Memoire Online - Flood vulnerability assessment of donstream area in Mono basin in Yoto district, south-eastern Togo

FLOOD VULNERABILITY ASSESSMENT IN DOWNSTREAM AREA OF MONO BASIN, SOUTH-EASTERN TOGO: YOTO DISTRICT

ABSTRACT

Flood Vulnerability Assessment in Downstream Area of Mono Basin,

South-Eastern Togo: Yoto District. (November, 2014)

The Mono River in the Yoto district, presents a challenge in terms of repeated flood hazard. The eight selected communities lie in majority in the floodplains of the Mono River and experience year after year flood disaster. This study focuses on flood vulnerability assessment of the downstream part in the Mono River basin in the Yoto district. It analyses the trend in rainfall and river discharge series (1971-2010); it assesses the determinants of flood vulnerability; and it equally computes Flood Vulnerability Index (FVI).

The result reveals a clear evidence of change in precipitation and river discharge patterns during the period of record. It shows an extreme variability in terms of flood magnitude and frequency in the Mono River. Besides, the closeness of households' farmlands to the river body, the type of construction and the position of settlements, the household size, the low level education of household head, the lack of diversification of livelihood strategies, the lack of adequate flood warning system and lack of willingness and ability to take responsive actions coupled with inadequate emergency services, are identified as main determinants increasing communities' vulnerability to flood disaster. Furthermore, FVI offers easy comparison of communities' vulnerability to flood disaster.

Keywords: Trend analysis, Determinant of flood vulnerability, Flood Vulnerability Index, Downstream part of the Mono River basin.

RESUME

Le fleuve Mono, présente un défi majeur en terme d'inondations qui constituent un phénomène récurent dans la préfecture de Yoto. La majeur partie de la population des huit villages est localisée dans le lit supérieur et moyen du fleuve et fait face année après année aux inondations. A cet effet, cette étude a été initiée pour analyser la vulnérabilité des populations aux inondations dans la basse vallée du fleuve Mono dans la préfecture de Yoto. L'analyse a portée sur la variation des précipitations et débits du fleuve de 1971-2010, l'identification des facteurs de vulnérabilités et le calcul des indices de la vulnérabilité d'inondation

Le résultat révèle une claire évidence de la variabilité pluviométrique et des débits pendant la période considérée. Il montre une variabilité extrême quant à la fréquence et l'intensité des inondations. La proximité des terrains agricoles par rapport au fleuve, le type de construction, le faible niveau d'éducation, l'absence de système d'alerte précoce adéquat, et la faible capacité de la population à prendre des mesures appropriées pour faire face aux impacts des inondations sont identifiés comme les principaux facteurs de la vulnérabilité de la population aux inondations. En outre, le calcul des indices de vulnérabilité offre une comparaison facile de la vulnérabilité des communautés aux inondations.

Mots clés: Analyse de la variation; Les facteurs de vulnérabilité; Indice de vulnérabilité d'inondation; Basse vallée du fleuve Mono.

Dedication

who by his grace has seen and guided me in my entire life and through my academic course of study to this level, the glory be to him.

ACKNOWLEDGEMENTS

I would like to start by thanking to WASCAL ( West African Science Service Centre on Climate Change and Adapted Land Use) for offering me a scholarship, which made it possible for me to participate in the Master Programme in Climate Change and Human Security (MRP CCHS) at University of Lomé.

My special thanks go to my committee chair, Dr. Georges Abbevi ABBEY, and my committee members, Prof Amadou Thierno GAYE, and Dr. Komi AGBOKA, for his patience, unceasing and tireless efforts, his guidance and mentorship, and for his encouragement, constructive comments, remarks, suggestions and support during the writing process of this thesis.

I would like to thank our former Director Professor Adote Blim Blivi for his constructive comments and advices to give us the will and strength to be best.

I would also like to express my gratitude to my lecturers and colleagues who helped me and kept me during the two years of this Master.

I would like to sincerely thank our current Director Professor Kokou Kouami and our coordinator Dr Aklesso Egbendewe-Mondzodzo for their encouragement and support.

I would like to express my sincere gratitude to Togo Red Cross Society, Mr André Akpadja , to the team of research assistants for their effort during the process of collecting data for this study and to all the communities and households where the questionnaires were administered without whom this research would not have been possible.

I am grateful to all those who have helped directly or indirectly in the production of this thesis . I render my special thanks to my entire family especially, to my young sister, Ms Kissi Esther, to all my friends, especially, Mr Batadjaga Magloire, Mr Adjaho Iréné, Mr Kpotor Edguard, Mr Wilson-Bahun Noah, Mr Bruce Michel, Mr Etoh Kudzo Sena who listened to my complains, gave me advice and the will to go on, and made me laugh when I needed to. To you, all I owe my gratitude.

LIST OF FIGURES

LIST OF TABLES

LIST OF MAPS

LIST OF PHOTO

SREX: IPCC Special Report on Managing the Risks of Extreme Events and Disasters to Advance Climate Change Adaptation

CHAPTER I. INTRODUCTION

1.1. Problem Statement

Many countries worldwide, whether in Europe, America, Asia, Oceania, Australia or in Africa, are experiencing heavy rains, river overflows, hurricanes, typhoons, tsunamis causing unexpected floods which decimate entirely or partly some localities in all over the world. Floods are among the most recurring and devastating natural hazards, impacting human lives and causing severe economic damage throughout the world (Sadiq et al, 2011, p 85). Floods can be defined as hydrological events characterised by a rapid rise of water flow in the river. They are characterised by long, short and no warning, depending on the type of floods, speed or onset which may be gradual or sudden (Carter 1991, p1). Various elements either climatic or non-climatic influence flood processes resulting in different types of flood. Six types of floods are distinguished: coastal, flash, river, flood due to drainage problems, tsunamis, and tidal wave/bore floods (Jonkman, 2005).

Flood disasters are occurring as a consequence of either natural factors, such as climate change and climate variability or anthropogenic factors, such as socio-economic and land-use developments (Balica, 2009, p 2571). The frequency of those disasters has been increasing over the years, resulting in loss of life, damage to property and destruction of the environment.

Over the last 50 years, there has been a growing body of evidence pointing to the effect of human behaviour on the global natural environment and on the possibility that certain types of natural disasters such as floods may be increasing as a direct consequence of human activity (Guha-Sapir et al, 2004, p 15). Equally, the effects associated with global warming such as sea level rise, more intensive precipitation levels and higher river discharges may be consequences of this as well. Those effects may increase the frequency and the extent of flood hazards on a worldwide scale and make the number of people at risk in developing countries more vulnerable to flood disasters due to high poverty level.

In Africa, floods of different kinds are one of the most common type of disastrous events, and they account for the biggest losses inflicted by natural disasters. The UN Office for the Coordination of Humanitarian Affairs (OCHA) recently stated that, compared with previous years, 2010 has seen the largest number of people affected and dying from flooding. This is consistent with the dramatic rise in flood events that have battered the world, with West Africa being a case in point.

It is understood that flood risks will not subside in the future, and with the onset of climate change, flood intensity and frequency will threaten many regions of the world (Sadiq, 2011, p 85). The Fourth Assessment Report (AR4) by the Intergovernmental Panel on Climate Change (IPCC, 2007) projects that warming in Africa in the 21st century is likely to be greater than the average global warming and does find that extremely wet seasons, high intensity rainfall events, and associated flooding in West Africa are expected to increase by 20% over the next decades. However, it is noted that responses of local communities to the impact of extreme climatic events in many cases in West Africa have mostly been reactive instead of proactive.

Between 1925 and 1992, Togo has recorded 60 urban and rural floods that caused damages and casualties (MERF, 2013, p 13). Flood disasters are not then a recent phenomenon in the country but have become a frequently recurring problem that occur mainly between July and October which inflicts significant environmental, social, and economic damages and affects population safety. According to "EM-DAT" (Emergency Events Database) of the Centre for Research on the Epidemiology (CRED), a number of flood disasters have been recorded during the last 20 years, with particularly severe events occurring in 2007, 2008 and most recently in 2010 "figure 2". According to MERF (2010), the 2010 flood, in Togo has impacted both urban and rural areas throughout the entire country, affecting 82,767 people; 21 persons were reported to have lost their lives, 85 to have been injured, 12,382 houses have been impacted and 7744,24 hectares of land to have been destroyed. Damage and losses were amounted to an estimated 1.1 percent of GDP, amounting to US$38 million.

Figure 1: Number of people affected by flood during the period of (1994-2010)

The study area is made up of eight villages (Mawussou, Djrekpon, Batoe, Tofacope, Atikpatafo, Logokpo, Tchakponou-kondji, Kpodji) from three counties Sedome, Esse-Godjin and Tokpli located along the river in a severely flood prone area. The population has experienced occurrence of floods. Flood events were frequent during the last decade, causing loss of lives, extensive damages to property, including houses, destruction of transport infrastructures, agricultural land, breakdown in education system and food production. In sum flood affects human security in these communities. The number of reported flood disasters during the last 20 years in the Yoto area, occurred mainly in 1995, 1999, 2007, 2008 and 2010, with the 2010 flood being the most severe (UN, 2010) and recently in October 2014. During the 2010 flood, six counties, 35 villages, including the study villages, were impacted; 2081 people were affected and 1496 hectares of crops were destroyed in the area. Supplied by a set of sub-branches, the Mono River with 21,300 km² often undergoes during torrential rains period, the rising of water level followed by high flows causing the overflowing of the river which inundates the selected villages and makes the population more vulnerable to flood disasters. To this are added environmental factors such as fragmentation of the river banks due to erosion effect digging and widening the river channel, the anthropogenic pressure like the construction of Nangbeto dam at the upstream of the Mono basin, deforestation, the demographic explosion and the socio-economic constraints that exacerbate the vulnerability of the population located at the downstream part of the basin (AGO et al., 2005, p 1).

The country's vulnerability is expected to increase as a result of climate change. It follows that both the frequency and the severity of flood hazards have the potential to increase (MERF, 2009). Regardless of the current and the future trend of flood hazards combined with socio-economic constraints of the area, about 74.8% of people are below the poverty line (IMF, 2010, p 17), the occupancy or use of flood-prone areas may involve a degree of vulnerability.

For communities to be protected against damage due to floods, it exists four main type of flood measures that have to be taken into account: the non-structural measures, the structural measures, land planning measures and conducting flood vulnerability and risks assessments, the latter being the first step in disaster risk reduction process. Therefore, to enable decision makers to implement appropriate flood policies in the right place, there is a need to conduct flood vulnerability assessment with a vulnerability score for a systematic understanding of an area; its characteristics related to flood disasters and easily interpret and compare vulnerability of different communities. Thus, the focus of this study is to conduct flood vulnerability assessment of the downstream area in the Mono River basin in the Yoto district through indicator-based vulnerability assessment as proactive response to floods.

1.2 Research Objectives

The overall objective of the current study is to conduct indicator-based flood vulnerability assessment of the downstream part in the Mono River basin in the Yoto district to compute a Flood Vulnerability Index in order to assess the conditions which influence flood damage in the study area and pinpoint the most vulnerable villages to flood for an effective flood risk reduction. More specifically, the present study attempts to:

1. examine the long term trends in rainfall and discharge data for a record period of (1971-2010);

2. identify the determinants of communities' vulnerability to floods under the three factors of vulnerability (Exposure, Susceptibility and resilience);

3. apply FVI methodology to compute Flood Vulnerability Index of the target area.

1.3. Research Questions

· Do the occurrence of floods hazards relate to change in rainfall and river discharge?

1.4. Research Hypothesis

While the frequency and the intensity of flood hazards in the study area may be related to change in rainfall and river discharge patterns, interaction between the human-environment or socio-ecological system could be the major determinant of households, and communities' vulnerability to such hazard. Analysing trend in rainfall and discharge time series and understanding the conditions which influence flood disaster in the study area should be reliable information to pinpoint local hotspots of flood vulnerability.

1.5. Thesis Structure

Chapter I covers the background information, the problem statement, the objectives, the justification, significance of the study and its objectives and the scope of the study. Chapter II discusses the definition of related concepts to the research topic and literature review on index approach to measure vulnerability to natural hazards. Chapter III includes both the research methodology and data collection process to answer the research questions and to test the research hypothesis. Chapter IV explains the empirical findings on the assessment of climate change, flood frequency analysis and flood vulnerability assessment. Finally, the last chapter includes conclusions as well as way forward for future research and the limitation of the current study.

CHAPTER II: LITERATURE REVIEW

2.1. Hazards, Disasters, and Vulnerability

The concepts of hazard, disaster and vulnerability have been extensively used in various disciplines with different meanings. Even for natural hazards, such as floods, no unique definitions and assessment procedures have been widely accepted (Pistrika and Tsakiris, 2007, p 1). Hazard is the probability of occurrence within a specified period of time and within a given area of a potentially damaging phenomenon (Maiti, 2007, p 10). This definition adds both spatial and temporal components to the definition of hazards while another definition from UNISDR (2009, p 17) refers hazard to "a dangerous phenomenon, substance, human activity or condition that may cause loss of life, injury or other health impacts, property damage, loss of livelihoods and services, social and economic disruption, or environmental damage." Hazard is, in the case of river-floods, a natural event that is perceived as a threat and not as a resource by humans (Fekete, 2010, p 31). For the author, hazard is revealed in the state of exposure, when the natural event actually hits the vulnerable elements. In technical settings, hazards are described quantitatively by the likely frequency of occurrence of different intensities for different areas, as determined from historical data or scientific analysis.

Hazard becomes a disaster when it hits a vulnerable community. It causes disaster when large numbers of people are killed, injured or affected in some ways (Maiti, 2007, p 10). In the same line of thought FAO (2008, p 16) points out that disasters of all kinds happen when hazards seriously affect communities and destroy temporarily or for many years the livelihood security of their members. Another definition from ISDR refers disaster to «a serious disruption of the functioning of a community or a society causing widespread human, material, economic or environmental losses which exceed the ability of the affected community or society to cope using its own resources". A disaster results then from the combination of exposure to a hazard, socio-ecological vulnerability that are present, and the limited capacities of households or communities to reduce or cope with the potential negative impacts of the hazard.

