THE DEVELOPMENT OF A GLOBAL CONFLICT INDEX AND WEB GIS TO MAP THE INDIRECT IMPACTS OF ARMED CONFLICT

The aim of the research is to create a Conflict Index and web GIS that would enable humanitarian agencies to increase their conflict competencies on a global scale. This idea of creating a ‘Conflict Index’ was initiated by Concern Worldwide as a result of an in-house investigation into how they might become better at responding to the effects of armed conflict, and prevent it becoming a mitigating factor in their ability to deliver effective programs.

Currently, there is no method of quantifying the number of people affected by conflict or the severity of their needs. There are several databases and indices that track the location, duration and intensity of conflicts worldwide, however, there is no estimate of the overall human cost of violent or armed conflict on civilian populations.

As conflict cannot be fully captured by one individual indicator, a simple three-dimensional topology; Security & Social, Coping Capacity, and Consequences, is proposed to measure the indirect impacts of armed conflict on a population/area. This multilayer topology builds up a conflict profile by bringing together ten components and twenty-eight different indicators.

The conflict score is then calculated using the Weighted Sum tool in the Spatial Analyst Toolbox of ArcMap, where each of the three dimensions are treated with equal weight and importance.

The final web GIS will act as an interactive conflict tool that will aid in informing humanitarian organisations on their security planning, decision making and program design, and in turn increase their program competencies in conflict affected countries.

1           Introduction

1.1         Background

Conflict, in comparison even to hunger, climate change or natural disasters, is an unparalleled hazard in its destructive impact on human development. It holds around one fifth of the world’s population under the threat of large-scale organised violence (Beatty 2016).

Armed conflict in particular, which is defined by the Uppsala Conflict Data Program as “a contested incompatibility that concerns government and/or territory where the use of armed force between two parties, of which at least one is the government of a state, results in at least 25 battle-related deaths in one calendar year”, can cause catastrophic impacts in a country by “trapping populations in situations of protracted disaster and the cumulative impacts of impoverishment and vulnerability” (Raleigh et al. 2010; Wallensteen and Sollenberg 1998).

Multiple studies have shown that the impact of armed conflict goes far beyond battle related deaths, and will extend to a breakdown of supply lines, increased endemic hunger and malnutrition by depleting food stocks and absent harvests, loss of assets, mass displacement, deterioration of health services, as well as serious psycho-social trauma to civilians (De Groeve et al. 2016b).

These impacts, if left unaided, pose catastrophic humanitarian risk, which in turn can lead to debilitating transgenerational impacts with the potential to reverse development gains by up to twenty years (Beatty 2016). In a very simplistic view, the poorest regions on the world will be those most affected by armed conflict, and as approximately half of the worlds global poor are currently living in conflict affected zones (The World Bank 2011), it is clear that the need to work more effectively with conflict data has never been more crucial.

The idea of creating a ‘Conflict Index’ using the extensive amount of available conflict data, was initiated by Concern as a result of an ongoing in-house investigation into how field teams might become better at responding to the effects of armed conflict.

Concern, through partnership with the International Food Policy Research Institute (IFPRI), has an existing data-driven publication; The Global Hunger Index (GHI), (Towey et al. 2016). This index is a flagship document in the hunger/nutrition world that uses data gathered from a cross-section of indicators to generate ‘hunger’ scores for every country (Ward and Beatty 2016).

There is currently no definitive method of quantifying the number of people affected by conflict or the severity of their needs. Several NGO’s and Research Agencies have successfully developed databases and indices that track the risk, type, location, duration and intensity of conflicts worldwide, and will be discussed later in this report. However, there is no estimate of the overall human cost of violent or armed conflict on civilian populations. Many people go unaccounted for and the real impact of conflict on the lives of those affected is difficult to determine (Lattimer et al. 2016).

The title of the research project proposed for the MSc is “The Development of a Global Conflict Index and web GIS to Map the Indirect Impact of Armed Conflict.” The aim of the proposed research is to create a Conflict Index web GIS that could be accessed by humanitarian agencies to increase their conflict competencies on a regional or global scale.

