Towards More Advanced, Equitable Natural Hazard Risk Metrics

: With the aim of identifying gaps in the state of the art in natural hazard risk metrics, we performed a literature review on some of the risk metrics frequently employed in risk assessments of earthquakes, floods, and hurricanes. Academic researchers use risk metrics to develop risk models, new framework and explore new approaches to assess risks from natural hazards, while policy makers use risk metrics to make informed decisions. Reviewing risk metrics of natural hazards is crucial for understanding and ensuring the equitable distribution of resources and policies in natural hazard management. Direct economic loss, indirect economic loss, casualties, well-being loss, mental health loss, environmental loss, population displacement and recovery time are the risk metrics that will be examined in this research. The report is organised into sections each focusing on a different risk metrics. In each section, the risk metric will be introduced, defined, and discussed, followed by a review of how the risk metric is used theoretically and in practice. Theory papers include academic journal and practice papers include news articles and policy papers. Our review found that direct economic loss is the most prevalent risk metric used in risk assessments of the three natural hazards. This report concludes with discussions of the equity of the risk metrics reviewed, the limitations to our research and proposal of potential risk metrics that are more equitable for future use.


Introduction
Natural hazards such as earthquakes, floods, hurricanes, and more can cause enormous amounts of property damage, injury, disruption to socioeconomic system, loss of livelihoods, environmental damage, and displaces people from their homes worldwide (Bankoff, 2003).A natural hazard can be defined as "environmental phenomena that have the potential to impact societies and the human environment" (Natural Hazards,2023) Some parts of the world are more at risk of certain types of natural hazards than others; for example, areas around the 'ring of fire' are more prone to earthquakes.It is important to understand the risk in order to better prepare and mitigate the negative impact of natural disasters.According to the UNISDR (United Nations Office for Disaster Risk Reduction) terminology, the definition of "risk" is "the combination of the occurrence probability of an event and its negative consequences'' (UNISDR Terminology on Disaster Risk Reduction, 2009) (refer to figure 1).A risk assessment is defined as "a qualitative or quantitative approach to determine the nature and extent of disaster risk by analysing potential risk hazards and evaluating existing conditions of exposure and vulnerability that together could harm people, property, services, livelihoods and the environment on which they depend".Results from a risk assessment can be used to develop strategies to mitigate or manage those risks.

Figure 1. Visualisation of risk as a product of probability and impact with green representing low risk while red representing high risk
A risk metric is a method used to measure the risks of natural hazard and serve as an essential link between the quantitative risk assessment and decision-making (Johansen and Rausand, 2014).Risk metrics are used in both theory and in practice.In theory, risk metrics serves as quantifiable indicators to examine the relationship between natural hazard events, human vulnerability, and resilience.In practice, risk metrics are used by policy makers to make informed decisions and to develop new polices or strategies to mitigate the risks posed by natural hazard.Furthermore, they are also used in news articles to provide readers with information on the impact of natural hazards (Vahdat et al.,2014).For example, a risk metric that quantifies the number of fatalities can be used to identify the type of natural hazard events that lead to the highest number of fatalities which will help in decision making on types of interventions and mitigation strategies to reduce casualties in similar future event.
This report aims to conduct a literature review of risk metrics that have been used in natural hazard risk assessments with a view to identify gaps in the state of the art of the risk metrics, exploring the equity between risk metrics and proposing new metrics to fill the gaps (refer to figure 2).
Reviewing risk metrics of natural hazards is crucial for understanding and ensuring the equitable distribution of resources and policies in natural hazard management.Academic researchers also use these metrics to develop risk models and explore new approaches to assess risks from natural hazards.Investigating equity of risk metrics will help us investigate if the metric can potentially distinguish the differences between the experiences of different social and economic groups in a natural hazard.Through this we can identify the metrics that are more equitable and use them to propose inclusive and equitable disaster preparedness and response strategies.This study will only focus on risk assessments of three natural hazards: earthquake, floods, and hurricane, as they are three of the most devastating and deadly natural hazards (Hannah et al., 2022).Earthquake can damage and even destroy homes through ground shaking, soil liquefaction, and landslides.The sudden influx of water from floods can cause damage to homes, infrastructure, disrupt transportation and communications systems, as well as spread diseases by contaminating drinking water.Hurricanes cause widespread damage and loss of life through strong winds, heavy rainfall, and storm surge.
The report will be organised into sections each focusing on different risk metrics.This report will focus on economic loss, casualties, mental health impact, wellbeing/welfare loss, environmental loss, population displacement and recovery time as these were found to be the most prevalent metrics used in earthquake risk assessments in a preliminary literature review.In each section, the risk metric will be introduced, defined, and discussed, followed by a review of how the risk metric is used theoretically and in practice.
To define a clear scope for this report, criteria will be set out for selecting both theory and practice papers.For theory, journal papers that uses the metrics mentioned above to measure risks of earthquake, floods or hurricane will be reviewed.To ensure the report is up to date, only journal papers that are published from the year 2000 onwards will be included, with the more recent studies with the most citations being prioritised.On the practice side, policy papers and news articles will be reviewed.In the context of this report policy papers are identified as documents that provide evidencebased information to policymakers who will make decisions on disaster risk mitigation policies.As for news articles only articles of the of major media outlets will be reviewed to maintain a level of credibility (Majid, 2023).
To keep the scope of our report clear we will focus on specific country for each hazard for practice papers.For earthquake we will be focusing on Turkey, for floods policy papers and news in the United Kingdom will be reviewed, and finally for hurricane only United States will be reviewed.