Assessing and measuring vulnerability in the context of natural hazards and climate change requires first and foremost a clear understanding of the concept (s) of vulnerability (Birkmann, 2013, p 9 ). Vulnerability is an important concept in human environment research, its conceptualization has been interpreted in many different ways, according to the perception of the researchers. The word "vulnerability" has created important links between different research communities, particularly disaster risk management (DRM), climate change adaptation (CCA), development and resilience research (Birkmann ,2013, p 9).

Cannon (1990) refers vulnerability only to biophysical exposure, where vulnerability is described as a measure of the degree and type of exposure to risk generated by different societies in relation to hazards.

Some studies found that vulnerability only refers to the susceptibility of a given system; United Nation/ISDR (2004) and the United Nation Development Programme (UNDP, 2004) view vulnerability as a human condition or process resulting from physical, social, economic and environmental factors, which increase the susceptibility of a system to be damaged from impact of a given hazard.

Other authors, like Blaikie et al.(1994) and Wisner, et al. (2004) relate vulnerability of a system or a community only to its capacity to anticipate, cope with, resist and recover from the impact of a hazard.

Adger (1999) views vulnerability as a function of two components: the effect that an event may have on humans, referred to as social vulnerability and the risk that such an event may occur, often referred to as exposure.

According to Chamber (1983), vulnerability has two sides: an external side of risks, shocks to which an individual or household is subjected to climate change and an internal side, which is defencelessness, meaning a lack of means to cope without damaging loss.

Numerous studies define vulnerability as being a function of exposure, susceptibility or sensitivity, coping capacity or resilience. Watson et al. (1996), defines vulnerability as the extent to which climate change may damage or harm a system, depending not only on a system's sensitivity but also on its ability to adapt to new climatic conditions. Kasperson et al., (2000) defines vulnerability as the degree to which an exposure unit is susceptible to harm due to exposure to a perturbation or stress and the ability or lack of the exposure unit to cope, recover or fundamentally adapt to become a new system or to become extinct. According to Tuner et al. (2003, p 8075), vulnerability refers to the degree to which a system, subsystem or system component is likely to experience harm due to exposure to a hazard be it perturbation or stressor. For Balica (2007 p 26), vulnerability is the extent to harms, which can be experienced by a system under certain conditions of exposure, susceptibility and resilience. For Damm, (2010), the term vulnerability is taken as a function of exposure, susceptibility, and capacities. According to Fekete (2010, p 31), vulnerability is both a state and a degree: everyone is vulnerable in the state of exposure to a hazard and is vulnerable to a certain degree: vulnerability changes in time and space and aims at identifying and explaining why the object of research is at risk and how risk can be mitigated.

While, IPCC (2007) relates vulnerability to the character, the magnitude and the rate of climate change and variation in addition to the susceptibility and limited coping capacity of a system and IPCC (2012a, p 32) shows how the concept of vulnerability has served as a guiding element to address disaster risk in the context of climate change and climate variability.

The similarity between all of these studies is that they agree on the three factors that define vulnerability. Thus, the vulnerability of a system is not only a function of exposure to hazards, perturbations and stresses alone but also resides in the sensitivity or susceptibility and in resilience or capacity of the system experiencing such hazards. Birkmann (2013, p 10) reviews vulnerability concept from various researchers and concludes that the concept of vulnerability stresses the fundamental importance of examining the preconditions and the context of societies and communities and elements at risk to effectively promote risk reduction and climate change adaptation.

Based on the various views on vulnerability shown above, flood vulnerability in the current study is viewed as the degree of experienced flood harms under certain condition of exposure, susceptibility and resilience factors within the human-environment systems. Therefore, flood vulnerability is taken here as a function of exposure, susceptibility and resilience.

2.2. Flood Vulnerability Factors

The vulnerability of any system (at any scale) is a function of the exposure and susceptibility of that system to hazardous conditions and the ability, capacity or resilience of the system to cope, adapt and/or recover from the effects of those conditions (Smit and Wandel, 2006). Core factors of vulnerability encompass exposure, susceptibility or sensitivity and resilience or coping and adaptive capacities. Exposure generally refers to the extent to which a unit or a system of the assessment (community, city, building) falls within the geographical range of a hazard event (Birkman, 2013, p 25).

According to IPCC (2012a, p 559), exposure describes the presence of people, livelihoods, environmental services, resources and infrastructures or other valuable items in place that could be affected. Exposure to floods could be understood, then, as the presence of valuable items of human-environment, or socio-ecological systems that are present in floods-prone areas. The indicators for this component can be put in two categories; the first one covers the exposure of different elements at risk and the second one gives details on the general characteristics of the flood. While the first category of indicators supplies information about the location, elevation, population density, land-use, their proximity to the river, their closeness to inundation areas, the second category provides information about the frequency of floods in floodplains, their duration and magnitude (Balica, 2007, p 31).

Penning-Rowsell and Chatterton (1977) defines susceptibility as the relative damageability of property and materials during floods or other hazardous events. According to Turner et al. (2003), susceptibility is mainly defined by cross-scale interactions of multiple internal stresses and perturbations. The concept of susceptibility or sensitivity is the vulnerability factor that describes the human-environmental or socio-ecological conditions or current state that can worsen the hazard, or trigger an impact. So, flood susceptibility indicators evaluate the sensitivity of an element at risk before and during a flood event

Buckle (1998) defines resilience as "the capacity that people or groups may possess to withstand or recover from emergencies and which can stand as a counterbalance to vulnerability". According to UN/ISDR (2004), resilience is determined by the degree to which the social system is capable of organizing itself to increase its capacity for learning from past disasters for better future protection and to improve risk reduction measures.

For Turner et al (2003), resilience of the system is often evaluated in terms of the amount of change a given system can undergo and still remain within the set of natural or desirable states.

Based on the above definitions, flood resilience can be seen as the ability of a system or a community to mitigate or minimize threats of floods on itself. Resilience of a system to flood disasters can only be considered with past flood events as it focuses on elements encountered during and after the floods "figure 4"

2.3. Methodology for Measurement of Vulnerability to Natural Hazards

2.3.1. Theoretical and Conceptual Frameworks of Vulnerability

The different views on vulnerability are displayed in various concepts and frameworks on how to systematize it (Birkmann, 2013, p 41). The measurement of vulnerability requires for a model, which delivers the structure, context and objectives of the analysis (Fekete, 2010). The different concepts and models are essential to the development of methods for measuring and identifying relevant indicators of vulnerability (Downing, 2004).

According to Birkmann (2013, p 62), the different conceptual frameworks can be classified into at least six different schools of thought: (a) school of vulnerability frameworks that is rooted in political economy and particularly addresses issues of the wider political economy, such as root causes, dynamic pressures and unsafe conditions that determine vulnerability. It can be illustrated by , for example, the pressure and release (PAR) model published in Blaikie et al. (1994) and Wisner et al. (2004); (b) school of vulnerability that focus on the notion of coupled human -environmental systems and are linked to a socio-ecological perspective and socio-ecology as research school. The social-ecology perspective compared to political-economy, puts the coupled human-environmental system at the centre of the vulnerability analysis and stresses the transformative qualities of society with regard to nature. It can be represented by the framework developed and published by Turner et al.(2003); (c) school of vulnerability that sees vulnerability and disaster risk assessment from a holistic view. It has tried to develop an integrated explanation of risk and particularly differentiate exposure, susceptibility and societal response capacities. A core element of this approaches is a feedback-loop system that claims that vulnerability is dynamic and that vulnerability assessment cannot be limited to the identification of deficiencies. It can be represented by BBC framework published by Birkmann (2006a); (d) school of vulnerability that emerged within the context of climate change science and adaptation research. It focuses on exposure, sensitivity and adaptive capacities as key determinants of vulnerability including physical characteristics of climate change and climate variability. It can be illustrated by (Fussel and Klein, 2006); (e) school of vulnerability that integrates adaptation and coupling processes into a feedback-loop system and process-oriented perspective of vulnerability. It can be illustrated by Move framework published by Birkmann et al. (2013) and finally (f) the school of vulnerability that combines framework of disaster risk research and climate change adaptation represented by the IPCC SREX concept (IPCC, 2012a). It stresses the need to differentiate the physical event from vulnerability in order to maintain the analytic power of the concept vulnerability as a way to show and examine the social construction risk.

Despite the different points of views reveal by the different schools of thought, it is important to acknowledge that they also represent some similarities, such as the understanding that vulnerability is mainly concerned with the preconditions of a society or community that make it liable to experience harm and damage from a given hazard.

2.3.2 Indicators for Measuring Vulnerability

Indicators are widely recognized as useful measurement tools in distinct fields of research (Damm 2010, p 42) but researchers disagree on their definitions. According to Gallopin (1997, p 14) indicator is defined as a sigh that summarizes information relevant to a particular phenomenon. Some authors (Adriaanse, 1995) define indicators in relation to an aggregation process starting with variables or basic data, followed by processed information and indicators, finally ending up with highly aggregated indices. While others view them as a single variable or an output value from a set of data that describes a system or process. According to Birkman ( 2013, p 88), defining indicator in terms of the level of aggregation neglects an essential aspect: goals. For this researcher, every indicator-development process needs to be related to goals, or at least to a vision which serves as a basis for defining the state or characteristic of interest.

The Hyogo Framework for Action (HFA) 2005-2015 stresses the need to develop systems of indicators of disaster risk and vulnerability at national and sub-national levels that will enable decision-makers to assess the impact of disasters (UN/ISDR 2005). An indicator, or set of indicators, can be defined as an inherent characteristic that quantitatively estimates the condition of a system (Balica et al. 2012). «Indicators necessarily limit themselves to the sphere of the measurable» (Moldan and Dahl 2007: 9). A vulnerability indicator can be defined as a variable which is an operational representation of a characteristic or quality of an object or subject able to provide information regarding the susceptibility, coping and adaptive capacity and resilience of a system (Birkman, 2013, p 87).

Vulnerability indicators are widely used in vulnerability assessment. The first step in an indicator-based vulnerability assessment is the selection of the study area; second, one has to select indicators based on criteria, such as the availability of data, personal judgement or previous research. The procedures for indicator selection follow two general approaches. These are deductive and inductive approaches (Adger et al.,2004). In deductive approach, indicators are selected based on relationships established from theories and conceptual frameworks, whilst inductive approach involves statistical procedures to relate a large number of variables to vulnerability in order to identify the factors that are statistically significant. While a range of widely-accepted relevant characteristics and indicators is being presented in literature, (Adriaanse, 1995; World Bank, 2005.), the actual conditions that determine flood vulnerability are, to a certain degree, very site-specific, location, and hazard-dependent (Muller et al, 2011, p 2113). It can be expressed in terms of functional relationships between expected damages regarding all systems and exposure, susceptibility and resilience characteristics of the affected system, referring to all the different types of possible flood hazards (Balica, 2007).

A total of 30 indicators have been identified under the three factors of vulnerability through various literature. Exposure and susceptibility both have a positive influence on vulnerability, and resilience has a negative influence on vulnerability "Table 1"

Table 1: Flood indicators information
No	Defined indicator	Factors	Unit	Functional relationship with vulnerability (+ or -)	References
1	Flood frequency	Exposure	year	Higher is the number of flood events, higher is the vulnerability (+)	Balica (2007)
2	Flood duration	Exposure	days	The higher the flood duration, the higher the vulnerability (+)	Balica (2007)
4	Flood water depth	Exposure	m	The higher the flood water level, the higher the vulnerability (+)	Balica (2007)
5	Proximity of the village to the water body	Exposure	m	The Closer is the place to the river, the higher is the vulnerability (+)	Balica (2007)
7	population in the flood area	Exposure	#	The higher the number of population, the higher the vulnerability (+)	Balica (2012); Fekete (2009);
8	Heavy rainfall	Exposure	mm	The higher the value of the variance, the higher the vulnerability (+)	Balica (2012)
9	Maximum discharge in the past ten years	Exposure	m³/s	The higher the discharge, the higher the vulnerability (+)	Balica (2012);
10	Land use: Farmland	Exposure	%	The higher the %, the higher the vulnerability (+)	Balica (2012); Fekete (2010); Bowen and Riley (2003)
11	Gender	Susceptibility	%	The higher the % of women, the higher the vulnerability (+)	Wisner et al. (2004); Haki et al. (2004); Cutter et al. (2003); Muller et al. (2011)
12	Elderly	Susceptibility	%	The higher the % of elderly, the higher the vulnerability (+)	Clark et al. (1998); Muller et al (2011); Steinführer and Kuhlicke (2007); Thieken et al. (2007); Birkmann et al. (2008)
13	Children under 15	Susceptibility	%	The higher the % of children, the higher the vulnerability (+)	Schneiderbauer (2007); Cutter et al. (2003); Muller et al. (2011); Birkmann et al. (2008)
14	Agriculture workers	Susceptibility	%	The higher the % of household having agriculture activity the higher the vulnerability (+)	Fekete (2010)
15	Female headed household	Susceptibility	%	The higher the %, the higher the vulnerability (+)	McLanahan (1983); Snyder et al. (2006);
16	Literacy Level	Susceptibility	%	The higher the %, the higher the vulnerability (+)	Fekete (2010); Schneiderbauer (2007); Haki et al. (2004); Steinführer and Kuhlicke 2007
17	Household size	Susceptibility	%	The higher the %, the higher the vulnerability (+)	Haki et al. (2004); Cutter et al. (2003); Muller et al. (2011); Martens and Ramm (2007)
18	Number of houses with poor material (wall, roof, floor)	Susceptibility	#	The higher the number of houses with poor material, the higher is the vulnerability (+)	Schneiderbauer (2007); Clark et al. (1998); Cutter et al. (2003); Muller et al (2011)
19	Past experience	Susceptibility	%	The lower the %, the higher the vulnerability (+)	Balica (2007); Birkmann (2005a); Velasquez and Tanhueco (2005); Wisner et al (2004); Muller (2011)
20	Preparedness	Susceptibility	%	The lower the % of people with flood experience, the higher the vulnerability (+)	Balica (2012); Birkmann (2005a); Velasquez and Tanhueco (2005); Wisner et al. (2004); Cardona (2003); Muller (2011)
21	Awareness	Susceptibility	%	The lower the % of people, the higher the vulnerability (+)	Balica (2007)
22	Emergency services	Resilience	%	The higher the % of people reported to get help from government or institution during and after flood, the lower the vulnerability (-)	Balica (2007)
23	Ability to evacuate	Resilience	%	The higher the %, the lower the vulnerability (-)	Cardona (2003); Muller (2011); Balica (2012); Birkman et al (2013)
24	Knowledge about private protection measures	Resilience	%	The higher the %, the lower the vulnerability (-)	Muller et al (2011)
25	Knowledge about flood hazard	Resilience	%	The higher the percentage, the lower the vulnerability (-)	Cardona (2003); Muller (2011)
26	Warning system	Resilience	%	The existence of warning system lowers the vulnerability (-)	Balica (2007); Balica(2012); Veenstra (2013)
27	Recovery Time to flood	Resilience	%	The faster the recovery time, the lower the vulnerability (-)	Balica (2012)
28	Emergency service	Resilience	%	The higher the %, the Lower the vulnerability (-)	Balica (2012); Aall and Norland (2005); Veenstra (2013)
29	Long term residents	Resilience	%	The higher the %, the lower the vulnerability (-)	Fekete (2010)
30	Environmental recovery	Resilience	%	The higher the %, the lower the vulnerability (-)	Balica (2007)

2.4. The Index Approach to Study Vulnerability

In literature, quantitative assessment of vulnerability is usually done by constructing a vulnerability index. This index is based on several sets of indicators that result in the vulnerability of a region. It produces a single number, which can be used to compare different regions. Literature on index number construction specifies that there should be good internal correlations between these indicators.