The research project has five primary objectives outlined herebelow;

1. To Identify and Select Conflict Indicators;
2. To Develop a Standardised Score to Conflict Indicators;
3. Using the Standardised Indicators to Develop a Model that Generates Conflict Scores;
4. To Produce a Severity Scale;
5. To Produce an Interactive Web-Map;

1.2         Outline

Chapter two discusses the purpose of the research and an overview of how conflict is defined.

Chapter two presents a literature review of existing research considered to be the most appropriately related to this dissertation, as well as giving context to where the research lies within the existing literature. This chapter will also explore similar case studies and discuss in more detail the need for the research.

Chapter three will outline, examine and interpret the specific tools and datasets that have been sourced and are to be used throughout the project. The methodology for developing the indicators will also be explored in detail in this section, as well as demonstrating why a predominantly quantitative research approach was the more suitable in this instance. A detailed evaluation of the methodology applied in similar case studies is presented, followed by an analysis of existing indicators used to define conflict. The chosen indicators and proposed weighting scheme are introduced here, and the development of the conflict model is carefully detailed.

Chapter four combines the results from chapter three to suggest the more accurate and appropriate calculation of the final conflict scores. It will critically analyse and interpret the results of the methodology, while also examining the reliability of the results. The final web-GIS and severity scale will also be illustrated here.

Chapter five will describe the conclusions arrived at. The author will present the results achieved and interpreted throughout the project, compared against the initial aims and objectives. She will summarise the most critical issues or limitations encountered throughout the research, determine any unresolved ideals which could be followed up and developed outside of this project, and identify further areas for research.

2           Literature Review

The objective of this literature review is to understand whether an approach similar to the GHI can be applied to conflict data, and to assess where Geographic Information Systems (GIS) can be applied to address and quantify the impacts of armed conflict and how it has been used in related fields until now.

2.1         Existing Indicators and Weighting Schemes

2.1.1        Global Hunger Index (GHI)

Concern (through partnership with the International Food Policy Research Institute (IFPRI)) has an existing data-driven publication; The Global Hunger Index (GHI) (Towey et al. 2016). This index is a flagship document in the hunger/nutrition world that is designed to raise awareness and understanding of regional and national differences in the struggle against hunger by drawing attention to the issue. It uses data gathered from a cross-section of four indicators and three dimensions to generate hunger ‘scores’ for every country. It then applies GIS technologies to map these scores and create a hunger visualisation tool. These calculated global scores also illustrate drastic differences among regions and countries, enabling humanitarian agencies to improve their presence and resources in affected areas.

Figure 1: GHI Composition Chart (Towey et al. 2016)

Considerable progress has been made since the inception of the index in 2000, with a reported reduction of approximately 29% in the developing world, due in part to a reduction in each of the four indicators; prevalence of undernourishment, child stunting, child wasting, and child mortality (Towey et al. 2016).

In its 2015 publication, the GHI explored the relationship between armed conflict and hunger. It found that those countries with the highest GHI scores tended to be those engaged or recently engaged in armed conflict (Von Grebmer et al. 2015). This finding enforced the idea that humanitarian needs change and methods of delivery need to change when conflict becomes an additional factor with an existing crisis. If nothing is done, conflict runs the risk of becoming a mitigating factor in an organisation’s ability to deliver their work program (Beatty 2016).

This need to be able to work more effectively with conflict data and its impacts has informed the authors thinking, and has led the initial research for this project.

2.1.2        GIS and Population Based Mortality Survey in Iraq

A study was carried out in 2012 that attempted to measure the impact of conflict on a population. It adopted a two-stage cluster sampling method in a population-based mortality survey in Iraq (Galway et al. 2012). While this approach proved to be effective at illustrating the impact, its results were somewhat restricted and narrow as only one indicator, mortality, was considered in the methodology. However, given the complexity of the impact of conflicts on a civilian population, this approach may only represent a fraction of the actual human cost of war, and may fail to account for the fatalities resulting from the denial of access to food, health and water (United Nations Secretary General 2016).

