Factors In Predicting Landslide Occurrence Biology Essay

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

Landslides are responsible for considerable loss of both money and lives in the entire world. This environmental problem worsens with increased urban development, change in land use and abnormal weather particularly rainfall attributed to climate change. Landslides have been the focus of many studies on scientific research, engineering study and practices and land use policy. The effects of climate variability and climate change on geomorphic processes such as landslide has been recognized by scientist and researchers (Collison et al., 2000; Dehn et al., 2000).

Landslides are generally defined as complex earth slides where earth flows from rotational slides and develop downslope (Casagli et al. 2006). Landslide and slope failure are the same in the Philippine context. The former is characterized by a slow movement of soil material with gentler slope and can be observed at different stages while the latter is faster movement of debris or soil materials in steeper slopes (David and Felizardo, 2006). The occurrence of landslide varies depending of several factors such as relief, geology, tectonic, weathering, erosion and land use (Casagli et al. 2006). Types of landslide come in many forms (e.g. shallow and deep-seated landslides) and largely depend on the frequency and magnitude in terms of intensity and/or duration of rainfall events (Fukuoka, 1980; Canuti et al. 1985; Wieczorek, 1987). In general, deep-seated landslides and shallow landslides are triggered by rainfall in moderate intensity over long period of time and short duration intense rainfall, respectively (Corominas, 2001).

The municipality of Infanta, Quezon is located in the eastern portion of Luzon fronting the Pacific Ocean frequently visited by typhoons. Successive typhoons (Winnie, Violeta and super typhoon Yoyong) hit Infanta from November 13, 2004 to December 03, 2004 causing flood and damages to properties and agriculture sector. Typhoon Winnie brought extremely high amount of rainfall (370 mm) on November 29, 2004 that caused flashfloods and numerous landslides predominantly natural terrain landslides leaving Infanta, Gen. Nakar and Real, Quezon; and Dingalan, Aurora devastated claiming numerous lives and damages to properties. The Office of Civil Defense reported that more than 2.3 million people were affected and about P4.6 billion were lost in infrastructure and damage to agriculture sector. This extreme rainfall recorded is equivalent to 15 day rainfall of November based on the Philippine Atmospheric, Geophysical and Astronomical Services Administration (PAGASA) annual rainfall data from 1951-2002. A mission report by Dai et al. (2007) revealed that prior to November 29, 2004, an unusually large signal was recorded by the crustal stress instrument located at the ground level of Infanta's Municipal Hall indicating stress changes in the earth's crust. Their study also found that majority of the landslide was small in size (e.g. several m3 to hundred m3) and had a shallow failure depth of less than 1m to 3m. The deadliest landslide had a failure volume of 2 x 104 m3 and traveled long distance. Moreover, the mission report stated that according to the people, the disaster can be attributed to the illegal logging and slash and burn practices of the farmers that destabilized the top soil in the area; and the conversion of forests lands to agriculture. Geologist also noted that the unstable strata due to fault lines along Infanta, Quezon and adjacent municipalities contributed to the disaster.

Significance of the Study

While landslide are usually triggered by an external factor such as intense rainfall (Varnes, 1978), other important factors needs to be considered particularly the effect of human practices on the geophysical environment. Systems approach is needed to study dynamic and complex system process such as landslide. To better understand the process involved in landslide, we need to examine inter-relationship that exist between: (a) geophysical factors (geotechnical soil properties, topography, landuse); (b) human agents (agricultural and livelihood practices) and (c) climate factors (rainfall, frequency of typhoons). Based on the above inter-relationships, Landslide Risk Index (LRI) could be define as: LRI = ƒ(geophysical, human, climate).

Agent-Based Modeling (ABM) is a system modeling approach that can be used to understand the complex and dynamic inter-relationships of human, climate and environment. For this study, NetLogo (Wilensky, 1999) will be used as the modeling platform in the development of an ABM of landslide in Infanta, Quezon. NetLogo is an open-source programmable modeling environment for simulating natural and social phenomena particularly suited for modeling complex systems developing over time.