Different methodologies have been used to compute a Flood Vulnerability Index (FVI). All FVI equations have factors for exposure to hazard, sensitivity or susceptibility of the people, and their resilience or coping capacity to the hazard. Vulnerability is the result of the combination of exposure, susceptibility and resilience.

Atkins et al. (1998) studied the methodology for measurement of vulnerability and constructed a suitable composite vulnerability index for developing countries and island states. Their composite vulnerability indices were presented for a sample of 110 developing countries for which appropriate data were available. The index suggests that small states are especially prone to vulnerable events when compared to large states. Among the small states, Cape Verde and Trinidad and Tobago are estimated to suffer relatively low levels of vulnerability and majority of the states estimated to experience relatively high vulnerability; and the states like Tonga, Antigua and Barbados being more vulnerable to external economic and environmental factors.

Chris Easter (2000) constructed a vulnerability index for the commonwealth countries, which is based on two principles. First, the impact of external shocks over which the country was affected and, second, the resilience of a country to withstand and recover from such shocks. The analysis used a sample of 111 developing countries of which 37 small and 74 large for which relevant data were available. The results indicate that among the 50 most vulnerable countries, 33 were small states with 27 being least developed among them.

Moss et al. (2001) identified ten proxies for five sectors of climate sensitivities which are settlement sensitivity, food security, human health sensitivity, ecosystem sensitivity and water availability. They equally established seven proxies for three sectors of coping and adaptive capacity: economic capacity, human resources and environmental or natural resources capacity. These proxies are aggregated into sectoral indicators, sensitivity indicators and coping or adaptive capacity indicators and finally help in constructing vulnerability resilience indicators to climate change.

Dolan and Walker (2003) discussed the concept of vulnerability and presented a multi-scaled, integrated framework for assessing vulnerabilities and adaptive capacity. Determinants of adaptive capacity include access to and distribution of wealth, technology and information, risk perception and awareness, social capital and critical institutional frameworks to address climate change hazards. These are identified at the individual and community levels and situated within larger regional, national and international settings.

Katharine Vincent (2004) created an index to empirically assess relative levels of social vulnerability to climate change-induced variations in water availability that allow cross-country comparison in Africa. An aggregated index of social vulnerability was formed through the weighted average of five composite sub indices, which are economic well-being and stability, demographic structure, institutional and strength of public infrastructure, global interconnectivity and dependence on natural resources. The results indicate that using the current data, Niger, Sierra-Leone, Burundi, Madagascar and Burkina-Faso are the most vulnerable countries in Africa.

2.4.1.Existing Flood Vulnerability Index

Connor and Hiroki (2005) presented a methodology to calculate a Flood Vulnerability Index (FVI) for river basins, using eleven indicators grouped into four components. The index uses two sub-indices for its computation: the human index, which corresponds to the social effects of floods; and the material one, which covers the economic effects of floods. The purpose of the FVI is to serve as a tool for assessing flood risks due to climate change in relation to underlying socio-economic conditions and management policies.

An elaborated methodology to calculate FVI was developed by Balica (2007), using indicators which aims at assessing the condition that favour flood damages at various levels: river basin, sub-catchments and urban area. The methodology focused on two concepts: factors of vulnerability based on three elements, including exposure, susceptibility and resilience on one hand, and components of vulnerability including actual flooding and establishing the elements of a system that suffer from this natural disaster on the other hand. The methodology has been applied at different scales and has resulted in interesting observations as to how quantifiable indicators can reflect backs. Balica defines vulnerability as a function of exposure, susceptibility, and resilience.

The Seventh Framework Programme (2011) defined the FVI in terms of the following factors: exposure, susceptibility, and lack of coping capacity. The methodology included a step of converting the indicators into non-dimensional units, by interpolating the maximum and minimum of the series of data obtained. The FVI values oscillate between 0 and 1, where 1 means the highest flood vulnerability and 0 represents the lowest vulnerability to floods. The methodology was tested in Japan river basins and in 18 river basins in Philippines.

Depending on the equation used, the indicators will have to have a different format, but the result of the FVI remains the same. The goal of the equation of the FVI is to compare different communes to one another in overall vulnerability, but also in its separate factors exposure, susceptibility and resilience. To make it possible to visualize these separated factors, a summation relationship is more useful. Also, it is preferred if the resilience is negatively formulated, and a higher score causes the vulnerability to be higher, conform other factors. With the chosen equation, the indicators have to be measured on a scale from 0-100% or 0-1, like Balica et al. (2012). Then, the indicators have to be normalised. The method of normalization has to take into account the functional relationship between the variable and vulnerability. If the functional relation is ignored and if the variables are normalized simply, the resulting index will be misleading. After computing the normalized scores the index is constructed by giving either equal weights to all indicators/components or unequal weights. These factors are then summed up according to the equation, and the result is a 0-100% or

CHAPTER III: RESEARCH METHODOLOGY

3.1 The area of study

3.1.1 Localisation

Mono River system is the largest river system in Togo with catchments area of 21500 km²; it serves as eastern boundary between the Yoto district and the Republic of Benin. The district is located in South-Eastern Togo, North-East of the Maritime region. It is geographically bound by latitude 6°30` and 6°60`N, longitude 1°20` and 1°35`E. It is bordered by the Haho district to the north, Bas-Mono and the Vo districts to the south, the Zio districts to the west and Republic of Benin to the east. The study was conducted in the downstream area of the Mono River basin in eight villages of Sedome (Mawussou, Djrekpon, Batoe), Esse-Godjin (Tofakope, Atikpatafo), and Tokpli (Kpodji, Tchakponou-kondji and Logokpo) counties, in the Yoto district "Map 1" .The selected villages fall under the hazard prone area, where populations have been affected, especially during 2010 flood event, providing then a better study population who can help us to generate a better view on the assessment conducted.

3.1.2. Landscape, soil and vegetation

The study area is formed by hydromorphous soils which are rapidly saturates of water. The sand contents decreases, depending on the closeness of the area to the river. The geology consists of the continental shelf called the terminal plate which extends from Kouvé area to the north-western of Sedome.

The vegetation is a savannah and is composed of the classified and gallery forests and various grassland grasses.

The fauna consists of mammals (buffalo, warthogs, monkeys, deer, agouti etc.) and various birds of prey, aquatic life, crocodiles and hippos.

3.1.3. Climate and Hydrology

The study area, which is at an altitude that ranges from 17 to 55 meters above sea level, has Guinean sub-equatorial climate with two distinct rainy seasons separated by dry periods which are influenced by the movement of two (2) types of winds at different times of the year. The mean annual temperature ranges from 22°C to 30°C and precipitation varies between 800mm and 1200 mm/year; this usually peaks in May-June and September-October. The Mono River has a pluvial law which has changed in the downstream part of the basin due to the construction of Nangbéto dam in 1987 for hydroelectric purposes. Thus, it passed from the irregular to a relatively regular flow due to the release of water of the dam. Before the construction of the Nangbéto dam, the Mono River presented the phases of low water with null flow and height from mid-December to the third week of May, whereas from May until December the river experienced high flow with average maximum of (450 m3/s) in September. This is changed after the construction of the dam with a relatively permanent out-flow at the downstream part ( Ago et al 2005).

3.1.4. Population and Economic Activities

The study area is made up of three counties (Sedome, Esse-Godjin and Tokpli). According to the Togo Population and Housing Census Report in 2010, the total population of the three counties was estimated at about 34918 with 10803 in Sedome, 9261 in Esse-Godjin and 14854 in Tokpli. The majority of the population is located in the River floodplains. Agriculture is the most important activity being carried out in the area with a majority of the people living practising subsistence farming.

The fertile soils coupled with the abundant rainfall per year ensure ample yields of food crops. The main crops grown in the area include maize, cassava, sugarcane, beans, groundnut, palm trees and some vegetables.

The people in the targeted area also keep animals such as goats, cattle, pigs and chicken. Other activities in the targeted area include trading, fishing, palm oil production etc.

3.2. Methods

This chapter describes the methods that are used in executing this study. Construction of vulnerability index consists of several steps. First is the selection of study area which consists of several villages. In each village a set of indicators are selected for each of the three components of vulnerability.

3.2.1. Study Population and sampling

This study is carried out in eight villages (Mawussou, Drekpon, Batoe, Atikpatafo, Tofakope, Tchakponou-kondji, Logokpo and Kpodji) from three counties (Sedome, Esse-godjin, Tokpli) in the Yoto district, popularly known to be associated with flood. The choice of the counties and the villages is based on information obtained from literature and further confirmed from Togolese Red Cross institution which is highly involved in Disaster Risk Reduction and Adaptation to Climate Change in the Yoto district.

Data were collected through personal interviews from two hundred and twenty one (221) households randomly sampled from the selected villages.

3.2.2. Selected Vulnerability Conceptual Frameworks

The current study relies on Turner et al's vulnerability framework . It focuses only on the vulnerability part of the framework in red "figure 5"

Figure 4: Turner et al's Vulnerability Framework; Source: Turner et al., 2003, p 8076

The Turner et al's vulnerability framework is selected for the present study for many reasons. It illustrates the interactions involved in vulnerability analysis, drawing attention to the array of factors and linkages that potentially affect the vulnerability of the coupled human-environment system in a place. It facilitates the identification of critical interactions in the human-environment system that suggest response opportunities for decision makers. It is opened to the use of both quantitative and qualitative data. It also illuminates the nested scales of the vulnerability problem but provides an understanding of the vulnerability of a particular place. This study focused on the local level `village' as a unit of analysis. The main factors of the framework that were tackled in the present study are presented in "figure 6".

Figure 5: Vulnerability components extracted from Turner et al., 2003 framework

3.2.3. Flood Vulnerability Indicator Development

In this study, only the deductive approaches were used to select indicators to serve as proxies of human-environment vulnerability to flood disasters. The field survey and interviews that were carried out in the scope of this research showed whether the selected indicators are most relevant for flood vulnerability analysis in the study area taking into account the local knowledge and perception of the affected people. Those indicators which fitted the local conditions best were combined into the composite vulnerability index "Table 2".

No	Defined Indicators	Factors	Abbr	Functional relationship
1	Population in flooded area	Exposure	E1	(+)
2	Women (%)	Exposure	E2	(+)
3	Children (%)	Exposure	E3	(+)
4	Elderly (%)	Exposure	E4	(+)
5	Return period ( year)	Exposure	E5	(+)
6	Flood duration (days)	Exposure	E6	(+)
7	Flood depth (m)	Exposure	E7	(+)
8	Flood magnitude (m³/s)	Exposure	E8	(+)
9	Village proximity (m)	Exposure	E8	(+)
10	Farmland in flood area (ha)	Exposure	E10	(+)
12	Education : no schooling (%)	Susceptibility	S1	(+)
13	Household size (more than 10)%	Susceptibility	S2	(+)
14	Female headed (%)	Susceptibility	S3	(+)
15	Farmers (Solely) (%)	Susceptibility	S4	(+)
16	Poor building material (%)	Susceptibility	S5	(+)
17	Household with affected land (%)	Susceptibility	S6	(+)
18	Community Awareness (%)	Susceptibility	S7	(+)
19	Household Coping mechanisms (%)	Susceptibility	S8	(+)
20	Emergency service (%)	Susceptibility	S9	(+)
21	Household Past experience (%)	Susceptibility	S10	(-)
22	Household Preparedness (%)	Susceptibility	S11	(-)
23	Warning system (%)	Resilience	R1	(-)
24	Household perception on flood risk(%)	Resilience	R2	(-)
25	Household Evacuation capability (%)	Resilience	R3	(-)
26	Household flood Training (%)	Resilience	R4	(-)
27	Recovery capacity (%)	Resilience	R5	(-)
28	Recovery Time (%)	Resilience	R6	(-)
29	Long term resident 10 years + (%)	Resilience	R7	(-)
30	Environmental recovery (%)	Resilience	R8	(-)

3.3. Data collection and Analysis

This part describes data collection processes as well as data analysis methods.

3.3.1. Primary Data Collection

Field work plays a very important role in collecting primary data. By applying simple random sampling technique, information were collected through questionnaire-based interviews at household levels and personal observation. The questionnaires are designed based on selected indicators developed under Turner et al (2003) vulnerability framework "Table 2". One question for every indicator under each factor is displayed "Annexe 8".

They were obtained by directly talking to the interviewees at household level so as to get very reliable and accurate information because they were the ones directly affected by the flood disasters and whose livelihood was being disrupted. The households were interviewed from their individual homes.

To ensure the primary data quality, research assistants were recruited and trained on how to administer questionnaires and collect quality data. They were familiar with the study area and fluent in the local language (Ewe) and French. The questionnaires were pre-tested and edited to cover identified gaps. The researcher and the research assistants were together in the field during the data collection period. Additionally, supervision was done continuously and meetings were held with research assistants on a daily basis to address any challenges that were met during the data collection process.

3.3.2. Secondary Data

The secondary data included rainfall and river discharge data; topographical sheet. "Table 3 ".