Review of Risk Metrics
This following section will focus on the review of different current natural hazards risk metrics used in literature and in practice.They have been divided into the following categories: direct economic loss, indirect economic loss, mental health, casualties, environmental loss, population displacement and recovery time.
In each section, the risk metric will first be introduced and defined, followed by review of some its most cited academic literature, then review of the metric in practice which includes policy papers and news articles.Finally, the remaining academic journal and practice papers will be summarised in two separate tables.

Economic loss
Economic loss resulting from natural hazards such as earthquake, floods and hurricane can be defined as "a perturbation to the functioning of the economic system, with a significant negative impact on assets, production factors, output, employment, or consumption."(Hallegatte et al. 2010).Several authors in the past including Pelling et al., (2002), Cochrane (2004), and Rose (2004), have discussed economic loss and generally the typologies can be distinguished into direct and indirect losses.Economic loss is a very common method to measure the impact of natural hazard as it provides an objective and quantitative way to assess the direct and indirect costs of the disaster.Economic loss estimates consider the damage to buildings, infrastructure, and other physical assets, as well as the economic impact of the disaster on businesses, tourism, and other sectors.In this literature review we are also dividing economic losses into direct economic losses and indirect economic losses as there are significant distinctions between them.

Exposure and Vulnerability Model
Disaster Risk Assessment (Where risk metrics come in) Mitigation Measures

Direct economic loss
Direct economic loss is defined as "the monetary value of total or partial destruction of physical assets existing in the affected area.Direct economic loss is nearly equivalent to physical damage."(UNISDR Terminology on Disaster Risk Reduction, 2009).Physical assets are the basis for calculating direct economic loss.Some examples of physical assets include buildings such as homes, hospitals, commercial buildings, office buildings, schools, factories, public transport systems, telecommunications infrastructures, etc.Since direct economic losses are calculated from damage to physical buildings, most of the direct economic loss from hazards happen throughout the duration of the hazard itself.Direct economic losses are assessed soon after the natural hazard has happened, cost of repair or reconstruction, and insurance payments are estimated.Hence, they are tangible and usually easy to measure.In this section of the literature review we are including studies that produced results of natural hazard risks in terms of direct economic loss where a specific or range of monetary value is given.
One example of measuring direct economic loss in earthquakes is a web-based seismic risk assessment platform known as OpenQuake which is an open-source development of software for seismic hazard and risk assessment launched as part of the Global Earthquake Model (GEM) (Pinho, 2012).OpenQuake consists of a set of calculators with the capability to compute economic losses for a group of assets caused by an earthquake.OpenQuake is then able to output data which includes loss curves, loss maps, and damage distributions.
Many published academic journals papers include direct economic loss as an earthquake risk metric.To begin with, Silva et al (2015) carried out an earthquake risk assessment of entire mainland Portugal using the OpenQuake engine and produced results of earthquake risk in terms of economic losses, including economic loss map and chart of economic loss in terms of construction material, design level, and number of storeys of a building.Corbane et al (2017) utilized Earthquake Loss Estimation Routine (ELER) which is a process of assessing potential damage and economic loss caused by earthquakes to calculate earthquake risk to produce economic loss maps and a chart of top 20 cities in EU in terms of expected economic losses.
Direct economic loss as a risk metric also features prominently in flood risk assessments.For instance, in Huang et al. (2008) the authors analysed the economic losses caused by floods in 1998 in the province of Hunan, China.Their study found that floods have caused significant direct economic loss to the province, particularly in terms of damage to the region's infrastructure and agriculture.A total economic loss of US$ 8.925 million dollars was recorded from the 10,722 families investigated.The authors concluded that flood-related economic loss is closely linked to the severity and duration of the flood as well as the economic levels of the flood victims.The authors also suggested that policies aimed at preventing floods should prioritise measures to prevent river floods as this type of flood causes the most significant economic losses.
Hurricanes also cause direct economic loss mainly through property damage.Strong winds can damage roofs, windows, walls, and other parts of buildings, while heavy rain can cause flooding and water damage.All of these can result in direct economic losses for property owners, insurance companies, and the wider economy.As such, the economic impact of hurricanes can be significant and can have lasting effects on affected communities (After a Hurricane, 2023).Just like earthquake and floods, many journals also use direct economic loss to measure hurricane risk.For example, Pan (2015) estimated the direct economic loss caused by Hurricane Ike in the Greater Houston Region including economic losses from damages to houses and infrastructures.The author used a systematic approach that combines disaster models, a regional input/output model and a spatial allocation model to estimate the spatial distribution of property damage by census tract.The modelling results are reasonably close to the estimates from damage insurance claim after Hurricane Ike.The author also suggested that the modelling results can provide valuable insight for policymakers and planners in identifying cost-effective options for hazard mitigation and resource allocation for post hazard relief equitably.
Direct economic loss as a risk metric is also frequently used in practice.A few weeks after the devastating earthquake struck Turkey and Syria on 6th February 2023, the Financial Times reported that the main earthquake caused $34 billion worth of damage and that rebuilding in all the provinces affected by the main quake and its aftershocks could cost upwards of $68 billion (Samson, 2023b).In July 2020 the UK Department for Environment, Food & Rural Affairs published a policy statement which presented a plan for investment in flood and coastal defences from 2021 that will avoid £32 billion in future economic loss from floods (Eustice and Pow, 2020).More recently, CNBC reported that Hurricane Ian which struck Florida and South Carolina last year resulted in between $50-65 billion of insured damages (Newburger, 2022).
The remaining academic and practice papers reviewed that includes direct economic losses as a natural hazard risk metric are summarized in the tables below: 2.1.1.