The general objective of this study is to develop an Agent-Based Model (ABM) of landslide in Infanta, Quezon as a functional inter-relationship of: (a) geophysical factors (geotechnical soil properties, topography, vegetation); (b) human agents (agricultural and livelihood practices) and (c) climate factors (rainfall, frequency of typhoons).

Specifically, the study aims to:

Characterize and analyze the geophysical conditions of the study area by determining: a) geotechnical soil properties (e.g. particle size distribution, plastic limit, liquid limit, plasticity index); b) topography (elevation, slope, aspect); and c) vegetation cover.

Characterize the agricultural and livelihood practices of the selected communities in Infanta, Quezon;

Analyze the rainfall pattern and frequency of typhoons in the study area;

Model landslide occurrence as a functional inter-relationship of the geophysical environment, human practices and variability in climate;

Predict the threshold level and the degree of landslide occurrence through ABM; and

Give recommendations on community landslide risk planning and management

Scope and LimitationS of the study

This study is limited to the prediction of landslide occurrence in Infanta, Quezon as a functional inter-relationship between three (3) key factors: (a) geophysical (geotechnical soil properties, topography, landuse); (b) human agents (agricultural and livelihood practices) and (c) climate factors (rainfall, frequency of typhoons). Data will be gathered through household survey, field survey, soil sampling and laboratory analysis. The research will focus on the prediction of the thresholds and degree of landslide based on different scenarios and how it can serve as a decision tool in community landslide risk planning and management.


The study aims to answer the following research hypotheses:

Landslide in Infanta, Quezon is due to the geophysical, human and climate factors.

The variability in rainfall amount and frequency of typhoons will increase the probability of landslide in the upland areas of the study area.

The complex interaction of geophysical characteristics, human practices and variability in rainfall and typhoon frequency can be modeled through ABM.

The threshold level and occurrence of landslide in Infanta, Quezon can be predicted using ABM approach.


Characterization of the Study Area

Secondary data will be gathered from the Municipality of Infanta, Quezon as part of knowledge base development for the characterization of the study area. Municipal and barangay profiles will be gathered to determine the demographic profiles, agricultural and livelihood practices of the selected communities. Soil survey reports, geologic reports and maps, landslide and flood hazard maps will also be gathered. The maps should preferably be in 1:50,000 scale otherwise available maps will be used discussion purposes.

Rainfall data and frequency of typhoons will be collected in the nearby PAGASA weather station located in Infanta, Quezon. Other climatic data such as temperature and relative humidity will also be collected. Climatic data analysis will be carried out using spreadsheet, statistical software packages and Geographic Information System (GIS).

Primary Data Collection

Household survey and Focus Group Discussions (FGD)

A random household survey will be conducted to gather detailed information on landslide occurrence and community agricultural and livelihood practices. About 90 households will serve as respondent of the survey representing the three (3) barangays: (a) Magsaysay for upland; (b) Pinaglapatan for lowland; and (c) Ilog for Coastal. The result of the survey will be encoded in a MS Access database for easy access and retrieval of data. A typology will be developed through cluster analysis of agricultural and livelihood practices. Focus Group Discussion will also be conducted to gather perceptions and experiences on the causes of landslide in the study area.

Soil sampling

To characterize the soils of landslide-prone areas, soil samples will be gathered at various depths of the soil profile on identified sampling sites. Sampling sites would generally represent a recent landslide area. About 1-kg of soil samples at various depths of each sample sites will be taken for laboratory analysis.

Soil analysis

Soil samples will be brought to the Analytical Soils Laboratory (ASL), Agricultural System Cluster (ASC), College of Agriculture, U.P. Los Banos, College, Laguna for analysis using standard methods for Cation Exchange Capacity (CEC), particle size distribution, plastic limit and liquid limit. Plastic and liquid limit test were done following the standard procedure of American Society for Testing and Materials (ASTM) -D4318 (ASTM, 1996).

Model Development

Identification of model variables

The results of the household interview, FGD, soil analysis and gathered secondary data will serve as basis for the identification of key model variables. The relationships between identified variables will be established based on theories, concepts and principles related to landslide from various literature.