3.3.3. Data Analysis

Articles, theses, reports etc... processing consisted in the reading of documents in order to come up with a consistent literature review. After this first step, the trend detection analysis in the annual and seasonal datasets was accomplished to assess climate change over the study area. In addition, the analysis of vulnerability's determinants under exposure, susceptibility and resilience and computation of flood vulnerability index were carried out. Moreover, an accurate flood vulnerability maps were created using ArcGIS techniques.

3.3.3.1. Trend Analysis of Rainfall And River Discharge to Assess Climate Change.

This study examines trend in river discharge and rainfall in the downstream part of the Mono basin, using Mann-Kendall statistic test. Athieme flow gauging station (downstream) and Tabligbo rainfall gauging station were selected "Table 4". Each station had a long record of 40 years (1971-2010) of data to determine whether or not there have been any significant changes in those variables over the downstream part of the river basin using Mann-Kendall test run at 5% significance level on time series data. Available monthly rainfall and daily river discharge data were first grouped into monthly, seasonal and annual average data. Missing data were filled through linear interpolation of the same months data of the contiguous years on either side of the missing value.

Mann-Kendall test was formulated by Mann (1945) as non-parametric test for trend detection and the test statistic distribution was given by Kendall (1975) for testing non-linear trend and turning point. This test, is widely employed in various studies to ascertain the presence of statistically significant trend in hydrologic and climatic variables with reference to climate change (Yu et al.1993; Douglas et al. 2000; Hess et al.2001; Burn and Elnur 2002; Yue et al. 2003; Burn et al.2004; De Toffol et al., 2008; Singh et al. 2008). There are two advantages of using this test. First, it is a non-parametric test and does not require the data to be normally distributed. Second, the test has low sensitivity to abrupt breaks due to inhomogeneous time series. According to this test, the null hypothesis H0 assumes that there is no trend and under the alternate hypothesis, it is assumed that a significant change has occurred over time, or that an increasing or decreasing trend is evident in the time series.

In this study, trend analysis has been done by using non-parametric Man-Kendall test together with the Sen's Slope Estimator (Qi) for the determination of trend and slope magnitude to find out the annual and monthly variability of rainfall and discharge data over the Mono basin.

The null hypothesis is tested at 95% confidence level for both rainfall and discharge data. If the p value is less than the significance level á (alpha) = 0.05, H0 is rejected. Rejecting H0 indicates that there is a trend in the time series, while accepting H0 indicates no trend was obtained.

Positive value of Qi indicates an upward or increasing trend and a negative value of Qi gives a downward or decreasing trend in the time series. Statistical Mann-Kendall test and Sen's Slope Estimator Test were performed, using Addinsoft's XLSTAT 2014 software.

3.3.3.2. Analysis of the determinants of communities' vulnerability to flood

The magnitude of an extreme event is inversely related to its frequency of occurrence, very severe events occurring less frequently than more moderate events (Maiti, 2007, p 44). The objective of frequency of occurrence is obtained through the use of probability distributions. Some of the commonly used probability distributions are: Gumbel's or Extreme Value type 1 distribution (EV1); Log-Normal distribution; Log-Pearson type III distribution (LP3), and Method of plotting position.

For this study, further insight into flood frequency is provided by the return period analysis. The return period was obtained using the most efficient formula for computing plotting positions for unspecified distributions and now commonly used for most sample data: the Weibull equation (1). The objective of the method is to build the relation between the probability of the occurrence (return period) of a certain event and its magnitude. Frequency is how often an event of a given magnitude may be expected to occur in the log-run average.

The annual peak discharge data of the Mono River at Athieme station (1971-2010, N= 40 years) is selected for flood frequency analysis. A simple technique was to arrange the given peak in descending order of magnitude and assigned an order number (m). The probability of occurrence for each observation is given by:

Where: P= Probability of occurrence; m= order number of the event; N= Total number of events in the data; The return period for each observation was determined using the following formula:

Depending on the flood peaks recorded in 2010 for the study area and the average flood peaks for the examined period, floods are classified according to their magnitude.

Data on flood duration and flood water levels were obtained from each household from interview. The interviewed household could recall the peak duration of flooding during the latest more severe flood (2010). The average days recorded from household interviews was calculated for each village. Flood water levels were measured inside the house as revealed by marks on building walls with reference to the ground floor during the interviews. Only houses in the main village (populated area) were considered. The flood water levels were ranged from the lowest level to the highest level for each village.

3.3.3.3 Analysis of human-environmental condition

Statistical analyses were used as the methods for human-environmental condition components analysis. It includes descriptive statistics to describe all the data in general.

3.3.3.4 Computation of Flood Vulnerability Index (FVI)

The collected data were arranged in the form of a rectangular matrix with rows representing villages and columns representing indicators. In order to obtain figures which are free from the units and also to standardize their values, the indicators were normalized so that they all lie between 0 and 1. After computing the normalized scores the index is constructed by giving unequal weights to all indicators.

Two types of functional relationships are possible: vulnerability increases with increase (decrease) in the value of the indicator. The study used then two formula to normalise indicator, depending on their functional relationship with vulnerability. Then, in case that the indicator has an increase functional relationship with vulnerability (positive indicators), the normalisation is done using the following formula:

On the other hand, in case that the indicator has a decrease functional relationship with vulnerability (negative indicators), the normalized score is computed using the formula:

Xij denotes the value of j indicator (j=1, 2, .........30) in the i village (i=1, 2-, ....8).

It is obvious that the scaled values of Yij lies between 0 and 1. The value 1 corresponds to that village with maximum value and 0 corresponds to the village with minimum value. Through those formula the normalised scores for each indicator were obtained using MS-EXCEL Max() and Min() functions.

2) Method of Weighting and Aggregation of Indicators into Vulnerability Index

After computing the normalized scores, the index is constructed by giving an unequal weight to all indicators. In literature, several methods are used to give weight to indicators either equal weights (simple average of the scores and Patnaik and Narain Methods) or unequal weights (Expert judgement and Iyengar and Sudarshan's methods) or multivariate statistical techniques (Principal components and cluster analysis method).

The present study uses an unequal method of Iyengar and Sudarshan's to give weight to all indicators. Iyengar and Sudarshan (1982) developed a method to work out a composite index from multivariate data and it was used to rank the districts in terms of their economic performance. This methodology is statistically sound and equally suited for the development of composite index of vulnerability to climate change. In Iyengar and Sudarshan's method, the weights are assumed to vary inversely as the variance over the regions in the respective indicators of vulnerability.

The choice of the weights in this manner would ensure that large variation in any one of the indicators would not unduly dominate the contribution of the rest of the indicators and distort inter regional comparisons. It is well known that, in statistical comparisons, it is more efficient to compare two or more means after equalizing their variances.

The overall village index, Y_i , also varies from zero (0) to one (1) with 1 indicating maximum vulnerability and 0 indicating no vulnerability at all. the higher the district index, the more the level of vulnerability .

The composite indicator for flood vulnerability factors (exposure, susceptibility and resilience) for the i^th village was obtained as:

where: Yi is the composite indicator of i^th village; Wj is the weight for each indicator lies between 0 and 1; ?Wj= 1; and Yij is the normalised scores of indicators.

To ensure that the indices calculated for each vulnerability factor can be compared, the sum for each factor of exposure, susceptibility and resilience are divided by their respective number of indicators that describe each vulnerability factor. The composite vulnerability index for exposure factor is given as:

Wj is the weight a single indicator, ei is exposure indicators; Yij is the normalised value of exposure indicator; n is the number of indicators.

Any flood vulnerability analysis requires information regarding these factors, which can be specified in terms of exposure indicators, sensitivity indicators and resilience indicators. Finally, the vulnerability of a system to flood events can be expressed with the following general equation (Balica, 2007, p 37). This equation is used in the present study to compute Flood Vulnerability Index (FVI).

The composite index values of the three factors of vulnerability and total flood vulnerability index values were integrated in ArcGIS 10.1 software with all relevant input data being available in a digital spatial database (polygon shape file) to produce exposure, susceptibility, resilience and vulnerability maps. The maps were classified and colour coded green-yellow-red, indicating low-moderate-high areas, respectively.

CHAPTER IV: PRESENTATION AND DISCUSSION OF RESULTS

This chapter presents and discusses results of the research based on the primary and secondary data collected. The results presented in this chapter have been arranged in sections which include finding on trend in rainfall and river discharge of the study area, determinants of vulnerability and Flood Vulnerability index and mapping of vulnerable areas results.

4.1. Empirical Finding on Trend and Variability Analysis

4.1.1. Precipitation Time Series Analysis

Statistical properties of the annual and monthly rainfall series were tested and presented in "Annexe 1". The result shows that April, May, June and October represent the smallest Coefficient of Variation (CV): 0.486, 0.388, 0.353, and 0.439, respectively which means that they were the homogenous months in terms of rainfall variations during the period of record. On the other hand, December, January and February show the largest CV with 2.122, 1.586 and 1.06, respectively. The rest of the months present similar rainfall pattern representing similar variation during the study period. The annual maximum rainfall occurred in the year 1999 with the total precipitation of around 1341.5 mm approximately and the minimum rainfall occurred in the year 1977 with the total of around 674 mm.

On running the Mann-Kendall test on precipitation data, the Sen's slope shows an evidence of a positive trend in annual series. The rate of annual rainfall change is about 3.434 mm/year. The result indicates that the null hypothesis was accepted for the annual rainfall trend (p-value= 0.159). Thus, statistically significant positive trend is not found for annual rainfall over the time period.

On plotting the linear trend line for the 40 years rainfall data, the following results in "Figure 7" were obtained.

Figure 6: Linear trend line corresponding to rainfall data (1971-2010:Tabligbo station)

The "figure 7" represents the graph for the twelve (12) months average rainfall for the time period (1971-2010). It shows two peaks in the year one in June (159.647mm) and another one in October (130.06 mm) which reveals the bimodal pattern of rainfall in the study area.

In the Mann-Kendall test, the Sen's slope estimator reveals the trend of the series for 40 years for individual 12 months from January to December which are -0.073, -0.560, -0.069, 0.353, 0.285, 1.134, 0.980, 0.153, -0.480, 1.307, 0.699, and -0.191, respectively. For April, May, June, July, August, October and November, there is an evidence of a rising trend while the result is displaying negative trend in January, February, March, September and December. Thus Sen's slope estimator shows a positive trend for eight months and for other four months it shows a negative one representing almost non-significant condition. The Null hypothesis was accepted for all the twelve (12) months "annexe 2". Therefore, statistically significant trends are not found for precipitation on monthly basis, at 95 % confidence level, even though there are negative and positive trends for the record of period (1971-2010) considered.

Figure 8: Annual rainfall cumulative deviation (1971-2010) (Tabligbo station)

To be able to determine normal, wet and dry years, cumulative deviation from mean of rainfall pattern were computed for the periods of record. "The figure 8" reveals that a cyclic pattern of variations with alternating drier and wetter years is suggested. Two main phases of different lengths were detected (1971-1977 and 1978-2010). The first phase covering eight (8) years shows six negatives anomalies that let to conclude a dry phase. It is followed by a long period from 1978-2010 characterized by variations with alternating drier and wetter years. This result explains rainfall variability over the study area during the period under examination.

4.1.2. Discharge Time Series Analysis

Trend analysis of the downstream part of the Mono River basin has been done also with 40 years of river discharge data from 1971 to 2010. Statistical properties of the annual and monthly flow series were tested and presented in "annexe 3". The results show positive skewness which means the data were normally distributed. According to the results, all the individual months show the largest CV representing similar variation during the study period. The annual average discharge for these 40 years is 114.985 m³/s. During the record period,

the maximum discharge occurred in the year 2001 with the total discharge of 262.408 m³/s approximately and a minimum discharge in the year 1984 with the total of around 19.09 m³/s.

On running the Mann-Kendall test on river discharge data, the Sen's slope shows an evidence of a positive trend in annual series. The rate of annual rainfall change is about 2.462m³/s/year. The result indicates that the null hypothesis was rejected for the annual discharge trend (p-value= 0.002) "annexe 4". Thus, statistically significant positive trend is found for annual river discharge over the time period.

On plotting the linear trend line for the 40 years river discharge data, the following results in "Figure 9 " were obtained.

The "figure 10" shows the monthly discharge distribution of 40 years. It shows one peak in September (333.442 m³/s)

In the non-parametric Mann-Kendall test, monthly trends of river flow for 40 years have been calculated for each month individually together with the Sen's magnitude of slope (Q). The Sen's slope reveals the trend of the series for 40 years for individual 12 months from January to December which are 2.653, 2.13, 2.233, 2.031, 2.289, 0.910, -1.354, -3.391, -4.697, -0.047, 2.295, and 3.433 respectively. While July, August, September and October show an evidence of negative trend, the others months show evidence of positive trend. The Null hypothesis was accepted for July, August and October months and rejected for the others months "annexe 4".

The "figure 11" shows the cumulative deviation from mean that reveals a cyclic pattern of variations with alternating low and high discharge years. Three main phases of different lengths were detected (1971-1978, 1979-1997 and 1998-2010). The first phase shows negative anomalies that let to conclude a low discharge phase. It is followed by a period (1979-1997) characterized by variations with alternating low and high discharge years. The last period (1998-2010) shows a phase of high discharge years. This result explains river discharge variability over the study area during the period of record.

The application of the trend analysis reveals an overall upward trend in annual rainfall and river discharge. It is well known that rainfall is one of the major inputs into runoff processes while river discharge shows a composite response of the whole basin. The upward trend in the two variables may show a causal effect of rainfall on river discharge and an evidence of climate variability. The evidence of positive trend in the river discharge and the rainfall characteristics suggest that the study area may be exposed to either river flood or flash flood.

The result equally shows that even though there is an evidence of upward trend in annual rainfall and river discharge in the study area , the trend in the rainfall series is not significant compared to the one of the river discharge. This may be explained by these studies (Rossi 1996, Klassou 1996, Amoussou and al.2012, ) which show that the Mono River is under the influence of downstream and upstream rainfall and besides, under the effect of the Nangbéto dam put in service since 1988 for hydropower production reasons which confers to the river an artificial character at the downstream part. According to the interview carried out in the scope of this study, the increase of flood hazards in the area is not only due to change in precipitation patterns causing overflow of the river but also to man-made actions such as: the regular opening of Nangbeto dam at the upstream of the basin. Clearly, flood hazard in the study area is a natural phenomenon which was exacerbated by anthropogenic factors.