Floods
This paper studied the influence of extreme rainfall events on river drainage and sediment transport in highly urbanized catchment by continuously monitoring the discharge and turbidity.
Prior experience as a moderator of disaster impact on anxiety symptoms in older adults (Norris and Murrell, 1988)

Floods
This paper established a dynamic rainfall-runoff model (SWMM) in the study area and used the optimization model and cost-benefit analysis to determine the adaptation scheme and disaster loss.A local-scale analysis to understand differences in socioeconomic factors affecting economic loss due to floods among different communities.(

Hurricane
This paper investigated the effect of rising sea level on the direct economic damage from hurricane using the 2012 Hurricane Sandy as the case study.They found that approximately $8.1 billion out of Sandy's total $60 billion damages are attributable to climate-mediated sea level rise.

Indirect economic loss
Indirect economic loss is defined as "a decline in economic value added as a consequence of direct economic loss and/or human and environmental impacts".(UNISDR Terminology on Disaster Risk Reduction, 2009).Indirect economic losses can be categorized into several distinct groups according to their impact.The first is microeconomic impacts which as an example is when the revenue of a business decline due to the business interruption caused by a natural hazard.Secondly, meso-economic impact is when the revenue of a business declines because of interruptions to the supply chain or unemployment during the immediate days after the occurrence of a natural hazard.Lastly, macroeconomic impacts are impacts on a large scale, they may include increases in prices of products, downward stock market trend, and overall decline in GDP.Indirect economic losses aren't restricted to the area where the natural hazard has occurred and could impact the surrounding areas, on country-wide level or even globally.Indirect economic losses could also continue to have an impact long after the natural hazard's original occurrence, so they can be intangible and hard to estimate and measure.(UNISDR Terminology on Disaster Risk Reduction, 2009).

News article Hurricane
This article reported that the value of the agricultural production in the storm's path during each calendar year was around 8.1 billion dollars.Citrus and vegetables are the most affected, with the economic impact on Florida's agriculture alone estimated to be between 786 million and 1.56 billion dollars.
One example of published articles that uses indirect economic loss as way to measure earthquake risk is by Wu et al (2012), they measured risk by estimating the indirect economic losses caused by the 2008 Wenchuan earthquake in Sichuan.The authors used a model that can reflect earthquake related changes in production capacity and ripple effects within the economic system in the region.The results estimated the indirect economic loss in the housing and production sector to be around 40% of the direct economic loss of the hazard, which is calculated to be approximately 300 billion Chinese Yuan.
Floods, especially river floods have far-reaching and longlasting negative economic effects (Dottori et al., 2018).Therefore, indirect economic loss as a risk metric is also often used in papers regarding flood risk assessments.Tanoue et al. (2020) describes a global modelling framework developed to estimate the indirect economic loss from river floods.The authors applied the framework to the 2011 Thailand flood and found that the estimated economic loss due to business interruption in the industry and service sector was $14.7 billion.The indirect economic losses reduced Thailand's GDP by 4.81% in 2011 and its impact will still be felt up until 2030.
Indirect economic loss as a risk metric is also frequently featured in journals on hurricane risks.In Tatyana Deryugina, Laura Kawano, and Steven Levitt (2018), the paper quantified indirect economic losses in terms of lost income, measured through differences in tax data before and after the hurricane.The results indicate that Hurricane Katrina had a significant negative impact on the incomes of those affected by the disaster, with particularly large losses in the year of the hurricane.
Indirect economic loss as a risk metric is also used in practice.Deutsche Welle reported that the Turkish Enterprise and Business Confederation has put the estimated economic loss from loss of national income to be at $10.4 billion and an additional $2.9 billion from loss of working days (Harper, 2023d).In the UK, the BBC reported at the end of 2022 that severe flooding in Scotland caused road closure and rail disruption.This leads to people not being able to travel to work and disruption to supply chain thus resulting in indirect economic loss (BBC News, 2022).As for hurricane, in 2019 the Congressional Budget Office in the US reported the expected annual loss in the commercial sector including from revenue loss due to disruption of business is $9 billion (Dinan and Wylie, 2019).
The remaining academic and practice papers reviewed that includes indirect economic losses as a natural hazard risk metric are summarized in the tables below: 2.1.2.1 Academic

Hurricane
This paper examined the long-term economic impact of Hurricane Katrina on individuals and households in affected areas using tax return data.
Risk-based input-output analysis of hurricane impacts on interdependent regional workforce systems (Akhtar and Santos, 2013)

Hurricane
This paper discusses how loss of workforce from the aftermath of a hurricane can adversely impact the economic productivity of the affected region.

Practice
Table 4. Practice papers that uses indirect economic loss as a natural hazard risk metric Titles and authors Type Hazard General note What is the likely impact of the earthquakes on economic growth in Türkiye?

Earthquake
This article reported that the indirect economic downturn was brought on by supply chain disruptions, destruction of industries, equipment, and food supplies (loss of physical capital and inventories), worker fatalities and injuries (loss of labour force), and decreased investment in the days and weeks after the earthquake.

Counting the cost of flood
Policy paper

Floods
This report stated that in addition to having a direct impact on economic activity, floods will force the appropriate authorities to invest more money on flood mitigation and flood-prevention technologies.

Hurricane
This article reported that the Florida property loss of Hurricane Ian is nearly estimated to $1 billion.At the same time, it also greatly affected the normal operation of the company, causing employees' salaries to fluctuate to a certain extent.