2.1.3        Direct and Indirect Effects of Conflicts, Violence and War on Economic Growth in Sri Lanka

A study carried out in 2010 in Sri Lanka, explored the short and long term, direct and indirect effects of conflict on economic growth (Santhirasekaram et al. 2010). Using empirical findings based on Ordinary Least Squares (OLS) estimation, the study showed that conflict, violence and war, measured by a proxy measure of annual growth rate of tourist arrivals, scored from zero (peace) to ten (high presence of conflict), negatively and significantly affected the economic growth. In fact, the study estimated that approximately one third of all of the country’s outputs were lost due to the direct and indirect effects of conflict, violence and war in the country (Santhirasekaram et al. 2010). Like the previous study discussed however, these results were achieved using only one indicator, and again, a variety of indicators would yield a more comprehensive result.

2.1.4        Global Humanitarian Agency (GHA)

The Global Humanitarian Agency (GHA) proposes that the number of displaced people can provide an indication of the impact of conflict on an area or population (Lattimer et al. 2016). By comparing the age profile of the displaced demographic against the known or historical demographic of the area in question, humanitarian agencies would be better able to understand the risks, needs and capacities of the two sets of vulnerable populations (displaced and non-displaced) and target their efforts accordingly to save lives, reduce poverty and build resilience. Again, this approach may only represent a fraction of the actual human cost of war, and may fail to account for the wider impacts and consequences felt as a result.

2.1.5        Concern Worldwide Conflict Strategy

In its Emergency Directorate on Conflict Strategy, Concern Worldwide outlined the difficulties in tracing the impacts of conflict on civilian populations. As is shown in section 2.1.2 above, often battle related deaths or mortality surveys are relied upon as best-case proxies in measuring the broader indirect deaths that may stem from wartime malnutrition, severe wounding or preventable disease (Beatty 2016). The actual impact of conflict is much greater than this, and will extend to the loss of assets, mass displacement, interrupted agricultural production or education, serious psycho-social trauma and much more (Beatty 2016).

To this end, Concern has defined conflict across four sub-types, Political Conflict, Identity Conflict, Resource Conflict, and Criminality; to aid in developing a more accurate illustration of the impacts created as a result of armed conflict.

Figure 2: Concern Conflict Types (Beatty 2016)

• Political Conflict concerns the use of violence to further political goals
• Identity Conflict concerns conflict caused, driven or intensified by the relationships between groups; such as ideologies, ethnicities, ethno-nationalities, tribes or religions or any other collective identity
• Resource Conflict concerns violent competition over essential or material resources
• Criminality is caused by harmful anti-social behaviours, ranging from opportunistic/sporadic to sophisticated/organised; prompted or encouraged by the weakening or suspension of the rule of law during conflict

2.1.6        Irish Deprivation Index

Other agencies have employed the use of a multidimensional approach in the development and mapping of indicators to assist in strategic planning. The Irish Deprivation Index developed by the health services in conjunction with the Economic and Social Research Institute in Dublin has been examined (Haase et al. 2014). The index presents an area-based statistical deprivation model for the Island of Ireland based on the 2011 Census. It uses a technique known as the Confirmatory Factor Analysis (CFA) that allows for comparable scores, and underlying dimensions of deprivation to be conceptualised and fixed on theoretical grounds. Using this technique, it draws on a combined set of ten indicators in three dimensions to categorise levels of deprivation in a community, and subsequently form a single deprivation index that is used for planning of resources, and targeting of services etc.

Figure 3: All Island Deprivation Index – Model Composition (Haase et al. 2014)

2.1.7        Food Security Portal & Global Food Security Index (GFSI)

The World Food Summit defines food security as “the state in which people at all times have physical, social and economic access to sufficient and nutritious food that meets their dietary needs for a healthy and active life” (Harmon et al. 2003).

The Food Security Portal, facilitated by the IFPRI, builds on this definition, using a multidimensional approach to measure global food crises. The open access portal encompasses a global research based monitoring and capacity strengthening device for successful identification and implementation of the appropriate policy actions in response to food crises (Stewart et al. 2016). It brings together over forty indicators related to food security, economics and human wellbeing, and aims to meet a country’s immediate food security needs, as well as aiding in building long term global food security.