From the result, river flood peaks may occur in September. The increase of river discharge causing flood hazards in the study area calls up the need to describe past floods magnitudes in order to predict design floods for the targeted area. To this end, calculation of the return period as well as the probability of occurrence of past flood magnitude to estimate the likely values of discharges to expect in the river at various recurrence intervals based on the available historical record was carried out.

4.2. Determinants of Communities' Vulnerability to Floods

4.2.1. Flood Frequency and Magnitude Analysis

The maximum instantaneous flow of 736.70 m³/s was recorded at Athieme during the 1999/ 2000 hydrological year while the lowest flood flow of 69.16m³/s was recorded in 1983/1984 hydrological year. The 40-years mean instantaneous maximum flood flow is 372.34m³/s with a CV of 40.6% and a standard variate of 182.757m³/s "table 5". The coefficient of variation was applied to measure the consistency and the steepness for the frequency curves in the river flow data. The CV value obtained indicated that the distribution of flood flows was not highly variable.

The return period and the probability of occurrence for each observation of the Mono River have been computed, using Weibull's formula, for the period 1971- 2010 which generally starts to peak in the month of July with the maximum in the month of September. according to flood hydrology data of Athieme station.

The Mono River discharge at the downstream reveals that the study area has been affected 22 times by low intensity flood with return periods of 1 or 2 years with high probability of occurrence. The low intensity is ranged between (69.16-372.33m³/s). The study area experienced nine times moderate intensity flood with return period of 2, 3 or 4 years with probability of occurrence less than 50% and magnitude between (372.34-549.64 m³/s). The study area was challenged with high flood event nine times ranging between (549.65-736.70 m³/s) "figure 18" and "table 6".

The latest more severe flood for the downstream part of the Mono basin is the one that occurred in 2010. Its return period is 5 years and the probability of occurrence of the 2010 flood (as a same magnitude) would be once in five years (Probability=0.22) "annexe 5". During the period, the recurrence interval of high flood based on the 2010 flood magnitude has ranged from 5 to 41 years. There are eight recurrence intervals covering a total period of 40 years between the first and the last occurrence of high flood events.

The water level as well as flood duration were different for the targeted villages "table 7". The result reveals that the higher the flood water level, the higher the flood duration.

It was noticed that villages such as Kpodji, Tchakponou and Logokpo have the highest flood level and highest flood duration although they are the most distant of the Mono River. This can be explained by the fact that these villages are surrounded by Mono River's sub-branches. They are not directly inundated by the Mono river itself but rather by the Mono's sub-branches. Then, when water comes from all the sub-branches, the total areas is highly inundated. In addition, the area is made up by heavy soil which can decrease the flood water infiltration capacity and increase the duration of flood water in the area.What about the soil types in the affected villages?????

Table 6: Depths of flood water (2010) as revealed by marks on building walls and average flood duration (2010) from household

In addition, all the eight surveyed villages lie in the low altitude level of the Yoto district comprised between (12-58m) which makes them to be highly exposed to flood hazards "Map 2".

The study focuses on two main elements at risks: households and their agriculture croplands. The total population from the surveyed sample is about 2124 composed of children, young, elderly, and adult. 42.74% of the total population are children, 4.61% are elderly and 17.14% are women. As it is shown in various studies and confirmed by the majority of the respondents 72.36%, women, children and elderly are considered as the most exposed to flood due to their fragility. The little high proportion of children under 15 years in the study area may increase the communities' vulnerability to flood disasters.

Based on information collected through the simple random survey, it was found that those surveyed households were having agriculture croplands of about 506 ha of which the majority is very closer to the river body (less or equal to 200m). The crop production activity in the study area depends on rainfall and practices which is still traditional with rudimentary tools. Farmers are faced with problems of storage and preservation of harvested crops. Thus, when flood came, before and even after harvesting period, the majority of crops are destroyed.

Photo 1: Oil palm farm under water since the 2010 flood, Photograph taken during field work

4.2.4. Assessment of communities' vulnerability : Human-environmental conditions

This section consists in identifying the different characteristics of a Human-Environment system which would make the surveyed villages vulnerable to floods.

4.2.4.1. Socio-Demographic Characteristics of Households

The total sample comprised of 221 respondents who were household heads. The social demographic features of households are as shown in (Table 7). The majority of the respondents were male (86.9%) while females constituted 12.2%. Most of the respondents were aged between 40 and 59 years (47.1%). The level of education was assessed because it was an important factor in understanding household vulnerability to disasters. The majority of the respondents (39.8%) attained only primary education as their highest level with 38.9% having no schooling at all. 14.5% had secondary level education and only 0.09% had attained tertiary level of education. The majority of the households heads are married (90.5%) with 7.2% of widowed. Most of households that is 62.90% had more than 10 members.

4.2.4.2 Location of settlement and type of construction

The location of settlements in the targeted area led to increase household vulnerability to flood disasters. As observed during the field survey, most of the settlements are located near the river as the river is the main source of water in the study area. The proximity to the river body facilitates the access to water for their various activities.

The field survey carried in the scope of this study shows that the vulnerability of the building structure depends on the building materials. A total sample comprised of 221 respondents were household heads .The majority of the households interviewed (74.7%) lived in building made up of clay walls with thatched roof. Of the households, 13.1% lived in clay walls with iron/tiles sheet roof's building, while 4.5% and 5.9% of the respondents lived in brick walls with iron/tiles sheet roof and hurdle or banco walls with thatched roof buildings, respectively. The majority of households therefore lived in the type of houses that make them susceptible to floods.

Photo 3: House made in Banco and straw Photo 4: House made by clay wall with destroyed by the 2010 flood in Tokpli county thatched roof, Photograph taken during Source: PDNA, 2010 field work

4.2.4.3. Livelihood patterns of respondents

The socio-economic status of this community constitutes another source of vulnerability. The social economic status of households is an important factor in assessing their vulnerabilities to disasters (Wisner, et al. 2004:12). Almost all people of the community in the study area depend on agriculture. The interview reveals that the main source of income for the assessed households are agriculture activities (crop production) 90% and 7.2 % of respondents who do not have agriculture as main activities have it as secondary activities. 65.05% of the total respondents depend solely on agriculture activities. Most of the surveyed households have a limited livelihood options, for most of them indicate having no secondary livelihood sources. Those who have a secondary activity mention second livelihood sources such as trading, breeding, fishing, hunting, palm oil production.

The marital status of household head played an important role in determining the livelihood strategy. Those who are married have a diversity of livelihoods as opposed to singles, and widowed household heads "figure ".

4.2.4.4. Awareness and impacts of flood

The awareness to flood hazards may raise the attention of population on how to manage flood risks. According to the key informants (NGOs and Public institutions) that were interviewed in the field, the occurrence of flood in the area does not take the residents by surprise because even though it is a natural phenomenon, the residents know when the river may overflow and they received warning on radio and through the sign posts implanted to indicate the water level. Equally, they are aware of flood-related risks in the area. The analysis shows that the majority, 94.1% is aware of the flood risks, while only 4.5% said they are not aware. A majority of the households (84.6%) said they had warnings about the threat of flood and how to handle its effects, while 13.6% said they did not have any prior warning about the impending threat of flood. Looking at the 2010 flood which was exceptional in the area according to the population, 86.9% were aware of the 2010 flood occurrence while only 9.5% said they were not aware.

Regarding the different ways for passing information on flood threat, 77.8% of the total respondents said they received information by radio, while 2.7% and 0.09% said they received warnings about floods by traditional ways and through volunteers of the Red Cross, respectively. Furthermore, despite the threats caused by flood in those communities, they still live in the same area for several reasons. The majority of the respondents (47.5%) reported they have nowhere to go, 22.2%, 12.7%, and 8.6% reported respectively that it is hard to find farmland elsewhere, that the lands are for their ancestors, and the lands of the area are fertile.

In terms of the impact of 2010 flood, 96.4% of the respondents had been affected in various ways (loss of crops, destruction of houses, loss of animals, loss of important assets etc.) and 2.7% said they had not been affected.

4.2.4.5 Environmental conditions

The extent of environmental degradation has also an important role to play in exacerbating community vulnerability. The field survey revealed that the study area is subjected to the fragmentation of the river bank digging and widening the river channel, the soil degradation "photo 6 and 7" and removing of vegetation along the river bank which leaves the soil exposed and increase surface runoff and then flood extent. AGO et al (2005) observed that, the soil degradation, the deforestation of the floodplain, the increase in the number of human settlements in the river boundary increases the vulnerability of population to flood hazard.

Photo 5: Dehydrate soil in Batoe village, Photo 6: Mono River bank fragmentation Photograph taken during field work in Mawussou village, Photograph taken during field work

4.2.4.6. Household coping mechanisms

The occurrence of flood hazard affects people in different ways. The affected people are forced to employ coping strategies. Various preparedness and coping mechanism are adopted by the surveyed households. Advanced preparation, training and planning will facilitate the evacuation processes (Ekotu, 2012:101). The majority of the respondents of the total sample (58.4%), reported they employed preparedness measures to reduce flood impact and (48%) have used reactive measures to cope during. They employed pro-active measures such as the use of short cycle seeds (6.3%), early harvest (13.1%), preparation of evacuation place (9%), protection of food supply (5%), putting of important assets in safe place (12.2%). Respondents reported to employ reactive measures, such as tire track (1.8%), sandbag track (11.8%), backfilling with sand (22.6%), backfilling with palm kernel shells (5.9%), and pumping (0.09%). The use of those pro-active measures as well as the reactive measures by the population were confirmed by Red cross institution of the Yoto district that is highly involved in disaster risk reduction in the targeted areas.

During and after flood disasters, most of households do not have any assistance from government and there is no policy to help the communities out. Each household, then, relies on itself and on their families and relatives in non-flooded area to cope with flood.

Despite the existing anticipatory information in the study area there is no adequate strategy to increase population's resilience to flood disaster. This calls to think on alternative solution to increase adaptive capacities of population. According to the key informant interview, population in vulnerable areas should firstly change their mentality in order to accept their relocation in a safe areas, they should build strong houses and adopt an appropriate agricultural practices adapted to their area and the use of short cycle seeds to increase their resilience.

4.2.4.7. Anticipative measures of flood occurrence

The anticipative measures play a great role in the reduction of flood impacts. The field survey reveals some anticipative measures used by the population to predict the occurrence of flood. The population combines sign post markers to control the level of water with local indicators of flooding (massive presence of ants coming out from ground, and snail climbing trees, transport of mud, presence of hippopotamus and some type of bird so called "Tolem" in local language) to anticipate the occurrence of flood. The majority of the respondents of the total sample (82.8%) reported they are aware of anticipative measures such as local indicators (78.7%) and sign post marking possible flooding levels (4.5%)

4.2.4.8. Training on flood hazard management

In terms of training on flood hazard management, 82.8% of the total respondents said they have received some training on flood hazard management while, 14% have never received any training . Of those who have some training, 1.8% indicated they received information on how to control river water level, 20% reported they received information on hygiene and water purification measures, 12.21% reported they received information on local indicators and a small number 3.1% said they were informed on evacuation areas.

4.2.4.9. Household recovery time and positive effects of flood on household

The majority of respondents (36.2%) said it takes long time for them to recover from flood disasters, while 24.9% reported they recover quick from flood disasters. 75.11% of respondents of the total sample said flood had no positive effects on their household while 24.9% reported flood had some positive effects on them. Of all those who reported having experienced positive effect of flood, 11.31% said their farmlands become more fertile after flood, 7.2 % said they experienced increase of crop yield after flood, 4.07% said they were able to practice garden activity after flood because the soil become suitable for such activity and a small number 2.3% said they experienced increase in fish population, which is good for fishing activity.

4.2.5 Household adaptation options

4.2.5.1 Household's perception of Government and NGOs role in flood management

The perception of households on the role of Government and NGOs in flood management was assessed, based on the three phases of disaster management cycle. The majority (15.83%) of the total respondents interviewed indicated that government and non government organisations should increase awareness and educate people about the causes, risks and warning signs of floods; 11.31% said government and NGOs should build dams and reservoirs or dikes and levees, and health centres. Others reported that government and non government organisations should build retaining ponds, flood channels, safe heaven places, construct roads, distribute boats; install more sign posts marking possible flooding levels in the community, plant trees along the river bank and control of Nangbeto dam opening, as mitigation and preparedness measures to control flood disasters. As response measures during flood, the majority (54.30%) of respondents reported that government and NGOs should assist the affected communities with food and non-food -items, however 36.20% and 9.50% said the government and NGOs should take appropriate measures to evacuate affected people in the safe havens and provide health assistance. 52.03% respondents of the total sample reported government and NGOs should provide seeds, fertilizers and animals; 13.57% suggested assisting people in returning to their home and distribution of building materials while 11.31% said government and NGOS should provide advice and training to flood victims and 9.50% proposed to assist affected people with financial support for their recovery after floods.

4.2.5.2. Household `s perception of communities role in flood management

A part from the role of government and NGOs in flood management, the interviewed households recognized the role of communities in managing flood disasters. 33.94% reported they should adopt early harvest option in order to reduce impact of flood on their livelihood; 22.63% said the importance of planting trees in reducing flood extent, while 13.57% reported they should install collective food storage in order to assist affected people with food items. Others suggested diversification of economic activities (11.31%), group farming (7.24%); building of strong house (3.16%), avoiding cultivation close to the river (2.26%), flood management committee (2.26%), collective saving (3.62%).

To sum up, besides the extreme variability in terms of flood magnitude and frequency in the Mono River in the study area, which may be due to the increasing in the precipitation and the river discharge patterns, the proximity of the villages and the closeness of households' farmlands to the river body, the type of construction and the position of settlements, the structure of the populations (high number of children; high household size), low level education of household, the lack of the diversification of livelihood strategies, the lack of adequate flood warning system and lack of willingness and ability to take responsive actions coupled with inadequate emergency services during and after flood, may increase the communities' vulnerability to flood disasters.