Casualties
A casualty of a natural disaster can be defined as a person suffering a physical or psychological injury as a result natural hazards (Watt and Weinstein,2013).This section will only focus on physical injuries as there is a separate section for risk metric concerning psychological injuries titled under "Mental Health Impact".People usually suffer injuries during earthquakes as a result of being trapped inside damaged or collapsed buildings, being hit by falling debris, or falling from high places (Watt and Weinstein,2013).Casualties from floods are often result from drowning, injuries caused by debris in floodwater, electrical injuries, and hypothermia (Fatalities associated with floods, 2022).Most casualties of hurricane are caused by indirect causes rather than the storm itself.Indirect casualties are injuries or deaths that cannot be directly attributed to the violent impact of the storm but would not have occurred if the storm hadn't happened.These include carbon monoxide poisoning from using generators indoor, cardiovascular failure, and power outages (Visé, 2022) (Fox, 2016).
Casualties as a risk metric is commonly featured in academic journals regarding earthquakes.Taking Haiti earthquake as an example, the strong earthquake devastated much of Haiti's capital city, Port-au-Prince, and its surrounding area.Many people became casualties of the disaster, around 220,000 people were reportedly killed and 300,000 people were injured because of the earthquake (Miks and Nesbitt, 2022).Majority of the fatalities were due to collapse of buildings with occupants still inside, stabilities of structures are heavily influenced by the stabilities of the ground they are built on, and many of the buildings in Haiti were built on unstable land (International Strategy for Disaster Reduction, 2010).
In Cousins et al (2008), researchers calculated the risk of an earthquake in New Zealand by estimating the number of casualties that result from a magnitude 7.5 earthquake.The result shows a relatively low casualty rate due to New Zealand's long tradition of earthquake resistant building design and policies to remove or upgrade buildings that are at risk.
Flooding accounted for 47% of all weather-related disasters during the previous 20 years, affecting 2.3 billion people and resulting in 157,000 deaths (CRED& UNDRR, 2015).Casualties as a natural hazard risk metric is also often used in academic papers regarding flood risks.Diakakis et al., evaluated the temporal and geographic distribution of flood episodes and casualties during the last 130 years in Greece (Diakakis et al,2012).There were 545 distinct occurrences in total, 686 of which resulted in fatalities and widespread destruction across the nation.The findings revealed seasonality tendencies, with November having the highest concentration of incidents.They also demonstrated that, compared to mountainous and rural areas, metropolitan settings often exhibit greater rates of flood recurrence.
Casualties as a risk metric are also frequently used in hurricane risk assessments.In Brunkard, Namulanda and Ratard (2008), the authors analysed the mortality rate data after Hurricane Katrina.They found that Hurricane Katrina resulted in 1,170 direct deaths and 705 indirect deaths in Louisiana.Majority of the direct cases were caused by drowning, blunt force, and injuries from sharp objects.Meanwhile, indirect deaths were mostly due to lack of access of healthcare and medicines.The paper also highlights the disproportionate impact of Hurricane Katrina on populations that are more vulnerable, such as those with limited access to healthcare and other resources.
Casualties as a risk metric is commonly used in practice.Just four days after February 2023 Turkey-Syria earthquake the BBC reported that the death toll in both countries have exceeded 25,000 with majority of the casualties being from Turkey (Gahagan and Moloney, 2023).In December 2014 the UK Department for Environment Food and Rural Affairs published a report titled "The National Flood Emergency Framework for England" that outlines a set of guideline and procedures for responding to flood emergencies.The report highlights casualties and fatalities as a major risk of flooding (The National Flood Emergency Framework for England, 2014).The National Hurricane Center in the US published a tropical cyclone report of Hurricane Katrina which reported that it caused a total of 1392 fatalities with 520 direct deaths, 565 indirect deaths and 307 indeterminate causes (Knabb, Rhome and Brown, 2006).
The remaining academic journal papers and practice papers reviewed that includes causalities as a natural hazard risk metric are summarized in the tables below: 2.2.1.Academic

Floods
This paper discussed that a flood warning system is essential because flooding has a high death rate.A case study in Orvieto (Italy) produced a model of the early warning system and the impact of casualties.The specialists discovered that a delay of 30 minutes in the warning's release can increase casualties by up to six times, and a successful early warning system can result in significantly reduced casualties.
A GIS for flood risk management in Flanders (Deckers et al., 2009)

Floods
This paper developed a risk-based technique to estimate the number of casualties using a Geographic Information System (GIS) to statistically evaluate flood risk using hydrologic models, land use data, and socio-economic data.This approach was used in the creation of LATIS, a GIS-based flood risk assessment tool, which provides the opportunity to execute risk assessments swiftly and efficiently by calculating the possible damage and number of casualties during a flood event.
Daily Although most did not cause casualties, some extreme cases showed potential for significant harm.The majority of casualties occurred when and where the physical risk of tornadoes was higher.