Causal loop diagram (CLD)

A causal loop diagram will be developed based on the conceptual framework, identified model variables and their relationships. The CLD will also serve as the model map prior to translation to NetLogo model language. Modification will be done as needed.

Model inputs and equation parameterization

Equations from published research work on landslide will be reviewed for possible adoption in the landslide model development process. The values of different variables, constants and equations based from literature, primary data and result of analyses will be tabulated in spreadsheet.

Map transformation, processing and validation

Maps needed as input in NetLogo will be transformed and processed to ASCII format using ArcGIS 9.2. Prior to map processing and transformation, digital maps (e.g. soil, elevation, slope, and aspect) will be validated in the field by ground truthing.

NetLogo landslide model programming

Equations, variables and their relationships will be translated to java language as set of commands in NetLogo. A model interface with data export function will also be developed in NetLogo to allow data processing in spreadsheet and statistical packages. Input maps will also be imported to the model.

Landslide scenario development and simulation

Different landslide scenarios will be developed for simulations by changing values of a particular model variable. The threshold levels and degree of landslide will be analyzed from the results of the simulations. Appropriate recommendations on community landslide risk planning and management will be drawn from the results of the simulation.

Statistical Analyses

Appropriate statistical analyses (e.g. correlation, regression) will be performed to compare results of the simulation. The parameters to be analyzed will also be identified prior to model simulation.

Conceptual framework of the study

The conceptual framework of the study is shown in Fig. 1. The geophysical environment includes topography (elevation, slope, aspect), vegetation and geotechnical soil properties (e.g. particle size distribution, plastic limit, liquid limit, plasticity index). As one of the factors of soil formation, topography has a profound effect on present vegetation or land use pattern as well as on the type and age of geologic materials. Vegetation on higher altitudes may differ from that of the lower altitudes. Moreover, slope stability for a given area is also related to the type of vegetation, human activities and geologic materials. The interplay with topography, vegetation and geologic material gives the soil unique properties.

On the other hand, the variability in climate such as the amount of rainfall and typhoon frequency can affect the geotechnical properties of the soil. Intense rainfall can cause drastic changes in soil moisture making the soil prone to sudden landslide. Depending of the intensity and amount of rainfall, landslide may be slow, creeping or fast. The human practices such as slash and burn may directly or indirectly influenced landslides. It may alter the physical landscape and vegetation to an extent that the stability of steep slopes is affected. The changes in vegetation can have both hydrological and mechanical influence on slope stability. These drastic changes create an imbalance in the system that could result to landslide occurrences.

Figure 1 The conceptual framework of landslide occurrence in the study area.

Review of Literature

The Landslide Process

Landslide can be described as the movement of mass of rocks, earth or debris down a slope (Cruden, 1991). Generally, it is defined as a complex earth slides where earth flows from rotational slides and develop downslope (Casagli et al. 2006). Landslide occurs as natural process of forming a relief. However, it can also occur as a result of human activities that disturbs the slope stability of an area. It may differ according to their size of displaced mass, shape, moving mechanism, velocity and other characteristics (Casagli et al. 2006). Landslide can also occur in generally low relief areas such as cut-and fill failures (roadways and building excavations), river bluff failures, lateral spreading landslides, collapse of mine-waste piles and other slope failure associated to quarries and open-pit mines.