The Global Food Security Index builds on the information gathered in the Food Security portal to measure the global drivers of food security, provide insight into those countries most vulnerable to food security, and finally to assess how resource risks increase vulnerability (Stewart et al. 2016). The index further develops the World Food Summit’s definition of Food Security by looking beyond hunger to the underlying factors affecting food security and considering its core issues of affordability, availability and quality.

The index is developed from a dynamic quantitative and qualitative benchmarking model, constructed from twenty-eight unique indicators, selected based on expert analysis and consultation with a panel of food security experts.

Scores for each of the three dimensions are calculated from reclassifying each indicator to a range of 0-100, where 100 is most favourable. Two weighting schemes can be applied at this point depending on the interest of the user. ‘Natural Weights’ assumes that all indicators are equally important and distributes weights evenly, or ‘Peer Panel Recommendation’ which averages the weightings suggested by five members of an expert panel (Stewart et al. 2016). The latter of the two is the default weighting applied.

In this instance, the final food security score for each country is calculated from a simple average of the three dimensions’ scores, which in turn are calculated as the weighted mean of the underlying indicators. GIS technologies are then applied to map these scores and create a food security visualisation tool.

Table 1: Global Food Security Index Composition

2.1.8        Human Development Index (HDI)

The Human Development Index was created by the United Nations Development Programme to assess a countries human development by measuring its average achievement in three key dimensions; Health, Education and Standard of Living as follows:

Table 2: Human Development Index Composition

Scores for each of the three dimensions are calculated from the equation:

Dimension Index= Actual Value-Minimum ValueMaximum Value-Minimum Value

Equation 1: HDI Normalisation

Note for the education dimension, Equation 1 is first applied to each of the two indicators, and then the arithmetic mean of the two resulting indices is taken (Griffith and Rose 2016).

The final human development score for each country is then calculated from the geometric mean of the normalized indices for each of the three dimensions as follows:

HDI=(IHealth IEducation IStandard of Living)13

Equation 2: HDI Geometric Mean

The HDI only captures a simplified part of what human development entails. It does not reflect on inequalities, poverty, human security, empowerment, etc (Griffith and Rose 2016).

2.1.9        Multidimensional Poverty Index (MPI)

Like the HDI, the Multidimensional Poverty Index was created by the United Nations Development Programme to assess the number of people who are “multidimensionally poor” i.e. suffering deprivations in 33% or more of the weighted indicators (Griffith and Rose 2016).

Using the same three dimensions as those in the HDI; Health, Education and Standard of Living, it identifies household deprivations across ten component indicators.

To calculate a household’s deprivation score, each indicator within a dimension, and each individual dimension is weighted equally to achieve a maximum deprivation score of 100. The final household deprivation score is then calculated by summing the deprivation scores for each indicator.

Table 3: Multidimensional Poverty Index Composition

This score is then inputted into the following equations to calculate the multidimensional poverty score:

Headcount Ratio, H	Intensity of Poverty, A

Equation 3: Headcount Ratio and Intensity of Poverty

Where:

• n is the total population
• q is the number of deprived people
• c_i is the deprivation score that the ith poor person experiences

The final multidimensional poverty score is then calculated from the following equation:

MPI = H x A

Equation 4: Multidimensional Poverty

2.1.10    Armed Conflict Location and Event Dataset (ACLED)

An article presented on the Armed Conflict Location and Event Dataset (ACLED), used a multidimensional approach, similar to that adopted in the Global Hunger Index, to measure the impact of internal conflict on a population (Raleigh et al. 2010). Its aim is to “capture the forms, agents, dates and locations of political violence and protest as it occurs within developing states” (Raleigh and Dowd 2017).

As is the case with the Concern Worldwide Conflict Strategy discussed above, the ACLED does not describe conflict indicators and risk indicators. Instead, it codes nine conflict event types including: three types of battles, violence against civilians, remote violence, rioting and protesting, and three types of non-violent events. See Table 4 for details.

This disaggregation of conflict events enabled research on local level factors such as: coding the actions of rebels, governments, and militias within unstable states, transfers of military control, headquarter establishment, civilian violence, and rioting. Its results were more comprehensive than the single indicator studies discussed above, and allowed the authors to conclude that while internal conflict might involve 15% of a state’s territory, almost half of a state can be directly affected by internal conflict and wars (Raleigh et al. 2010).