Mental Health Impact
One of the most important challenges that natural hazards survivors have to face is the impact on their mental health.The wellbeing of a person's mind is known as "mental health," which speaks about the consistency of thinking and feeling at peace since everything is well.According to one study on the mental health of earthquake survivors, 42.6% of them had moderate to severe mental health problems (Yokoyama et al., 2014).
One of the common methods for studying people's postdisaster mental health is the PTSD (posttraumatic stress disorder) index to measure psychological trauma.Posttraumatic stress disorder (PTSD), an anxiety disorder that can develop after exposure to any incident that causes psychological trauma (Neria et al., 2007), is the post-disaster mental condition that has received the most research in recent years."PTSD symptoms are broken down into three symptom clusters: five re-experiencing symptoms, seven numbing/avoidance symptoms, and five hyper-arousal symptoms (DSM-IV criteria B, C, and D, respectively)" (Dobie M.D. et al., 2002).Simply, it entails giving participants a questionnaire to fill out about a stressful experience they've had in the past.Adults with a score of 50 or higher are identified as having PTSD risk, and the total score ranges from 17 to 85 (Thepa&Hauff, 2005).
Mental health as a risk metric is also used in academic papers regarding earthquake risks.For instance, in Hong et al., 2009 the authors carried out the semi-structured diagnostic interview (CIDI) on survivors of the Zhangbei-Shangyi earthquake in 1998 based on DSM-IV diagnostic standard.It was found that the incidence of PTSD was 25.3% and 14.2% for 181 people in Duizhuang, two villages, three months after the earthquake, while the follow-up study nine months after the earthquake found that the incidence of PTSD was 30.3% and 30.3% respectively.
Floods also cause severe trauma for people who are impacted.Maltais and colleagues discovered that those afflicted two years after a flood in Canada had increased levels of depression, PTSD symptoms, psychological distress, and worse adjustment.South Korea was found to have a considerable decline in health-related quality of life 18 months after the flood (Fernández et al., 2015).Other concerns include the necessity to restore emotional ties to one's home as well as the stress and worry brought on by the reconstruction and rebuilding efforts made in flooded homes (Springett et al, 2017c).Mental health impact can be also observed four years after a significant flood in Banbury, UK (Tapsell and Tunstall, 2008).In short, most in-depth research conducted in the years following the floods looked at how flooding affected people psychologically, including suicide, PTSD, sadness, and anxiety.After flooding, social support is a protective factor that can enhance long-term health results (Zhong et al, 2018).
Mental health as a risk metric is also prevalent academic journals regarding hurricane risk assessments.In Schwartz et al. (2018), the authors provided a preliminary assessment of Hurricane Harvey's impact on mental health in affected areas.A survey was conducted on individuals impacted by Hurricane Harvey, with questions regarding their exposure to the hurricane and its aftermath.The survey found that very high percentage of people reported experiencing negative mental health outcomes including depression, anxiety, and PTSD, which falls under the definition of mental health impact in this report.The authors also highlighted the need to address symptoms of PTSD in affected people as early as possible and to provide long-term support.
In practice, after the February 2023 Turkey-Syria earthquake Al Jazeera reported the emotional impact of the earthquake and emphasised on the importance of providing mental health and psychological support to the victims (Uras, 2023).USA Today also highlighted the potential long-term psychological impact of Hurricane Ian and the importance of for offering mental health support to victims of Hurricane Ian as they may have experienced tremendous trauma and stress (USA Today, 2022).
The remaining academic journals and practice papers reviewed that includes mental health impact as a natural hazard risk metric are summarized in the tables below: 2.3.1.Academic

Floods
This paper investigates the characteristics linked to psychological distress and long-term mental health decline.In the long-term study, household income and flood depth were the two most significant predictors, as tough financial requirements could cause strong mental issues.According to the study, implementing mitigation strategies may improve the mental health outcomes for communities who are at risk.Psychological impact of the hurricane Mitch in Nicaragua in a one-year perspective (Caldera et  This document published by the National Health Institutes of Health (U.S) investigated the change in mental and physical health among low-income parents exposed to Hurricane Katrina.The study found that the prevalence of probable serious mental health issues doubled and close to half of the respondents exhibited probable PTSD.The study also found that higher levels of hurricane related loss were closely linked to worse mental health outcomes.

Wellbeing/welfare loss
Wellbeing of a person has no single definition, but it generally includes positive emotions, satisfaction with life, and positive functioning.Some disciplines also include aspects such as economic wellbeing, personal development, and engaging work (Well-Being Concepts, 2018).Welfare of a person can also be defined similarly as "the general state of health or degree of success etc." (Cambridge Dictionary, 2023).In the context of this report, wellbeing loss and welfare are defined as a negative impact on an individual's quality of life due to loss of income or livelihood and loss of access to essential services such as water, electricity, and shelter as a result of natural hazards.
Several academic papers have used the well-being as a risk metric to assess earthquake risks.Traditionally, scholars used assets to assess risks.On the other hand, Walsh and Hallegatte (2019) simultaneously evaluate earthquake risks using two metrics: socioeconomic resilience and well-being losses.According to a case study in the Philippines, about half of Filipinos are thrown into poverty each year after a disaster, this study determines that low-income earners suffer about a 9% loss in assets but a 31% loss in their well-being.The benefits of interventions may be easily quantified thanks to the use of metrics that take poverty and well-being into account.Besides, when studying the earthquake shocks and well-being outcomes based on the Yogyakarta earthquake that rocked Indonesia in 2006, present well-being, people's hopes for the future, the style of life they lead, and happiness both before and after an earthquake are the main subjects of this study.According to the data, there is enough proof that seismic disasters caused a noticeable reduction in well-being.The investigation revealed that the poor health persisted for years after the earthquake had passed (De & Thamarapani, 2022).
Floods also have a certain impact on subjective well-being indexes.Hudson et al. (2019) explored the intangible effects of flood risk on the subjective well-being of inhabitants in central Vietnam, when a flood occurs, there will be a temporary decline in the general public's well-being.Based on an empirical analysis of 320 counties impacted by the 1993 Midwest flood, Xiao and Feser (2014) give large-sample evidence of the unemployment impact of catastrophes.They discovered that the flood caused unemployment rates to surge in severely damaged counties but that the impacts rapidly subsided.The average person's pay has decreased somewhat after the floods.
Hurricanes also have major impact on people's welfare.Thompson et al (2005) examined how many individuals were unemployed following the natural hazard.In order to assess people's employment welfare levels following the hurricane, they discuss the impact of this windstorm on the period behaviour of the unemployment rate alongside recovery programmes implemented by the public and private sectors.
In practice, CNBC reported that thousands of people are left sleeping on the streets with no access to electricity and water after an earthquake struck Turkey in February 2023 (Khan, 2023).After Hurricane Ian struck Florida in October 2022 CNN reported that tens of thousands of people are likely to lose their jobs and file for unemployment benefit in the aftermath of the storm (White, 2022).
The remaining academic journal papers and practice papers reviewed that includes wellbeing/welfare loss as a natural hazard risk metric are summarized in the tables below: 2.4.1.Academic

Floods
This paper hypothesised that gender differences in the impacts of floods on welfare could lead to an initial decline in welfare through subjective well-being when a flood occurs within five years.Men responders often recovered 80% of their welfare losses, while female responses were linked to a welfare recovery of about 70%.people need to utilise their own money to make up for the welfare they are missing.