The commonly used categories of landslide types were proposed by Varnes (1978): falls, topples, slides, lateral spread and flows. The abrupt movements of geologic material mass, such as rocks and boulders that become detached from steep slopes or cliffs are called falls. It is strongly dictated by gravity, mechanical weathering and presence of interstitial water. Discontinuities like fractures, joints, bedding planes are the separations were movement occurs either by free-fall, bouncing and rolling. The second type topples is characterized by the forward rotation of a unit or units above some pivotal point, below or low in the unit. Like falls, it is influence by gravity and forces exerted by fluids in cracks. The more restrictive term, slides, applies only to mass movements. It is grouped into two sub-types: rotational and translational. Slides are characterized by the distinct zone of weakness which separates the slide material from the underlying more stable materials. When the surface of rupture is curved concavely upward it is called rotational slide. Its movement is roughly rotational about an axis that is parallel to the ground surface and transverse across the slide. On the other hand, in translational slide, earth mass moves along a roughly planar surface with little rotation or backward tilting. A block slide is a translational slide in which the moving mass consists of a single unit or a few closely related units that move down slope as a relatively coherent mass. Lateral spread usually occurs on very gentle slopes or flat terrains were the mode of movement is lateral extension accompanied by shear or tensile fractures caused by liquefaction. The fractures are usually initiated by ground motion either natural (earthquake) or artificial (earth moving constructions). The landslide type flows are divided into 5 subgroups: 1) debris flow; 2) debris avalanche; 3) earth flow; 4) mudflow and 5) creep. Debris flows are rapid mass movement with a combination of loose soil, rock, organic matter, air, and water as slurry flowing down slope. They are commonly caused by intense surface-water flow, due to heavy precipitation or rapid snowmelt that erodes and mobilizes loose soil or rock on steep slopes. Its movement is initiated from other types of landslides that occur on steep slopes, are nearly saturated, and consist of a large proportion of silt- and sand-sized material. Debris avalanche as the name suggest movement of debris typically associated to snows that are very rapid to extremely rapid. Earth flow is distinguished from other subtypes by its hourglass shape. It usually occurs in fine-grained materials or clay-bearing rocks on moderate slopes under saturated conditions. However, dry flows of granular materials are also possible. Mudflows commonly called mudslides are earth flows consisting of material that is wet enough to flow rapidly. It contains at least 50 percent sand, silt, and clay sized particles. The last subtype of flows is creep that is the imperceptibly slow, steady, downward movement of slope-forming soil or rock. The shear stress causes the movement sufficient to produce permanent deformation, but too small to produce shear failure. On the field, the occurrence of creep flow can be observed from the curved tree trunks, bent fences or retaining walls, tilted poles or fences, and small soil ripples or ridges. Complex is the last landslide type characterized by two or more of the above mentioned types. It is where multiple landslides are occurring and have occurred in the past parts of a landslide complex may be active whereas other parts may be dormant, stabilized or extinct). The multiple landslide processes could be occurring throughout a landslide complex. Large landslide complexes may also develop over time in areas where landslide-prone conditions exist (e.g. extended slopes, along mountainous sea coast, mountain ridge flank, along river valley, or along an extensive fault zone).

Landslides disaster events received significant attention over the past years particularly in the Philippines. Worldwide, landslides have serious impacts on social and economic aspects of human life. The recognition of the conditions that caused slope failure and the process that triggered the movement of mass is essential to understand landslide. Determining slope failure depends on the slope morphology, mechanical and hydrological soil properties and vegetation cover (Van Asch et al., 1999; Crosta 1998). These factors determine the decisive quantity of rainfall to cause slope failure. The process of landslide is a complex dynamic continuous series of events from cause to effect. It generally occurs as a result of the interaction of the geologic, topographic and climatic factors common to an area and cannot be attributed to a single causal factor (Varnes, 1958). Generally, landslides involve failure of earth materials under a shear stress. According to Varnes (1958) the instability of earth materials is affected by several contributory factors: (1) removal of lateral support leading to instability and actions of erosion, glacier ice, waves, and longshore tidal currents; creation of new slope by previous rockfall, slide, subsidence, or large scale faulting; (2) surcharge which includes natural and human factors; (3) transitory earth stresses such as earthquake, vibrations from blasting, machinery and traffic; (4) regional tilting which causes progressive increase in slope angles; (5) removal on underlying support by undercutting of banks by rivers and waves, sub-aerial weathering, subterranean erosion, human activities such as mining; and (6) lateral pressure due to water in cracks and caverns, freezing of water in cracks and swelling.