Table 4: ACLED Event Types (Raleigh and Dowd 2017)

As the ACLED seeks to capture reported fatalities in political conflict only, it does not capture wider homicides associated with inter-personal violence, criminality etc. As such, this approach may once again only represent a fraction of the actual human cost of war, and may fail to account for the fatalities resulting from wider impacts and consequences.

2.1.11    Heidelberg Institute for International Conflict Research (HIIK)

The Conflict Barometer developed by the Heidelberg Institute for International Conflict Research (HIIK), targets the concept of conflict intensity. It defines conflict intensity as “an attribute of the sum of conflict measures in a specific political conflict in a geographical area and a given space of time” (Werth et al. 2016). Its primary units of analysis are the “region-month”. i.e. the calendar month and the first-level subnational administrative unit of a country.

The HIIK uses a five-tier model to calculate conflict intensity as shown in Table 5 below.

Table 5: HIIK Conflict Intensity Model

These are Dispute, Non-Violent Crisis, Violent Crisis, Limited War, and War. As shown in Table 5, the first two tiers relate to non-violent conflicts where a political conflict is carried out without resorting to violence, or where one of the actors threatens to, but does not use violence. This includes violence against objects without taking the risk to harm persons, the refusal of arms surrender, pointing weapon systems against each other and sanctions (Werth et al. 2016).

The following three tiers relate to violent or armed conflicts, which are the focus of this research. Two dimensions; Means and Consequences; and five proxy indicators; Weapons, Personnel, Casualties, Destruction and Refugees; are used to quantify the intensity of these conflicts as shown in Table 6 below.

Weapons determines whether light or heavy arms are used (e.g. handguns or hand grenades vs. artillery or heavy bombs) (Werth et al. 2016)

Personnel measures the highest number of participants in an individual measure. All persons who, by their actions, collectively represent a conflict actor in the context of a violent measure are counted (Werth et al. 2016)

Casualties measures the overall number of casualties in the conflict in a region-month, comprising the number of deaths from violent measures or their direct consequences. Persons dying due to indirect effects, e.g. starvation or disease, are not counted (Werth et al. 2016)

Refugees & IDPs measures theoverall number of cross-border refugees and internally displaced persons (IDPs) in a region-month (Werth et al. 2016)

Destruction measuresthe amount of destruction resulting from the conflict during the whole month and within the subnational unit is determined in four dimensions considered essential for civil populations: infrastructure, accommodation, economy, and culture (Werth et al. 2016)

Table 6: HIIK Dimensions and Indicators

Each proxy indicator is scored on a ternary scale, which when aggregated, results in the total conflict intensity of a region-month.

2.1.12    Global Peace Index GPI)

The Global Peace Index, developed by Vision of Humanity, provides a comprehensive analysis on the state of peace in a country. Using GIS technologies, it creates a tangible picture of how peaceful a country is by compiling twenty-three core indicators across three dimensions. Each of these indicators were identified and examined by a panel of independent experts, who apportioned scores to each based on their relative importance. Each were scored on a scale of 1-5, where one is very low, and five is very high.

An additional two sub-components; ‘How at Peace Internally a Country is’, and ‘How at Peace Externally a Country is’; are then calculated from the twenty-three indicators, applying a weight of 60% to the measure of internal peace, and 40% to the measure of external peace. This weighting scheme was based on the innovative notion that a greater level of internal peace is likely to lead to, or at least correlate with, lower external conflict (Institute for Economics & Peace 2017).

Table 7: Global Peace Index Composition

2.1.13    Fragile States Index (FSI)

The Fragile States Index is a conflict risk assessment model that ranks countries based on the different pressures they face that impact their levels of fragility. It is based on both GIS technologies and a conflict assessment framework known as the “Conflict Assessment Tool” or “CAST” framework, which assesses the vulnerability of states to collapse (Messner et al. 2017).