Hurricane
This paper analysed the impact of hurricane risk on individual wellbeing in the United States.The author employed an approach that focuses on the indirect effects of natural hazards on individual wellbeing, such as through their impact on economic activity and job opportunities.

Practice
Table 10.Practice papers that use wellbeing/welfare loss as a natural hazard risk metric

Titles and authors Type of paper Hazard General note
Hundreds of thousands lost their jobs in Turkey-Syria quake: UN

News article Earthquake
This article reported that the earthquake had a terrible effect on employees and businesses.The hours wasted amounted to almost 657 000 employees' worth of labour.Thousands of workers lost their job, dividends, and fringe benefits.

News article Hurricane
This article reported that to address the inflation brought on by Hurricane Harvey in Texas, which resulted in an increase of nearly 50,000 in August 2017, citizens must apply for employment subsidies in massive numbers.
UK floods: Environmental Agency job cuts 'on hold'

News article Floods
This article reported that to ensure the budget is adequate for post-disaster reconstruction, employment was cut down around 1,550 due to a decrease in the amount of money received from the central government after the flood.

Environmental loss
Environmental losses relate to the negative effects that natural hazards have on the environment.The environment will undoubtedly be impacted by the occurrence of a natural hazard.For instance, the degradation of vegetation on both sides of rivers contributed to environmental damage by triggering landslides, debris flows, and erosion.The effects on the air, such as  emissions due to the natural hazard, is also an example of environmental impact.(Chen et al., 2012).
Environmental loss as a risk metric is frequently featured in academic journal papers regarding earthquakes.An example of an environmental loss due to earthquake is clean water and air contamination.Environmental losses are a type of non-financial loss in natural hazards People's well-being is directly tied to these factors required for survival (Sangha et al., 2020).Welsh-Huggins and Liel (2018) investigate the impacts of alternative structural concretes on the life-cycle sustainability and resilience of a reinforced concrete building in a high seismic region, and conduct a life-cycle environmental impact assessment, quantifying greenhouse gas emissions associated with building construction and seismic performance, accounting for potential earthquake damage and subsequent repairs.
Floods have a terrible effect on the ecosystem and the surrounding soil, as well as majority of animals (Istomina et al., 2004).Istomina et al. stated floods have an adverse effect on soils, cause fluvial morphological deformations, disrupt plant cover, and harm wildlife.The floods also change the chemical composition of water and dramatically worsen its quality；Hickey and Salas (1995) analyse the impacts of short-and long-term environment after extreme floods; Based on 835 instances of floods brought on by excessive rainfall in Hungary over the course of two decades, Czigány et al.(2009) examined the area's environmental effect and forecasted where flash floods will occur using GIS and other technologies.
Academic research papers also use environmental loss to measure hurricane risk.Mallin and Corbett (2006) stated that common adverse effects on water quality caused by hurricanes include excessive nutrient loading, increased biochemical oxygen demand and subsequent hypoxia and anoxia, fish and invertebrate deaths, aquatic animal scale replacements, chemical pollutants and debris from damaged human structures, and pollution of water；Fran (1996) and Floyd (1999), the two greatest hurricanes, had effects on hydrology, nutrient loads, fish reproduction, soil erosion, and biotic composition that lasted from several months to many years (Paerl et al., 2006).
Environmental loss as a risk metric is also used in practice.Le Monde reported that the February 2023 Turkey-Syria earthquake may have an impact on the water quality nearby.The absence of access to bathrooms and drinking water for the survivors living in temporary camps is greatly increasing the risk of an epidemic.Moreover, poor sanitary may cause rapid transmission of diseases (Cazorla, 2023).The UK Environment Agency has warned that climate change will lead to more frequent and intense floods.The published document points out that floods caused by weather changes will greatly affect environment, such as the surrounding crops and ecological environment (Environment Agency, 2018).The Washington Post reported that because the hurricanes generated a lot of waste, people flouting laws governing gasoline use in car fuel tanks, plastic linings, and other contaminants, Hurricane Ian will leave behind a trail of environmental hazards (Mufson, 2022).
The remaining academic journal papers and practice papers reviewed that includes environmental loss as a natural hazard risk metric are summarized in the tables below: 2.5.1.Academic This paper analysed the effect of Hurricane Ivan on water quality in Pensacola Bay, Florida.The authors assessed the quality of water using parameters such as dissolved oxygen, nutrients concentration and salinity.

General note
Geological impact of Turkey-Syria earthquake slowly come into focus (Ravilious, 2023)

News article
Earthquake This article reported that there are several landslides and rockfalls in the area where the earthquake occurred, it will have a significant impact on the future agricultural production in the region.
Flooding in UK: Ecological impacts and an ecosystem approach News article Floods This article layout that humans gain several benefits from the freshwater system.Extreme weather conditions may harm the ecology permanently.Extreme flooding has been proven to significantly reduce plant biomass as well as fish and pearl mussel abundance.Hurricane Ian damages homes and the environment (Penick, 2022) News article Hurricane This article reported that Hurricane Ian caused environmental damage in Florida, including the release of thousands of gallons of diesel and water that "looks like root beer and stinks like dead fish rolled into compost."