Present Methods in Landslide Risk Assessment

Different methods on risk, hazard and vulnerability assessment to landslide have been presented and implemented by several researches (Guthrie et al., 2008; Casagli et al., 2006; Dehn and Buma, 1999; Frattini et al, 2004). The methods are usually based on theory, concepts and establised equations. Methodological approaches generally include spatial modeling, geographic information system (GIS), remote sensing and modeling (cellular automata, general circulation models, deterministic modeling). Landslide hazard map is intended to show the location of possible slope instability. The lack of reliability tests on the procedures and predictions in estimating the probabilities of future landslide occurrence is common in most spatial models.

In a study by Chung and Fabbri (2008) in predicting landslide for risk analysis, they employed two analytical steps: (1) relative-hazard mapping; and (2) empirical probability estimations. Their mathematical model generates a prediction map by dividing an area into prediction classes according to the relative likelihood of future landslide occurrence and conditional by local geomorphic and topographic properties. Their study concluded that good approaches to landslide risk analysis would result to better assessment of uncertainties of explanatory variables. Moreover, spatial probability maps would be very useful for evaluation of data quality and suggestive of best inputs for hazard prediction eventually leading to landslide risk evaluation.

Hydrological models are also used in assessing landslide risk and hazard. However, hydrological physically-based models commonly used to simulate saturated and unsaturated flow in natural slopes requires high data inputs. This type of model is commonly used to perform two- or three-dimensional analyses of single landslide sites, whilst their application to wider and more complex systems is still difficult. Several hydrological models can be linked to slope stability models to enable accurate simulations of the plausible stability conditions to be performed. An example of a hydrological-slope stability model is the CHASM model produced by Bristol University, which combines a finite difference hydrological model, where Darcy's Law is employed for saturated conditions and the Millington Quirk method for unsaturated hydraulic conductivity, with a slope stability model using Bishop's method of analysis (Anderson and Lloyd, 1991).

Cellular automata

Cellular automata (CA) models can be understood as mathematical models that involves grid or lattice of cells in which each cell evolves though a series of discrete time steps based on the value of neighboring cells (Guthrie et al., 2008). It was first constructed 40 years ago (von Neumann 1966). The local neighborhood of cells in a CA models affects the evolution of all cells based on same set of rules and values. CA models are predisposed to exhibit complex, self-organized behavior, though the underlying rules are implied. Recent studies have used CA models to study the behavior of landslides (Guthrie et al., 2008). Research suggests that CA will work well particularly at the regional scale in characterization of landslide (Bak et al., 1988; Avolio et al. 2000; Clerici and Perego 2000; D'Ambrosio et al., 2003; Iovine et al., 2003; Turcotte et al., 2002). In general, most of these landslide modeling research to date deals with the biophysical characteristics and does not include the behavior of human as a factor or as agent in the model.

Systems Approach for Complex Dynamic Environmental Problems

Complexity is a property of a system that referring to the simultaneous presence of simple and complicated behaviors. The big challenge in studying complex system processes is to find ways to identify and model the commonalities of simple behaviors from a broad range of system and at the same time acknowledging the diverse behaviors that distinguishes one system from another. Studies have shown that there is a strong coupling interaction between landscapes and humans as hierarchical complex systems (Werner and McNamara, 2007). These interactions may spread to become regionally or globally persistent at multi-year to decadal time scales. The heterogeneous agents can be modeled through prediction models to determine what actions that represent non-linear behavior of the system and understand the processes involved.

Werner and McNamara (2007) hypothesized that human behavior can be treated as a dynamic system across broad range of scales. It may include stream of consciousness, feelings, communication, emotions, moods, rational thought and analysis, personality, patterns of economic relations, beliefs, world views, customs, culture, decisions, and genetic evolutions. Human impacts the environment in a variety of ways such as enhancing soil erosion by agriculture and construction or deposition of damming streams and flood control, altering vegetation through harvesting and manipulation and modifying the chemical and microbiological context of environmental processes. On the other hands, the environment impacts humans at short time scales by natural disasters (e.g. hurricanes, flood, landslide and earthquakes) that would cause economic damage and lost of lives. In a longer time scales, environmental processes provide human settlement patterns, cultural development and genetic evolution (Davis, 2002; Fagan, 2004; Hewitt, 2000).