It combines twelve equally weighted key risk indicators and over 100 sub-indicators to measure the state of a country at any given moment in time. By calculating it in this way, the indicators produce a snapshot of a given moment in time which can then be compared against previous snapshots to determine whether conditions in any country are improving or worsening (Fund for Peace 2017).

Table 8: Fragile States Index Composition

Using the CAST framework, data from the three streams outlined hereunder is triangulated, before it is subjected to a critical expert panel review.

Content Analysis – The twelve indicators are broken down into sub-indicators aggregated from qualitative data points collected from over 10,000 different English-language sources around the world. provisional scores are then applied to each country.

Quantitative Data – Raw data sets are normalized and scaled for comparative analysis against pre-existing quantitative datasets. Depending on the degree to which the Content Analysis and the Quantitative Data agree, the provisional scores are confirmed, or where they disagree, are reconciled based on a set of rules (Fund for Peace 2017).

Qualitative Review – Independent reviews of each of the country’s scores are carried out based on events from that year compared to the previous one.

2.1.14    Global Conflict Risk Index (GCRI)

The Global Conflict Risk Index (GCRI) is an ‘early warning system’ designed to give policy makers a global risk assessment based on macro-economic factors (Smidt et al. 2016). Essentially, it is an index of the risk of violent conflict in the immediate future and is based on quantitative indicators from open sources.

While Concern Worldwide has defined four conflict types (Political, Identity, Resource and Criminality), the GCRI distinguishes three dimensions of conflict every country faces: the risk of confrontation with other states, the risk of internal conflicts over government control, and the risk of internal conflict over issues apart from government power itself, such as secession or resources (Smidt et al. 2016). It then defines five distinct risk areas for conflict within these dimensions. These are Political, Conflict Prevalence, Social Cohesion and Public Security, Geography and Environment, and finally Economy. It then further classifies these risk areas into ten component concepts and twenty-four individual variables that are based on the results of previous projects (INFORM) and a series of interviews with country and conflict experts.

Table 9: GCRI Composition (Smidt et al. 2016)

These variables are then plugged into two statistical regression models (logistic regression and linear regression respectively) to calculate the probability and intensity of the violent conflict occurring in a country. Combining these two outputs (as per Figure 4) allows a one-stop figure to be created that could be used to separate out the different degrees of civilian impact that each of the five conflict-types might produce i.e. Lack of democracy, food insecurity, stress on infrastructure and resources, transport mobility etc. (Smidt et al. 2016).

Figure 4: GCRI Weighting Composition (Smidt et al. 2016)

Between September 2015 and February 2016, the organisation focused on constructing a composite conflict indicator, and explored several methods of weighting and grouping variables. While it did not produce a definitive method in constructing a composite conflict indicator, it did conclude that a multidimensional approach gives a much more realistic portrayal of conflict (Smidt et al. 2016).

2.1.15    Index for Risk Management (INFORM)

The Index for Risk Management (INFORM) is a joint initiative started by a merging of interests of United Nations agencies, humanitarian agencies and the European Commission. The INFORM model is a risk-assessment tool to aid those engaged in resilience, emergency preparedness, disaster management and humanitarian response in the decision-making process with regards to crisis and disaster prevention, preparedness and response (De Groeve et al. 2016a). It identifies countries at risk from humanitarian crises and disasters that could overwhelm national response capacity, and applies GIS technologies to illustrate the results.

It uses a multidimensional approach to categorize risk into three dimensions; Hazards & Exposure, Vulnerability, and Lack of Coping Capacity. These dimensions are based on “risk concepts” that have been published in scientific literature, and are conceptualised in a counterbalancing relationship. i.e. the risk of what (natural and human hazard), and the risk to what (population) (De Groeve et al. 2016b). It builds up a final picture of risk by compiling fifty-three core indicators across these three dimensions.

The model utilises a balancing system to quantify the risk, pitting the hazard & exposure dimension against the vulnerability and lack of coping capacity dimensions. In this way, factors dependant on hazards are treated solely in the hazard and exposure dimension, while factors independent of hazards are split between the vulnerability and coping capacity dimensions.