The low level education coupled with the limited livelihood strategies and the low incomes have resulted in poor agricultural practices. In addition, since the crop production is the main source of income and food added to the high number household member, increased exposure to floods will exacerbate the population vulnerabilities to flood hazards by compromising their food security. This situation proves that the research hypothesis is verified.

4.3 Computation of Flood Vulnerability Index

The Flood Vulnerability Index (FVI), in the present study, aimed to identify the most vulnerable village related to flood events in the three selected counties in the downstream area of the Mono River basin in the Yoto district.

4.3.1 Identifying key indicators of developed FVI

Thirty (30) indicators were used in the present study. Those indicators were categorised under the three factors of vulnerability and were included in the FVI computation.

4.3.2. Normalised Scores and Weight Values of Indicators

A system at risk is more vulnerable when it is more exposed to a hazard. However, it will be less vulnerable the more resilient it is. From the vulnerability equation, high exposure and high susceptibility lead to increases in vulnerability. On the other hand, high resilience levels decreases vulnerability. To this end, the normalization method takes into account the functional relationship between the variable and vulnerability in order to avoid misleading issue in the construction of the indices. The normalised values of each indicator is given in annexe (6). Iyengar and Sudarshan (1982) method were used to calculate weight of each indicator. The calculated weights for each of the flood vulnerability indicators are given in "annexe "7.

4.3.4. Composite vulnerability index of vulnerability factors

4.3.4.1. Exposure factor

Exposure considers the indicators which explain how social entities such as households or economic activities like agriculture, etc., are exposed to flood events. Ten (10) indicators are used to explain the determinant of communities' vulnerability to flood disaster under exposure factor. Two main determinants are found: flood characteristics composed of flood frequency, magnitude, depth and duration as well as elements at risks composed of households and their farmland. Flood characteristics are quite the same for all the surveyed village but the difference is related to the elements at risk in each village. The composite vulnerability index of this factor is calculated for each village. By considering the composite index of exposure factor, the most exposed villages are Djrekpon and Kpodji with high indices ranging from 0.0451 to 0.0651 "Map 3". The most exposed villages have the highest scores for most of the considered indicators under the elements at risk compared to the other villages : high population and farmland in flooded area, high percentage of women and children and elderly in flooded area "annexe 6".

4.3.4.2 Susceptibility Factor

Susceptibility considers the indicators which evaluate the sensitivity of an element at risk before and during a flood event. Eleven (11) indicators are also equally used to explain the determinants of communities' vulnerability to flood disasters under susceptibility factors. The composite index of susceptibility factor for each village is computed. By considering the composite index of susceptibility factor, the most susceptible villages are Batoe and Atikpatafo with high indices ranging from 0.0058 to0.0065 "Map 4". The most susceptible villages have the highest scores for most of the considered indicators :high female headed household, low education level, limited livelihood strategies, high household size, low coping capability, low access to emergency service, low preparedness capability "annexe 6".

4.3.4.3 Resilience Factor

Resilience factor considers indicators which clarify the ability of a Human-Environment system to persist if exposed to flood by recovering during and after the event. Eight (08) indicators have been used to explain the determinants of communities' vulnerability to flood disasters under resilience factor. The composite index of susceptibility factor for each village is computed. By considering the composite index of resilience factor, the least resilient villages are Djrekpon, Tchakponou and Kpodji with indices ranging from (0.0059- 0.0069 ) "Map 5". The least resilient villages have the lowest scores for most of the considered indicators :Low knowledge on warning system, low evacuation capability, low recovery capacity "annexe 6".

The values of the indicators were used in the following general equation of vulnerability to determine the overall flood vulnerability index. Among the eight villages surveyed, Djrekpon and Kpodji are found to experience relatively high vulnerability with indices ranging from (0.0048-0.0225) , Batoe, Atikpatafo, Logokpo and Tchakponou are found to be moderately vulnerable with indices ranging from (0.0256-0.461) and Mawussou and Tofacope are estimated to suffer relatively low level of vulnerability to flood disaster with indices ranging from (0.462-0.668) "Map 6".

This section compares the vulnerability of the selected villages by computing the composite vulnerability index of the three factors of vulnerability, using indicators identified under the different determinants of Human-Environment system and the overall flood vulnerability index. The overall FVI has shown two most vulnerable villages: Djrekpon and Kpodji. It is found that Djrekpon and Kpodji villages were highly exposed, Djrekpon was highly susceptible and Kpodji was moderately susceptible while the two villages were found to be least resilient. Some justification can be found in these results by looking at the number of households affected by floods during the last ten years, the high percentage of household heads with no education level, the lack of livelihood strategies option of those households, the highly susceptible building materials, the lack of adequate coping capacity and recovery capacity from floods. However, the low values found in the different results of the three factors of flood vulnerability as well as in the overall vulnerability index for some villages can be misinterpreted as not being vulnerable to floods. This may not be the case since all determinants of human-environment or socio-ecological system can be damaged under certain conditions.

CHAPTER V: CONCLUSION AND POLICY RECOMMENDATION

This chapter presents the conclusions, and policy recommendation, of the current study and areas for further study based on the discussion of results presented in Chapters 4. This study was conducted in the district of Yoto, South-eastern Togo, in three counties: Sedome, Esse-godjin and Tokpli. The study covered the villages of Mawussou, Djrekpon, Batoe, Tofacope, Atikpatafo, Logokpo, Tchakponou-kondji and Kpodji. The main objective of the study was to analyse the long term trend of rainfall and river discharge series for the study area, to identify the determinant of communities' vulnerability to flood disasters and to compute flood vulnerability index in order to easily understand and compare the vulnerability of the different villages.

The data analysis results suggested the evidence of change in precipitation and in the river discharge which may be the major contributing factor of vulnerability to flood hazards. The increase of river discharge over the years calls up the need to describe flood hazards by computing the return period of the different intensities recorded over the record period of 1971-2010 and especially for 2010 flood which was exceptional in the targeted area and in the whole country. Flood disasters in the target area is not only due to the increase of precipitation and river discharge, but also to the interaction between human and the environment; the vulnerability analysis reveals that the communities' vulnerability to flood in the targeted area may mostly be caused by lack of coping mechanisms, the insufficient emergency response or service during and after floods from public and private institutions, the closeness of households' farmlands and settlements to the river body, lack of diversification of livelihood strategies, poor building materials, low education level etc...

The computation of Flood Vulnerability Index (FVI) suggests that communities' vulnerability to flood can be reflected by the three factors (exposure, susceptibility and resilience). The FVI offers easy comprehensive results, with the use of a composite values to characterise high, moderate and low vulnerability communities. It is found that out of the eight surveyed village, two were highly vulnerable (Djrekpon and Kpodji); Four were moderately vulnerable (Batoe, Atikpatafo, Logokpo and Tchakponou-kondji) while two were least vulnerable (Mawusssou and Tofacope).

To alleviate the vulnerability of those surveyed villages, the study recommends some prior actions that can be taken to prevent homes and communities from the damage caused by flooding. Based on the current research, it is recommended to build awareness, preparedness and knowledge of communities about the importance of flood risk management and particularly, what actions should be taken as preparation and coping mechanisms. It recommends to educate the populations in risk reduction by putting in a place local flood management committee in order to enable preparation of community action plans that explain what to do in flood case. The study recommends putting in place a good early warning system, that local and regional weather information can use to sensitize about flooding. Regarding advance warning, it is suggested to associate the public and private media in the diffusion of the alerts related to the seasonal forecasting of flood hazards, to install sign posts marking possible flooding levels in the community and educate people about warning signs of floods in order to limit the impacts of flood in the communities prone to flood disasters.

This study is an attempt to assess communities' vulnerability to flood in the Yoto district. For timing and limited funding purposes, it was difficult to expand the area of assessment beyond the eight selected villages. The results may not be totally extrapolated for the whole district as each community has its own conditions. Since the methodology is based on indicators, its main weakness is the accuracy of data to compute the equation. The results of Flood vulnerability index in this study depend in majority on information from communities. Some information was derived from sources that can be considered as non-reliable, for example the village distance of contact with a river, which was taken from Google Earth, by computing the distance using the ruler tool in the software. For the results to be valid, all data must be derived from other reliable sources.

Based on the present study, there is clearly a need for more research into communities resilience, and adaptive options to the flood hazard, particularly communities traditional knowledge and perceptions, experiences and historical processes used to mitigate floods. Also, the aspect of communities' behavioural response toward flood awareness especially household private decisions in flood risk management should be investigated. It is also recommended to conduct local risk assessment for building and agriculture croplands using the engaging technique of flood-depth-analysis to work out the structural and agricultural vulnerability that different levels of flood water could bring.

REFERENCES

1. Aall, C., and I.T. Norland .2005. Indicators for Local-Scale climate vulnerability assessments, Oslo: University of Oslo.

2. Adger, W.N. .1999. Social Vulnerability to Climate Change and Extremes in Coastal Vietnam. World Development 27: 249-269.

3. Adger, W. N., N. Brooks, G. Bentham, M. Agnew, and S. Eriksen .2004. New Indicators of Vulnerability and Adaptive Capacity, Tyndall Centre for Climate Change Research: Technical Report 7, Norwich.

4. Adriaanse, A..1994. In Search of Balance: A Conceptual Framework for Sustainable Development Indicators, London: Network seminar on sustainable

5. Ago E. E., F. Petit and P. Ozer .2005. Analyse des inondations en aval du barrage de Nangbéto sur le fleuve Mono (Togo-Bénin). Géo-Eco-Trop 29 :1-14

6. Atkins, J., S. Mazzi, and C. Ramlogan .1998. A Study on the Vulnerability of Developing and Island States: A Composite Index, Commonwealth Secretariat: UK

7. Amoussou, E., P. Camberlin and G. Mahé .2012. Impact de la variabilité climatique et du barrage Nangbéto sur l'hydrologie du système Mono-Couffo (Afrique de l'Ouest), Hydrological Sciences Journal: 805-817.

8. Balica, S. F. . 2007. Development and Application f Flood Vulnerability Indices for Various Spatial Scales, Unpublished Masters thesis: UNESCO-IHE Institute for Water Education, the Netherlands.

9. Balica, S. F., N. Douben, N. G. Wright .2009. Flood vulnerability indices at varying spatial scales. Water Science and Technology 60(10): 2571-2580.

10. Balica, S. F., N. G. Wright, F.Vander Meulen .2012. A flood vulnerability index for coastal cities and its use in assessing climate change impacts. Natural Hazards 64(1): 73-105.

11. Birkmann, J.. 2006. Measuring vulnerability to natural hazards: Towards disaster resilient societies, UNU press. http://i.unu.eduf (consulted on 20 march 2014).

12. Birkmann, J. .2008. Assessing Vulnerability Before, During and After a Natural Disaster in Fragile Regions. Case Study of the 2004 Indian Ocean Tsunami in Sri Lanka and Indonesia, Research Paper No. 2008/50, United Nations University: World Institute for Development and Economics Research (UNU-WIDER).

13. Birkmann J. .2013. Measuring Vulnerability to Natural Hazards: Towards Disaster Resilient Societies, United Nation University (second edition)

14. Blaikie, P., T. Cannon, I. David and B.Wisner .1994. At Risk: Natural Hazards, People' s Vulnerability, and Disasters, Routledge : London.

15. Bowen, R. E., and C. Riley .2003. Socio-economic indicators and integrated coastal management. Ocean and Coastal Management: 199-312.

16. Buckle, P. .1998. Re-defining Community and Vulnerability in the context of Emergency Management. Australian Journal of Emergency Management 13(1):21-26.

17. Burn, D.H. and M.A. Hag Elnur. 2002. Detection of Hydrologic Trends and Variability Journal of Hydrology 255 (2002):107 -122.

18. Cannon, T. .1990. Vulnerability Analysis and the Explanation of Natural Disasters. In A. Varley, Disasters, Development and Environment, New York: John Wiley and Sons Ltd.

19. Cardona, O. .2003. A need for Rethinking the concept of vulnerability an risk from a holistic perspective: A Necessary Review and Criticism for effective risk management, http://www.deservredando.org/public/articulos/2003/nrcvrfhp/nrcvrfhp_ago-04-2003.pdf (consulted on 20 August 2013)

20. Carter, W. N. .1991. Disaster Management, Disaster Managers' Handbook, Manila, Philippines: Publication of the Asian Development Bank.

21. Chamber, R. .1983. Rural Development: Putting the Last First, Essex: Longman.

22. Chris, Easter. 2000. The Common Wealth Vulnerability Index, Ministerial Conference on Environment and Development in Asia and the Pacific, Kitakyushu: Japan.

23. Clark, G. E., S. Moser, S. Ratick, K. Dow, W. B. Meyer, S. Emani, W. Jin, . J.X., Kasperson, R. E. Kasperson, and H. E. Schwarz .1998. Assessing the vulnerability of coastal communities to extreme storms: the case of Revere, MA., USA , Mitigation and Adaptation Strategies for Global Change 3( 1), 59-82.

24. Connor, R.F., and K. Hiroki .2005. Development of a method for assessing flood vulnerability, Water Science and Technology 51(5):61-67

25. Cutter, S.L., B.J. Boruff, and W.L.Shirley.2003. Social vulnerability to environmental hazards. Social science quarterly 84(22):242-261 http://www.colorado.edu (Consulted on 05 August 2014).

26. De Toffol , S., A.N. Laghari, and W. Rauch .2008. Are extreme rainfall intensities more frequent? Analysis of trends in rainfall patterns relevant to urban drainage systems, 11th International Conference on Urban Drainage, Edinburgh, Scotland: UK.

27. Damm, Marion. 2010. Mapping Social-Ecological Vulnerability to Flooding. A sub-national approach for Germany, Graduate Research series, PhD dissertation (Vol3), Bonn, Germany: United Nation University-IEHS.

28. Downing, T. .2004. What Have we Learned Regarding a Vulnerability Sciencein Support of Adaptation to Climate Change? Recommendations for an Adaptation Science Agenda and a Collection of Papers Presented at a Side Event of the 10th Session of the Conference of the Parties to the United Nations Framework Convention on Climate Change: Buenos Aires. www.aiaccproject.org (Consulted on 20 March 2014)

29. Dolan, A.H., and I.J. Walker .2003. Understanding Vulnerability of Coastal Communities to Climate Change Related Risks Journal of Coastal Research SI 39: 0749-0208.