Tierney (2001) noted that bulk of disaster research has been done by sociologist, focusing on social dimensions. Practitioners have already taken the lead in using multidisciplinary thinking through incorporating physical conditions, vulnerable groups and organizational issues. The introduction of dynamic modeling is needed to elaborate further system response as a function of time. Agent-Based Modeling (ABM) is a system methodological approach that combines the intuitive appeal of verbal theories and rich qualitative information from surveys with the rigor of analytically tractable mathematical modeling to understand social process, behaviors, emergent properties and their functional inter-relationships. In an ABM, the decision making entity is called an agent. It is a computational entity such as a software program that can perceive and act upon its environment and that is autonomous in its behavior and at least depends on its own experience partially (Weiss, 2001). Agents can represent vegetation, animals, people and organizations; sensor the environment; can communicate with other agents; and can learn, remember, move and have emotions. Using ABM, the interactions between agents and individual decisions of agents can be explicitly modeled.

Geotechnical Properties of the Soil

Geotechnical soil properties can be grouped into basic, index, hydraulic and mechanical (Hunt, 2007). Basic properties include the fundamental characteristics of the materials and provide a basis for identification and correlations. Some are used in engineering calculations. On the other hand, index properties define certain physical characteristics used basically for classifications, and also for correlations with engineering properties. The hydraulic properties, expressed in terms of permeability, are engineering properties. They concern the flow of fluids through geologic media. Lastly, mechanical properties such as rupture strength and deformation characteristics are also engineering properties, and are further grouped as static or dynamic.

The measurement of geotechnical soil properties can be done in situ and in the laboratory. Soil samples and rock cores are, for the most part tested in the laboratory. Rock cores are occasionally field tested. Rock cores are tested in the laboratory primarily for basic and index properties, since engineering properties of significance are not usually represented by an intact specimen. On the other hand soil samples are tested for basic and index properties and for engineering properties when high-quality undisturbed samples are obtained (generally limited to soft to hard intact specimens of cohesive soils lacking gravel size or larger particles).

The basic and index properties of soils are generally considered to include volume-weight and moisture-density relationships, relative density, gradation, plasticity, and organic content. In most cases, the most commonly measured geotechnical properties are liquid limit, plastic limit and shrinkage limit to define the plasticity characteristics of clays and other cohesive materials. This is popularly known as the Atterberg limits that defines the ranges of moisture content that a soil behave as solid, plastic and liquid (reference, xxxx). Liquid limit is defined as the moisture content of the soil as it passes from the liquid to the plastic state as moisture is removed. More discussion of liquid limit. On the other hand, plastic limit is the moisture content as the soil passes from the plastic to the semisolid state as moisture is removed. Shrinkage limit is the moisture content at which no more volume change occurs upon drying.

Add studies pertaining to atterberg application to landslide studies….

Rainfall-Induced Landslides

For most of the landslide occurring not only in the Philippines, rainfall is the common triggering factor. However, to date there is no clear scientific study in the Philippines on the threshold level of rainfall amount and duration to cause landslide. However, the Mines and Geoscience Bureau (MGB) in the Philippines showed that 150mm of rainfall per day is enough to cause landslide. Moreover, an unpublished study by Daag et al., (2006) as cited by Ollet (2008) stated that rainfall threshold to initiate landslide is site specific in the Philippines.

According to Daag et al. (2006), the threshold rainfall to trigger landslides in the Philippines is site specific. The critical rainfall level per day before a landslide is triggered is about 150 mm which is higher than the 100 mm/day global critical threshold value.

Worldwide established thresholds levels.

Different approaches (rainfall thresholds, hydrological models and coupled methods) to explain the relationship between rainfall and slope failures are well presented in many studies worldwide (Croizer 1986; Wilson 1997; Pasuto and Silvano, 1998). Numerous studies have already been done on combining hydrological models with slope stability analyses to understand the mechanism of rainfall-induced landslide. This approach is based upon the different responses of the terrain according to the geological, physical, hydrogeological and soil mechanical characteristics. In addition, the method may incorporate morphological and anthropogenic conditions, which control the surface and subsurface water flow and hence, influence the stability of the slope (Terlien 1998).