Population displacement
Population displacement in the context of natural hazards means the forced removal of individuals or communities from their homes or places of residence due to the immediate or long-term impact of natural hazards (UNISDR Terminology on Disaster Risk Reduction, 2009).Natural hazards can lead to population displacement directly through damage to homes and infrastructures or indirectly through loss of livelihoods and disruption of essential supplies and services.Population displacement can be temporary or permanent, the duration depends on the extend of the damage and the ability of the displaced population to return and rebuild (Mallick and Vogt , 2014).People's choice to evacuate during a crisis is strongly influenced by how risky they perceive the situation to be.Other people perceive catastrophe danger strongly, while others perceive it weakly, and some people even perceive disaster risk indifferently (Lindell et al.2000).Therefore, some individuals who live in disaster risk zones will decide to evacuate in the face of prospective catastrophe hazards, whilst others are unwilling to relocate for a variety of reasons.
Natural hazards can severely damage the local infrastructure and has a great impact on the local social economy.A serious societal impact is that people may need to relocate their homes and assets following a natural disaster.According to a report released by the Japanese National Police Agency on October 9, 2015, the Tohoko Japan earthquake left 2567 people missing, over 400million people had to leave their homes due to the need to evacuate and about 400 000 properties sustained damage, more than half of which were irreparable (Tsuchiya et al., 2017); Notably, following the 2015 earthquake in Nepal, an estimated 390,000 more people than usual left the Kathmandu valley, with the majority migrating to nearby places and the densely populated regions in the country's central and southern areas (Wilson et al., 2016) ; Moreover, a significant earthquake that year rocked northwest Turkey in August.A needs assessment was conducted to determine the population's urgent requirements.Interviews were conducted with 230 household representatives from the four earthquake-worst areas.84% of families were uprooted from their homes and were living in temporary shelters (Daley, Karpati and Sheik, 2001).
Population displacement also features prominently in flood risk assessments.In Kakinuma et al. (2020b) the paper presents an analysis of flood-induced population displacement in the world.The authors highlighted that flood could result in significant population displacement of populations both temporary and permanent.Various data sources on flood-induced displacement were analysed and South Asia, Southeast Asia, and Sub-Saharan Africa were regions most vulnerable to that type of displacement.
Population displacement is also a major risk metric of hurricane and frequently used in academic research.In Acosta et al., 2020 the paper analysed data from mobile phones and social media to estimate the number of people at risk and the migration pattern in Puerto Rico after Hurricane Maria.The authors observed a population loss across all data sources throughout the study period.However, the number differed significantly across different data sources.
Population displacement is also featured prominently in practice.Al Jazeera reported that nearly 530,000 people have been evacuated from the disaster area after the February 2023 Turkey-Syria earthquake and many survivors have settled in tents, container homes and other governmentsponsored accommodation (Jazeera, 2023).After Hurricane Ida in 2021, CNN reported that one region in Louisiana saw over 14,000 people left homeless after the hurricane damaged or destroyed 75% of the buildings there (Holcombe, Levenson and Selva, 2021).
The remaining academic journal papers and practice papers reviewed that includes population displacement as a natural hazard risk metric are summarized in the tables below: 2.6.1.Academic Table 13.Academic journal papers that include population displacement as a natural hazard risk metric

Titles and authors
Hazard General note Agent-based model simulations of future changes in migration flows for Burkina Faso.Global Environmental Change.(Kniveton et al., 2011 Earthquake This paper states that there are two kinds of original site resettlement, one is to rebuild new houses in the destroyed houses; The other refers to the maintenance and reinforcement of houses on the premise that they are damaged by earthquakes but do not constitute dangerous houses.Be proactive for better decisions: Predicting information seeking in the context of earthquake risk.International Journal of Disaster Risk Reduction.(Li & Guo, 2016) Earthquake This paper defined relocation as moving away from the current residence to build a new house in town.This way can grasp the opportunity to upgrade the infrastructure construction, and then improve the living environment of residents.
A generic decision model for developing concentrated rural settlement in post-disaster reconstruction: a China study.Natural Hazards.(Peng et al., 2013) Earthquake This paper discussed that the operation of decentralized resettlement is simple, but it is not conducive to sustainable development and infrastructure construction, which is of little significance to improving residents' quality of life.
The risk perception paradox-implications for governance and communication of natural hazards.(Wachinger et al., 2012) Earthquake This paper states that residents' satisfaction with the living environment also has an impact on residents' final mode choice.Studies show that people prefer to live in places with comfortable environment, harmonious neighbourhood and far away from disasters.

Floods
This paper states that in high-income countries like the UK, flood occurrences often result in few direct fatalities; instead, the rise in mental health impact brought on by constant relocation causes the most significant health impact.
Relocation was linked to greater levels of anxiety and depressive symptoms.
Hurricane Events, Population Displacement, and Sheltering Provision in the United States (Mitchell, Esnard and Sapat, 2012) Hurricane This paper focused on the impact of hurricane events on population displacement in the Unites States.The authors studied the patterns of population displacement during hurricane events using data from the Federal Emergency Management Agency (FEMA) and the U.S. Census Bureau, the study found that hurricane events result in high population displacement especially for people in low-income households.
Population Displacement and Housing Dilemmas Due to Catastrophic Disasters (Levine, Esnard, and Sapat, 2016) Hurricane This paper analysed the impact of natural hazards on population displacement and housing dilemmas.The authors examined the patterns of population displacement and housing dilemmas during natural hazards using data from various sources.The authors also suggested that effective disaster response and recovery effort should focus on development of proactive housing strategies that account for the diverse needs of impacted communities.
The Long-Term Recovery of New Orleans' Population After Hurricane Katrina (Fussell, 2015) Hurricane This paper analysed the long-term recovery of New Orleans' population after Hurricane Katrina.The authors examined the change in population, demographic shifts, and housing market dynamic within the city after the hurricane.The paper also suggested that post hazard recovery plan should prioritize the development of affordable housing.