30. Douglas, E. M., R. M. Vogel, and C. N. Knoll .2000. Trends in flood and low flows in the United States: impact of spatial correlation. Journal of. Hydrology: 90-105

31. Ekotu John Juventine. 2012. Landslide Hazards: Household Vulnerability, Resilience and Coping in Bududa District, Eastern Uganda, Unpublished Master's thesis, University of the Free State

32. EM-DAT: The OFDA/CRED International Disaster Database, Université catholique de Louvain: Brussels, Belgium

33. FAO .2008. Disaster risk management systems analysis, A guide book, Rome: Italy. (consulted on 16 February 2014).

34. Fekete, A. 2009.Validation of a social vulnerability index in context to river-floods in Germany. Natural Hazards Earth System. Sciences: 393-403.

35. Fekete, A. .2010. Assessment of Social Vulnerability for River Floods in Germany, Graduate Research Series vol. 4, UNU-EHS: Bonn

36. Füssel, H. M. , and R.J.T. Klein .2006. Climate change vulnerability assessments: an evolution of conceptual thinking. Climatic Change 75 (3): 301-329.

37. Guha-Sapir, D., D. Hargitt, P. Hoyois .2004. Thirty Years of Natural Disaster 1974-2003: The Numbers. Belgium: Presses Universitaires de Louvain. http://www.cred.be. ( Consulted on 15 October 2013).

38. Gallopín, G.1997. Indicators and Their Use: Information for Decision-Making. In: Moldan, B.; Billharz, S. ed: Sustainability Indicators: Report of the project on indicators of sustainable development. New York:John Wiley.

39. Haki, Z., Z. Akyuerek, and S. Duezguen .2004. Assessment of social vulnerability using geographic information systems: Pendik, Istanbul case study, In 7th AGILE Conference on Geographic Information Science 413-423, Heraklion, Greece: Middle East Technical University of Ankara, Turkey.

40. IPCC .2007. Climate Change 2007: Impacts, adaptation and vulnerability. Contribution of Working Group II to the fourth assessment report of the Intergovernmental Panel on Climate Change, Cambridge: Cambridge University Press.

41. IPCC.2012a. Managing the risks of extreme events and disasters to advance climate change adaptation

42. Iyengar, N.S. and P. Sudarshan. 1982. A Method of Classifying Regions from Multivariate Data , Special Article: 2047-2052.

43. Jonkman, S. N. .2005. Global Perspectives on Loss of Human Life Caused by Floods. Natural Hazards 34(2):151-175

44. Kasperson, J.X., R.E. Kasperson, B.L. Turner, W. Hsieh, and A. Schiller .2000. Vulnerability to Global Environmental Change, The Human Dimensions of Global Environmental Change, Cambridge: MIT Press.

45. Kendall, M.G. .1975. Rank Correlation Methods, 4th edition, Charles Griffin: London U.K.

46. Katharine , Vincent .2004. Creating an Index of Social Vulnerability to Climate Change for Africa, Tyndall: Centre for Climate change Research Working

47. Klassou K. S. .1996. Evolution Climato-hydrologique récente et ses conséquences sur l'environnement: l'exemple du bassin versant du fleuve Mono (Togo-Bénin),Phd Dissertation (Géog. Phys.), Bordeaux III:Université Michel de Montaigne

48. Maiti, Sreyasi. 2007. Defining a Flood Risk Assessment Procedure Using Community Based Approach with Integration of Remote Sensing and GIS- Based on the 2003 Orissa Flood, Unpublished Master's thesis, India Institute of Remote Sensing.

49. Mann, H.B .1945. Non-parametric test against trend. Econometrica 13: 245-259

50. Martens, T.; Ramm, K. (2007): Newsletter 2: http://innig.rtens.net/modules/Downloads/hjiu3443/INNIG_TP3_Newsletter_2.pdf, accessed: 25 April 2014.

51. McLanahan, S. .1983. Family Structure and Stress: A Longitudinal Comparison of Two Parent and Female-Headed Families. Journal of Marriage and Family 45(2): 347-357.

52. MERF, World Bank and UNDP.2010. Evaluation Des Dommages, Pertes et Besoins de Reconstruction Post Catastrophes des Inondations de 2010 au Togo

53. MERF and UNDP. 2013. Strategie Nationale de Reduction des Risques de Catastrophes Naturelles 2013-2017.

54. Moldan, B., A. L. Dahl .2007. Challenges To Sustainability Indicators in Hák. Sustainability Indicators: A Scientific Assessment ed, Moldan, B and Dahl, A. L. Washington DC: Island Press.

55. Moss R.H., A.L. Brenkert and E.L. Malone .2001. Vulnerability to Climate Change: A Quantitative Approach, Department of Energy: U.S.

56. Muller, A., J. Reiter, and U. Weiland .2011. Assessment of urban vulnerability towards floods using an indicator-based approach - a case study for Santiago de Chile Natural. Hazards Earth System. Sciences 11: 2107-2123.

57. Pistrika Aimilia and George Tsakiris.2007. Flood Risk Assessment: A Methodological Framework Centre for the Assessment of Natural Hazards & Proactive Planning, National, Greece :Technical University of Athens.

58. Rossi G. .1996. L'impact des barrages de la vallée du Mono (Togo-Benin). La gestion de l'incertitude. Géomorphologie:relief, processus, environnement 2(2): 55-68.

59. Sadiq I. K., Y. Hong, J. Wang, K. Koray, J. Yilmaz, J. Gourley, F. Robert, G. Adler, R. Brakenridge, P. Fritz, H. Shahid, and D. Irwin .2011. Satellite Remote Sensing and Hydrologic Modeling for Flood Inundation Mapping in Lake Victoria Basin: Implications for Hydrologic Prediction in Ungauged Basins. Transactions on Geoscience and Remote Sensing 49(1): 85-95

60. Schneiderbauer, S. 2007: "Risk and vulnerability to natural disasters -rom broad view to focused perspective. Theoretical background and applied methods for the identification of the most endangered populations in two case studies at different scales", Phd, Freie Universit· at Berlin,.

61. Sen, P. K. .1968. Estimates of the regression coefficient based on Kendall's tau. J. Am. Stat. Assoc. 63:1379-1389.

62. Seventh Framework Programme. 2011. Review and evaluation of existing vulnerability indicators in order to obtain an appropriate set of indicators for assessing climate related vulnerability, Seventh Framework Programme: project CLUVA.

63. Singh, P., V. Kumar, T. Thomas, and M. Arora. 2008. Changes in rainfall and relative humidity in different river basins in the northwest and central India. Hydrol. Process 22: 2982-2992

64. Snyder, AR. DK. McLaughlin, and J. Findeis .2006. Household Composition and Poverty among Female-Headed Households with Children: Differences by Race and Residence. Rural Sociology 71(4): 597-624.

65. Smit, B. and J. Wandel .2006. Adaptation, capacity and vulnerability. Global Environ Change 16:282-292

66. Steinführer, A.; Kuhlicke, C. (2007): Social vulnerability and the 2002 flood. Country Report Germany (Mulde River), Floodsite Report Number T11-07-08, UFZ Centre for Environmental Research, Leipzig.

67. Thieken, A.; H. Kreibich, M. Müller, B. Merz .2007. Coping with floods: preparedness, response and recovery of flood-affected residents in Germany in 2002. Hydrological Sciences - Journal 52(5):1016-1037.

68. Turner, B.L., R.E. Kasperson, P.A. Matson, J.J. McCathy, R.W. Corell, L. Christensen, B. Wisner, P. Blaikie, T. Cannon, and I. Davis. 2003. At Risk: Natural Hazards, Peoples' vulnerability and Disasters, London: Routledge.

69. UN/ISDR .2004. Living with Risk, a Global Review of Disaster Reduction Initiatives, Geneva: United Nations.

70. UN/ISDR .2005b. Hyogo Framework for Action (HFA), 2005-2015: Building the resilience of nations and communities to disasters, Geneva: United Nations,.

71. Veenstra, Jelmer .2013. Flood vulnerability assessment on a commune level in Vietnam, Unpublished Bachelor thesis, Vietnam: VNU University of Science in Hanoi,

72. Velasquez, G., and R. Tanhueco .2005. Know Risk. United Nations `World Conference on Disaster Reduction', Chapter: Incorporating social issues in disaster risk assessment.

73. Villagran de Leon . 2006. Vulnerability - A conceptual and Methodological review, Germany: UNU - EHS, no 4/2006, Bonn,

74. Wisner, B., P., Blaikie, T. Cannon, and I. Davis .2004. At Risk: Natural Hazards, Peoples' Vulnerability and Disasters, London: Routledge.

75. World Bank .2005. African Development Indicators http://www4.worldbank.org/afr/stats/adi2005/adi05_booklet_rev_061505.pdf, (Consulted on 23 July 2014).

76. Yamane T. .1967. Statistics, an introductory analysis, 2nd ed, New York: Harper and Row.

77. Yue, S., P. Pilon, B. Phinney, G. Cavadias .2002. The influence of autocorrelation on the ability to detect trend in hydrological series Hydrological Processes 16(9): 1807-1829.

78. Yu, Y. S., S. Zou, and D., Whittemore .1993. Non-parametric trend analysis of water quality data of rivers in Kansas. J. Hydrol 150: 61-80.

ANNEXES

Annexe 1 : Statistical summary of Annual and monthly precipitation for Tabligbo

Tabligbo station	Minimum	Maximum	Median	Mean	Stdev	Variance	CV	Skewness
Annual	674	1341.5	1005.1	1016.258	179.610	32259.59	0.177	-0.165
Jan	0	50	1.45	7.967	12.64	159.77	1.586	1.851
Feb	0	121.5	24.15	31.8	33.716	1136.79	1.06	1.313
Mar	0.1	309.6	93.850	96.262	60.612	3673.82	0.630	1.050
Apr	24.3	264.3	107.3	118.378	57.584	3315.97	0.486	0.641
May	45.3	325.5	142.75	152.712	59.196	3504.16	0.388	0.514
Jun	44.8	288.9	165.95	159.647	56.391	3179.93	0.353	0.015
Jul	4.8	209.7	79.75	89.19	52.361	2741.71	0.587	0.526
Aug	2.2	213.6	46.55	53.655	44.726	2000.44	0.834	1.562
Sept	0.2	291.3	109.500	120.9	69.310	57.122	35.206	33.334
Oct	28	313.3	135.55	130.06	57.122	3262.92	0.439	0.701
Nov	0	161.4	29.950	39.980	35.206	1239.45	0.881	1.347
Dec	0	194.1	3.5	15.705	33.334	111.123	2.122	4.263

Annexe 2: Mann-Kendall test results of annual, monthly and seasonal precipitation

	Mann Kendall Statistic (S)	p-value (two tailed test)	alpha	Sen's slope estimate		Test Interpretation
Jan	10	0.153	0.05	-0.073	Accept H0
Feb	-123	0.291	0.05	-0.560	Accept H0
Mar	- 31	0.651	0.05	-0.069	Accept H0
Apr	39	1.000	0.05	0.353	Accept H0
May	53	1.000	0.05	0.285	Accept H0
Jun	114	0.291	0.05	1.134	Accept H0
Jul	115	1.000	0.05	0.980	Accept H0
Aug	57	0.651	0.05	0.153	Accept H0
sept	-20	0.451	0.05	-0.480	Accept H0
Oct	122	0.175	0.05	1.307	Accept H0
Nov	119	0.760	0.05	0.699	Accept H0
Dec	-33	0.532	0.05	-0.191	Accept H0
Annual	122	0.159	0.05	3.434	Accept H0
High rainy season	-16.	0.861	0.05	-0.063	Accept H0
High dry season	119	0.169	0.05	0.703	Accept H0
Small dry season	61	0.484	0.05	0.349	Accept H0
Small rainy season	52	0.552	0.05	0.468	Accept H0

Annexe 3 : Statistical summary of Annual and monthly Flow for Athieme

Athieme Station	Minimum	Maximum	Median	Mean	stdev	Variance	CV	skewness
Annual	19.089	262.408	102.970	114.985	58.998	3480.749	0.513	0.725
Mar	0.180	132.952	20.762	40.509	43.749	1913.998	1.080	0.662
Apr	0.380	136.863	21.648	369.902	39.073	1526.688	0.979	0.560
May	0.790	141.390	28.752	47.229	47.705	2275.808	1.010	0.779
Jun	1.928	257.758	56.581	67.245	59.815	3577.842	0.890	1.213
Jul	6.094	359.385	105.551	128.895	91.771	8421.840	0.712	1.055
Aug	21.945	588.352	221.125	230.096	141.150	19923.299	0.613	0.953
Sept	35.190	731.677	302.805	333.442	189.934	36074.800	0.570	0.435
Oct	19.369	736.705	214.869	237.667	155.241	24099.804	0.653	1.153
Nov	1.589	281.567	75.072	91.686	75.570	5710.826	0.882	0.867
Dec	0.966	490.532	31.336	66.880	89.258	7967.078	1.335	2.934
Jan	0.429	248.383	35.203	51.203	56.458	3187.500	1.094	1.269
Feb	0.389	225.486	37.751	44.683	49.629	2463.063	1.111	1.409

Annexe 4: Mann-Kendall results of annual , monthly and seasonal flow for the study area

Months	Mann Kendall test
	Mann Kendall Statistic (Zc)	p-value (two tailed test)	alpha	Sen's slope estimate	Test Interpretation
Mar	62.000	< 0,0001	0.05	2.233	Reject H0
Apr	50.000	< 0,0001	0.05	2.031	Reject H0
May	52.000	< 0,0001	0.05	2.289	Reject H0
June	26.000	0.014	0.05	0.910	Reject H0
Jul	14.000	0.202	0.05	-1.354	Accept H0
Aug	-14.000	0.202	0.05	-3.391	Accept H0
Sept	-26	0.014	0.05	-4.697	Reject H0
Oct	6.000	0.624	0.05	-0.047	Accept H0
Nov	30.000	0.0004	0.05	2.295	Reject H0
Dec	48.000	< 0,0001	0.05	3.433	Reject H0
Jan	50.000	< 0,0001	0.05	2.653	Reject H0
Feb	56.000	< 0,0001	0.05	2.135	Reject H0
Annual	262	0.002	0.05	2.462	Reject H0
High rainy season	456	0.000	0.05	3.416	Reject H0
High dry season	436	0.000	0.05	2.663	Reject H0
Small dry season	80	0.357	0.05	1.392	Accept H0
Small rainy season	54	0.537	0.05	1.125	Accept H0
Feb	56.000	< 0,0001	0.05	2.135	Reject H0
Annual	262	0.002	0.05	2.462	Reject H0
High rainy season	456	0.000	0.05	3.416	Reject H0
High dry season	436	0.000	0.05	2.663	Reject H0
Small dry season	80	0.357	0.05	1.392	Accept H0
Small rainy season	54	0.537	0.05	1.125	Accept H0