Different approaches (rainfall thresholds, hydrological models and coupled methods) to explain the relationship between rainfall and slope failures are well presented in the literature (Croizer 1986; Wilson 1997; Pasuto and Silvano, 1998). Numerous studies have already been done on combining hydrological models with slope stability analyses to understand the mechanism of rainfall-induced landslide. This approach is based upon the different responses of the terrain according to the geological, physical, hydrogeological and soil mechanical characteristics. In addition, the method may incorporate morphological and anthropogenic conditions, which control the surface and subsurface water flow and hence, influence the stability of the slope (Terlien 1998).

For rainfall-induced landslides, a threshold may define the rainfall, soil moisture, or hydrological conditions that when reached or exceeded, are likely to trigger landslides (Crozier, 1996; Reichenbach et al., 1998; Guzzetti et al., 2007). Rainfall thresholds can be defined physically (process-based, conceptual thresholds) or empirically (historical, statistical thresholds) (Corominas 2000; Aleotti 2004; Wieczorek and Glade 2005, Guzzetti et al. 2007).

Literature Cited

Aleotti, P. 2004. A warning system for rainfall-induced shallow failures. Eng Geol 73:247-265.

Avolio, M.V., Di Gregorio, S., Mantovani, F., Pasuto, A., Rongo, R., Silvano, S., W. Spataro. 2000. Simulation of the 1992 Tessina landslide by a cellular automata model and future hazard scenarios. JAG 2:41-50.

Bak, P., Tang, C., and K. Wiesenfeld. 1988. Self-organized criticality. Phys Rev A 38:364-374.

Canuti, P., Focardi, P., and C.A. Garzonio. 1985. Correlation between rainfall and landslides. Bull Int Assoc Eng Geol 32:49-54.

Casagli, N., Dapporto, S., Ibsen, M.L., Tofani, V. and P. Vannocci. 2006. Analysis of the landslide triggering mechanism during the storm of 20th-21st November 2000, in Northern Tuscany. Landslides 3: 13-21.

Clerici, A. and S. Perego. 2000. Simulation of the Parma River blockage by the Corniglio landslide (Northern Italy). Geomorphology 33:1-23.

Corominas, J. 2000. Landslides and climate. In: Keynote lecture of Proceedings of the 8th International Symposium on Landslides. Bromhead E, Dixon N, Ibsen ML (eds.) A. A. Balkema, Cardiff, Wales, 4, pp 1-33.

Corominas, J. 2001. Landslides and Climate. In: Bromhead EN (ed) Keynote Lectures delivered at the 8th International Symposium on Landslides, Cardiff, June 2000.

Chung, C.J. and A.G. Fabbri. 2008. Predicting landslides for risk analysis-Spatial models tested by a cross-validation technique. Geomorphology 94 (2008) 438-452.

Crozier, M. 1996. The climate-landslide couple: a southern hemisphere perspective. Paleoclimate Res vol. 19 (ESF Special Issue 12):329-350.

Croizer, M.J. 1986. Landslides: causes, consequences and environment. Croom Helm, London.

Crosta, G. 1998. Regionalization of rainfall thresholds: an aid to landslide hazard evaluation. Env Geol 35:131-145.

D'Ambrosio D., Di Gregorio, S., Iovine, G., Lupiano, V., Rongo, R., W. Spataro. 2003. First simulations of the Sarno debris flows through cellular automata modeling. Geomorphology 54:91-117

David, R.V. and J.C. Felizardo. 2006. Alternative local disaster countermeasures for Agos River, Infanta, Quezon. pdf file available online: www.internationalfloodnetwork.org. Date accessed: 06 August 2008.

Davis, M., 2002. Late Victorian Holocausts. Verso, London.

Fagan, B.M. 2004. The Long Summer: How Climate Changed Civilization. Basic Books, New York.

Fukuoka, M. 1980. Landslides associated with rainfall. Geotech Eng 11:1-29.