As with previous studies outlined above, the INFORM model reclassifies each of the fifty-three indicators to an identical range of 0.0 – 10.0, where zero is very low, and ten is very high for each. It then tested several different methods of aggregating the indicators and components including:

• Minimum: the best indicator only
• Maximum: the worst indicator only
• Arithmetic mean
• Geometric mean
• All variables together, in groups, different combinations of weights on different levels

The final model implements a combination of the arithmetic and geometric averages to the fifty-three indicators at each level of the model. The final risk score is then calculated with the equation outlined hereunder.

Risk = Hazard & Exposure ^1.3 x Vulnerability ^1.3 x Lack of Coping Capacity ^1.3

Equation 5: INFORM Risk Equation

The graphic below illustrates the individual dimensions and composite indicators that make up the INFORM index. Each of these dimensions and indicators are explained in detail thereafter.

Figure 5: INFORM Composition (De Groeve et al. 2016a)

Hazards & Exposure Dimension

The Hazard and Exposure dimension is focused on the probability of the physical exposure to a specific hazard. i.e. there is no risk if there is no physical exposure, no matter how severe the hazard event is. Therefore, the hazard dimension and the exposure dimension are merged into a singular hazard & exposure dimension (De Groeve et al. 2016b). The dimension is split into two equally weighted parts; natural hazards and human-induced hazards.

Figure 6: INFORM Hazard & Exposure Dimension (De Groeve et al. 2016b)

Vulnerability Dimension

The Vulnerability dimension considers the strength of the individuals and households in an area relative to a crisis (De Groeve et al. 2016b). It represents the damaging effects to the economic, political and social characteristics of a community that are susceptible to destabilisation in the event of a hazardous event. All the indicators considered in this dimension are hazard independent characteristics, apart from physical vulnerability which is included in the Hazard and Exposure dimension discussed above.

The dimension is split in two, the Socio-economic category which refers to the general demography of the country, and the Vulnerable Groups category which refers to the smaller social groups with “limited access to social and health care systems” (De Groeve et al. 2016b).

Figure 7: INFORM Vulnerability Dimension (De Groeve et al. 2016b)

Lack of Coping Capacity Dimension

The final dimension considered in the INFORM index is the Lack of Coping Capacity dimension. It considers the resilience of the society, institutional strength, and the country’s ability to cope with disasters in terms of formal organized activities and the existing infrastructure which contribute to the reduction of disaster risk. The dimension is again split in two; the Institutional category which focuses on the existence of disaster risk reduction (DRR) programmes for mitigation and prevention, and the Infrastructure category which measures the county’s capacity for emergency response and recovery (De Groeve et al. 2016b).

Figure 8: INFORM Lack of Coping Capacity Dimension (De Groeve et al. 2016b)

The final aggregation composition of each of the dimensions is shown in

Table10 below.

Table 10: INFORM Aggregation Compostion

2.2         Use of GIS

It is clear from the Literature Review above that GIS has successfully been used to address and quantify complicated elements in the humanitarian field until now. It’s ability to produce unbiased reliable results make it a fundamental tool whose technologies are sure to be grossly expanded.

In choosing the most appropriate method to apply GIS technologies to this field of research, the author strictly adhered to the following considerations taken from the ArcGIS platform:

1. What needs to be analysed?
2. Can the problem be broken into sub-problems?
3. Are there any significant layers?
4. Does the data need to be reclassified or transformed within any layer?
5. Do the input layers need to be weighted?
6. Do the layers need to be added or combined?
7. Do the best/worst locations need to be selected as a result?

Previous studies have adopted a Regression Analysis method to examine the complex humanitarian datasets. The advantage of this method is that the exploring and examining of spatial relationships can help to explain the driving factors behind particular spatial patterns (e.g. GCRI). However, as one of the core objectives of this research study is to allow the user to weight the various indicators according to a particular method, this analysis is not considered to be the most suited.

The Overlay Analysis toolkit on the other hand allows the user to apply weights to several input layers, and combine them to create an integrated analysis. It’s five analysis tools are more commonly used for suitability modelling but were considered to be the most appropriate toolset for this research. The Weighted Sum tool in particular was chosen as it allows the user to “overlay several rasters, multiplying each by their given weight and summing them together”, a key objective in this research study.