Annexe 5: calculation for Return Period of 2010 Flood, Mono River

Water Year	Rank	Ranked discharge	Return period	Exceedence probability
2000	1	736.70	41	0.02
1971	2	731.68	21	0.05
2007	3	709.11	14	0.07
2001	4	665.12	10	0.10
2003	5	615.93	8	0.12
1981	6	591.77	7	0.15
1986	7	588.35	6	0.17
1980	8	552.91	5	0.20
2010	9	549.65	5	0.22
1999	10	526.23	4	0.24
2005	11	471.37	4	0.27
1992	12	463.27	3	0.29
1975	13	444.23	3	0.32
1988	14	440.94	3	0.34
1990	15	436.73	3	0.37
1972	16	425.99	3	0.39
1989	17	424.81	2	0.41
2008	18	402.60	2	0.44
1982	19	350.72	2	0.46
1976	20	337.38	2	0.49
2002	21	331.09	2	0.51
1974	22	329.98	2	0.54
2006	23	315.35	2	0.56
1978	24	289.35	2	0.59
1977	25	257.10	2	0.61
2004	26	256.97	2	0.63
1979	27	252.48	2	0.66
1995	28	251.82	1	0.68
1973	29	247.67	1	0.71
1985	30	240.60	1	0.73
1991	31	227.13	1	0.76
1998	32	225.49	1	0.78
1983	33	212.32	1	0.80
1987	34	200.91	1	0.83
1994	35	196.56	1	0.85
1993	36	166.30	1	0.88
1997	37	139.64	1	0.90
1996	38	137.94	1	0.93
2009	39	80.17	1	0.95
1984	40	69.16	1	0.98

Annexe 6:normalised scores of flood vulnerability indicators of each village

Indicators	Djrekpon	Batoe	Logokpo	Tchakponou	Kpodji	Tofacope	Atikpatafo	Mawussou
E1	0.9007	1.0000	0.8278	0.0000	0.6821	0.4967	0.4967	0.8675
E2	0.6855	0.3918	0.4378	0.0000	0.3877	0.1526	0.4127	1.0000
E3	1.0000	0.2454	0.3152	0.2179	0.3865	0.9689	0.5672	0.0000
E4	0.0946	0.0054	0.0000	1.0000	0.4482	0.0518	0.2750	0.1679
E5	1.0000	0.5000	0.0000	0.0000	0.5000	1.0000	1.0000	0.0000
E6	0.0000	0.7593	0.8704	0.7222	1.0000	0.3148	0.5556	0.1852
E7	1.0000	0.5932	0.2712	0.3729	0.8729	0.0508	0.2542	0.0000
E8	0.2031	0.1080	0.4523	1.0000	0.2915	0.0000	0.0301	0.1834
E9	0.5986	0.9010	1.0000	0.7790	0.8236	0.8605	0.0000	0.0906
E10	0.8581	1.0000	0.6782	0.0000	0.6910	0.4334	0.0800	0.1172
S1	0.5101	0.1376	0.5694	0.1376	1.0000	0.0000	0.7974	0.5926
S2	0.5101	0.1376	0.5694	0.1376	1.0000	0.0000	0.7974	0.5926
S3	0.1601	0.8204	0.8521	0.4392	0.0000	0.6931	0.7949	1.0000
S4	0.3811	0.0000	0.8147	0.4072	1.0000	0.2264	0.2769	0.0983
S5	0.0645	0.5163	0.7737	1.0000	1.0000	0.4627	0.7099	0.0000
S6	0.7921	0.7850	1.0000	0.7850	1.0000	0.8808	0.8711	0.0000
S7	0.6235	0.4833	0.0000	0.0000	0.0000	0.1788	0.0000	1.0000
S8	0.0000	0.6071	0.3928	0.7500	0.5714	0.3928	1.0000	0.5357
S9	0.0000	0.6071	0.3928	0.7500	0.5714	0.3928	1.0000	0.5357
S10	0.8645	1.0000	0.0000	0.5803	0.0939	0.8252	0.8321	0.4392
S11	0.7742	1.0000	0.0624	0.1334	0.0000	0.0740	0.0000	0.3448
R1	0.5913	0.6613	0.0471	0.0233	0.0000	0.4768	0.9418	1.0000
R2	0.8461	1.0000	0.3903	0.5435	0.0000	0.7925	0.5102	0.5205
R3	0.9033	1.0000	0.2031	0.0000	0.6364	1.0000	0.0400	0.3276
R4	0.9033	1.0000	0.2031	0.0000	0.6364	1.0000	0.0400	0.3276
R5	0.5482	1.0000	0.0000	0.4712	0.5672	0.3538	0.6299	0.9361
R6	0.2944	0.1667	0.4141	0.0625	0.2898	1.0000	0.0000	0.2995
R7	1.0000	0.8307	0.1272	0.1538	0.0000	1.0000	0.7969	0.7375
R8	0.9240	0.8821	0.4475	0.4500	0.0000	0.9122	1.0000	1.0000
R9	0.4300	1.0000	0.1040	0.0000	0.0000	0.6173	0.4000	0.9197

Annexe 7: Calculated weights of flood vulnerability indicators

No	Indicators	Weight
1	E1	0.00234
2	E2	0.03905
3	E3	0.0162
4	E4	0.06061
5	E5	0.12353
6	E6	0.00601
7	E7	0.5302
8	E8	0.0002
9	E9	0.01099
10	E10	0.00449
11	S1	0.01011
12	S2	0.01217
13	S3	0.01398
14	S4	0.00812
15	S5	0.02167
16	S6	0.01148
17	S7	0.01462
18	S8	0.00434
19	S9	0.00483
20	S10	0.00596
21	S11	0.01078
22	R1	0.00459
23	R2	0.00404
24	R3	0.0143
25	R4	0.0058
26	R5	0.00698
27	R6	0.00371
28	R7	0.00972
29	R8	0.00846
30	R9	0.02333

Annexe 8: Questionnaire for Household Interview

I am a student at the University of Lome , Togo pursuing a Masters in Climate Change and Human Security. I am collecting data for a research study in Yoto District. The study focuses on assessing population vulnerability to flood in the area. I would like to ask you some questions about your family. The data that you provide is for academic purpose and it will be kept strictly confidential. This is voluntary, you can refuse to answer to some of the questions but I hope you will accept as your views are very important.

1.What are the main climate hazards that have affected your community during the last ten years?

1. heavy rainfall |__|; 2. overflow of Mono river |__|; 3. Other, specify..... ........................|__|

4.Do you think the frequency of occurrence and impacts of flooding have increased during this decade compared to previous decades ?

.................................................................. ..................................................................................

..............................................................................................................................................

............................................................................................................................................

9.How could you appreciate the 2010 flood magnitude compared to the others floods?

.............................................................................................................................................

................................................................................................................................................

..............................................................................................................................................

1. No schooling |__| ; 2. Functional literacy |__|; 3. Primary schooling |__|; 4. Secondary schooling |__|; 5. Tertiary schooling|__|; 6. University |__|

.................................................................................................................................................................

1. brick walls with iron/tiles sheet roof |__|; 2. Mud walls with iron/tiles sheet roof |__|;

1. None |___|; 2. Agriculture |___|; 3. breeding |___| ; 4. fishing |___|; 5. handcraft |___|

6. Palm oil production |___|; 7. Trading |___| other specify ...................................... |___|

1. None |___|; 2. Agriculture |___|; 3. breeding |___| ; 4. fishing |___|; 5. handcraft |___|

6. Palm oil production |___|; 7. Trading |___| other specify .................................. |___|

......................................................................................................................................................

24.Was there any information or announcement or warning about the threat of floods?

29. Do you use any method to reduce the effect of flood disasters to your household when they occur?

..................................................................................................................................................................

....................................................................................................................................................................

33.How do you value the government response during and after the flooding in your area?

1. belated|__| ; 2. immediat|__|; 3. inadequate|__|; 4. adaquate|__|, other specify.................. |__|

..............................................................................................................................................................

1. Local indicators|__|; 2. early warning system |__|; 3 other specify................................................ |__|

1. Public school building |__|; 2. Neighbours or relatives in non flooded area |__|;

3. church building|__| 4. public evacuation site |__|; 5. migrate temporarily to other areas less vulnerable |__|; 6. Other Specify...................................................................................... |__|

45.Does the government or others institutions provide prevention and protection measures?

.......................................................................................................................................................

47.How do you value the ability of anticipation and preparation of the government or other institution to floods?

..................................................................................................................................................................

................................................................................................................................................................

53.Do experience environment recovery after flood ? ( positive effect of flood on the environment)?

...................................................................................................................................................................

55.What do you think the government or NGOs should do as prevention measures to reduce flood impacts on your community?

................................................................................................................................................................

56.What do you think the government or NGOs should do to respond to flood during flood disaster

.................................................................................................................................................................

57.What do you think government or NGOs should to after flood to help you recover?

....................................................................................................................................................................

58.What do you think the government or other institution should do to control flood disaster

....................................................................................................................................................................

59.What do you think your community itself should do to reduce flood impact on your locality?

Annexe 9: Key Informants Interview Guide.

5. Why do you think your community was more affected than any other communities?

12. Do they ever seek your ideas before, during and after helping in disaster situations?

VITA

KISSI Abravi Essenam received his Bachelor of Science degree in Environment Sciences at the Faculty of Sciences from Université de Lomé (Togo) in 2011. After three months of Proficiency in English at University of Cape Coast (Ghana) in 2012, she entered the Climate Change and Human Security program at Université de Lomé October 2012 and received her Master of Science degree in November 2014.

Her research interests include Environmental impact assessment, Disaster Risk Reduction, Risk and Vulnerability Assessment, Climate Change, Human security and Mapping.

Flood vulnerability assessment of donstream area in Mono basin in Yoto district, south-eastern Togo

ABSTRACT

Flood Vulnerability Assessment in Downstream Area of Mono Basin,

South-Eastern Togo: Yoto District. (November, 2014)

RESUME

Dedication

ACKNOWLEDGEMENTS

TABLE OF CONTENTS

LIST OF FIGURES

LIST OF TABLES

LIST OF MAPS

LIST OF PHOTO

CHAPTER I. INTRODUCTION

1.1. Problem Statement

1.2 Research Objectives

1.3. Research Questions

1.4. Research Hypothesis

1.5. Thesis Structure

CHAPTER II: LITERATURE REVIEW

2.1. Hazards, Disasters, and Vulnerability

2.2. Flood Vulnerability Factors

2.3. Methodology for Measurement of Vulnerability to Natural Hazards

2.3.1. Theoretical and Conceptual Frameworks of Vulnerability

2.3.2 Indicators for Measuring Vulnerability

2.4. The Index Approach to Study Vulnerability

2.4.1.Existing Flood Vulnerability Index

CHAPTER III: RESEARCH METHODOLOGY

3.1 The area of study

3.1.1 Localisation

3.1.2. Landscape, soil and vegetation

3.1.3. Climate and Hydrology

3.1.4. Population and Economic Activities

The fertile soils coupled with the abundant rainfall per year ensure ample yields of food crops. The main crops grown in the area include maize, cassava, sugarcane, beans, groundnut, palm trees and some vegetables.

The people in the targeted area also keep animals such as goats, cattle, pigs and chicken. Other activities in the targeted area include trading, fishing, palm oil production etc.

3.2. Methods

3.2.1. Study Population and sampling

3.2.2. Selected Vulnerability Conceptual Frameworks

3.2.3. Flood Vulnerability Indicator Development

3.3. Data collection and Analysis

This part describes data collection processes as well as data analysis methods.

3.3.1. Primary Data Collection

3.3.2. Secondary Data

3.3.3. Data Analysis

3.3.3.1. Trend Analysis of Rainfall And River Discharge to Assess Climate Change.

3.3.3.2. Analysis of the determinants of communities' vulnerability to flood

3.3.3.3 Analysis of human-environmental condition

3.3.3.4 Computation of Flood Vulnerability Index (FVI)

CHAPTER IV: PRESENTATION AND DISCUSSION OF RESULTS

4.1. Empirical Finding on Trend and Variability Analysis

4.1.1. Precipitation Time Series Analysis

4.1.2. Discharge Time Series Analysis

4.2. Determinants of Communities' Vulnerability to Floods

4.2.1. Flood Frequency and Magnitude Analysis

4.2.4. Assessment of communities' vulnerability : Human-environmental conditions

4.2.4.1. Socio-Demographic Characteristics of Households

4.2.4.2 Location of settlement and type of construction

4.2.4.3. Livelihood patterns of respondents

4.2.4.4. Awareness and impacts of flood

4.2.4.5 Environmental conditions

4.2.4.6. Household coping mechanisms

4.2.4.7. Anticipative measures of flood occurrence

4.2.4.8. Training on flood hazard management

4.2.4.9. Household recovery time and positive effects of flood on household

4.2.5 Household adaptation options

4.2.5.1 Household's perception of Government and NGOs role in flood management

4.2.5.2. Household `s perception of communities role in flood management

4.3 Computation of Flood Vulnerability Index

4.3.1 Identifying key indicators of developed FVI

4.3.2. Normalised Scores and Weight Values of Indicators

4.3.4. Composite vulnerability index of vulnerability factors

4.3.4.1. Exposure factor

4.3.4.2 Susceptibility Factor

4.3.4.3 Resilience Factor

CHAPTER V: CONCLUSION AND POLICY RECOMMENDATION

REFERENCES

ANNEXES

Annexe 1 : Statistical summary of Annual and monthly precipitation for Tabligbo

Annexe 2: Mann-Kendall test results of annual, monthly and seasonal precipitation

Annexe 3 : Statistical summary of Annual and monthly Flow for Athieme

Annexe 4: Mann-Kendall results of annual , monthly and seasonal flow for the study area