Guzzetti, F., Peruccacci, S., Rossi, M. C.P. Stark. 2008. The rainfall intensity-duration control of shallow landslides and debris flows: an update. Landslides 5:3-17.

Guzzetti, F., Peruccacci, S., Rossi, M., Stark, C.P. 2007. Rainfall thresholds for the initiation of landslides. Meteorology Atmospheric Physics (in press).

Guthrie, R.H., Deadman, P.J., Cabrera, A.R. and S.G. Evans. 2008. Exploring the magnitude-frequency distribution: a cellular automata model for landslides. Landslides 5:151-159.

Hewitt, G. 2000. The genetic legacy of the Quaternary ice ages. Nature 405, 907-913.

Iovine, G., Di Gregorio, S., V. Lupiano. 2003. Debris-flow susceptibility assessment through cellular automata modeling: an example from 15-16 December 1999 disaster at Cervinara and San Martino Valle Caudina (Campania, southern Italy). Nat Hazards Earth Syst Sci 3:457-468.

Pasuto, A. and S. Silvano. 1998. Rainfall as a trigger of shallow mass movements. A case of study in the Dolomites, Italy. Environ Geol 35(2-3):184-189.

Reichenbach, P., Cardinali, M., De Vita, P, F. Guzzetti. 1998. Regional hydrological thresholds for landslides and floods in the Tiber River Basin (Central Italy). Environ Geol 35(2-3):146-159.

Schuster, R. L. 1996. Socioeconomic significance of landslides. In: Investigation and mitigation. Turner, A. K. and R.K. Schuster (eds.). Transportation Research Board Special Report, 247, National Academy Press, Washington, D.C., U.S., 12-35.

Terlien, M.T.J. 1996. Modelling spatial and temporal variations in rainfall-triggered landslides. ITC Publ. 32, Enschede, Netherlands, 50 pp.

Terlien, M.T.J. 1998. The determination of statistical and deterministic hydrological landslide-triggering thresholds. Environ Geol 35(2-3).

Tierney, K., Lindell, J., Perry, M.K. and W. Ronald. 2001. Facing the unexpected: Disaster Preparedness and Response in the United States. Joseph Henry Press. Washington, DC.

Turcotte, D.L., Malamud, B.D., Guzzetti, F., P. Reichenbach. 2002. Self-organization, the cascade model, and natural hazards. Proc Natl Acad Sci USA 19:2530-2537.

Van Asch, T.W.J., Buma, J., Van Beek, L.P.H. 1999. A view on some hydrological triggering systems in landslides. Geomorphology 30:25-32.

Varnes, D.J. 1978. Slope Movement Types and Processes. In: Landslides: analysis and control. Schuster RL, Krizek RJ (eds). Special report 176, TRB, National research Council, Washington DC.

Von Neumann, J. 1966. Theory of self-reproducing automata. University of Illinois Press, Urbana.

Weiss, G. 2001. Multiagent Systems: A modern approach to distributed artificial intelligence, The MIT Press, Cambridge.

Wieczorek, G.F., and T. Glade. 2005. Climatic factors influencing occurrence of debris flows. In: Debris flow hazards and related phenomena. Jakob M., Hungr O (eds.) Berlin, Springer, pp 325-362.

Wieczorek, G.F. 1996. Landslide triggering mechanisms. In: Landslides: investigation and Mitigation. Turner and Shuster (eds.). Transportation Research Board-National Research Council, Special Report, vol 247, pp 76-90.

Wieczorek GF (1987) Effect of rainfall intensity and duration on debris flows in central Santa Cruz Mountains, California. Geol Soc Am. Rev Eng Geol 7:93-104.

Wilensky, U. 1999. NetLogo. http://ccl.northwestern.edu/netlogo/. Center for Connected Learning and Computer-Based Modeling, Northwestern University. Evanston, IL.

Wilson, R.C. 1997. Normalizing rainfall/debris-flow thresholds along the U.S. Pacific coast for long-term variations in precipation climate. In: Debris flows hazards mitigation: mechanics, prediction and assessment. C.I. Chen (ed.) ASCE, 32-43.