and Ve maps. Fig. 10 proposes 17.78% of the total area classified as Very Low, 6.25% as Low, 28.56% as Moderate, 11.08% as of High, and 17:53% as Very High. Landslide Vulnerability at a "high" to "very high" degree can leave an impact on individuals and society. If the vulnerability level received only involves with individuals, the level of the vulnerability is not a great hazard. On the other hand, if a society bear landslide vulnerability level on the big stage and the rate of occurrence happens too often and may be at the expense of life and property, the level of vulnerability will be seen as a threat of a major disaster.
Measurement were taken to reduce the level of landslide vulnerability in the study area is to be more complex and difficult for many parties involved have to deal with it. Although the effects of hazards in the study area can be overcome, but the impact of large vulnerability may exist when exposure parameters vulnerability risk continues to rise and adaptation capacity continues to decline.

Conclusion
In light of available information, the following conclusions may be drawn from this study: 1. Landslide Vulnerability Assessment (LVAs) for Kota Kinabalu, Sabah indicates that 17.78% of the total area classified as Very Low, 6.25% as Low, 28.56% as Moderate, 11.08% as of High, and 17:53% as Very High.
2. Landslide Vulnerability at a "high" to "very high" degree can affect the economy and the daily activities of the population.
3. Residential, commercial, public and industrial infrastructure has higher vulnerability rather than the agricultural and forestry areas. It is because most of the population concentrated in the three regions.

This Landslide Vulnerability Assessment
(LVAs) approach is suitable as a guideline for preliminary development planning, control and manage the landslide hazard / risk in the study area and potentially to be extended with different background environments.
(SLAB) scholarship to the author 1 which has been of great financial assistance during this study. Highest appreciations also to the MOE for the fundamental research grant award (FRG0258/SS2-2010) to finance all the costs of this research.