Recovery time
Recovery time in the context of natural hazards refers to the time needed for the affected region and population to return to a state of normalcy and stability after the occurrence of the natural hazard.There are many ways to define recovery from a natural hazard.Smith and Wenger (2007b) defined the recovery process as "the differential process of restoring, rebuilding, and reshaping the physical, social, economic, and natural environment through pre-event planning and postevent actions".While the UN Office of Disaster Risk Reduction defined recovery as "decisions and actions aimed at restoring or improving livelihoods, health, as well as economic, physical, social, cultural and environmental assets, systems and activities, of a disaster-affected community or society, aligning with the principles of sustainable development, including build back better to avoid or reduce future hazard risk" (Proposed Updated Terminology on Disaster Risk Reduction: A Technical Review, 2015).For the purpose of this report, the key point is that recovery time should only be measured in terms of time and not in terms of money or other unit of measurements.
Many foreign countries have conducted in-depth research on the problem of earthquake recovery and reconstruction, among which Japan has a relatively high frequency of earthquakes due to its geographical characteristics.In 1995, a 7.5-magnitude earthquake struck Hanshin.By making an effective reconstruction plan, the economy in the disaster area returned to normal within a month, and so did the residents' lives.Restoration and reconstruction and earthquake resistance of buildings are one of the key research issues in Japan.Itsuki (2017) has made a comparative study on the restoration and reconstruction procedures in Japan, Taiwan Province, and Turkey.Martinez and Hirayama introduced and studied the problems of post-disaster planning and housing construction in El Salvador and Kobe.From research of American scholars on restoration and reconstruction, it can be found that the implementation and planning of restoration and reconstruction are the two main points in the study of rehabilitation and reconstruction in the United States.Olshansky believes that speed directly affects the reconstruction of disaster areas, so fast and efficient planning is the priority.Bryson introduced the operational research model into the reconstruction in order to optimize the reconstruction planning, which provided some reference for the decision-making department (Bryson, 2016).
Recovery time is also used as a risk metric in academic papers regarding flood risks.Asgary et al. conduct research about the effect of floods on small enterprises and the reasons for their recovery are examined six months after the 2010 floods in Pakistan.Small firms in disaster-prone areas of poor countries are less equipped to create and carry out business continuity plans because they lack the necessary personnel, funding, and awareness of their vulnerability.90% of the sample firms (sample size=500) reopened six months after the flood, according to the findings, but the bulk of them were doing so at a loss, and only a few were operating at or aboveaverage levels (Asgary et al, 2012).Besides， 8-9 months after the catastrophic floods in Central Europe in 2002.An investigation carried out by Thieken et al. ( 2007), over 2,000 homes were chosen at random and split up into several groups.During the study, it was discovered that those who were knowledgeable about flood management and self-defence were less negatively impacted by the floods and recovered more quickly.
The recovery of New Orleans following Hurricane Katrina has been the subject of numerous studies; this hurricane profoundly impacted people's lives and employment.Firstly, Zottarelli (2008) studied the post-disaster employment recovery process; the data in this article give a certificate to examine the short-and medium-term employment recovery roughly one month and one year after Hurricane Katrina.Secondly, Fussell also proposed a study of the problem of population displacement and environmental foundation recovery.The reconstruction of the city's businesses, homes, and infrastructure was completed in 2006, four years after the hurricane.Moreover, Kates predicted that the process of making the built environment functional after hurricane would take 60 weeks and that improving, replacing, or rebuilding the built environment would take between 8 and 11 years (Kates,2006).
Recovery time is also frequently used in practice.For instance, a news article by The Washing Post reported that one of the experts said rebounding from Hurricane Ian could take up to a decade, with one of the toughest hurdles being a shortage of affordable housing after the storm destroyed most of them (Sacks, 2023).
The remaining academic journal papers and practice papers reviewed that include recovery time as a natural hazard risk metric is summarized in the tables below: 2.7.1.Academic Table 15.Academic journal papers that include recovery time as a natural hazard risk metric

Titles and authors
Hazard General note Community vulnerability and capacity in post-disaster recovery: The cases of Mano and Mikura neighborhoods in the wake of the 1995 Kobe earthquake.(Yasui, 1970) Earthquake This paper introduced and studied the problems of post-disaster planning and housing construction in El Salvador and Kobe.Hanes had an in-depth discussion on Tokyo's post-disaster planning.
Post-earthquake housing reconstruction programme.Open House International.(Martinez, 2005) Earthquake This paper introduced the operational research model into the reconstruction in order to optimize the reconstruction planning, which provided some reference for the decision-making department.
After the Rain -learning the lessons from flood recovery in Hull.(Whittle et al. 2010)

Floods
This paper examined the health, economic, and social aspects of the longer-term experience of flood impact and recovery along with the identification and documentation of essential factors.In-depth research that was conducted after the flood by interviewing a large number of families and recording how they eventually returned to their pre-disaster living arrangements.Disaster recovery and business continuity after the 2010 flood in Floods This paper documented the recovery of Pakistani businesses six months after the flood.However, the findings indicate that 90% of the sample businesses were Pakistan: Case of small businesses (Asgary et al., 2012) operating at a loss when reopened, with only a small percentage doing at par or better.It will substantially improve the revival of the firm if relevant authorities can offer financial assistance to select small businesses.
Coping with floods: preparedness, response, and recovery of floodaffected residents in Germany in 2002 (THIEKEN et

Discussions of results
This report reviewed eight different risk metrics that are used to measure earthquake, floods, and hurricane risks.These risk metrics cover virtually every element of life, from economic to behavioural factors.From our review it is evident that economic loss is the most prominent risk metric for measuring natural hazard risks (refer to figure 3).Out of all the risk metric reviewed economic loss dominates in terms of the number of academic journal papers it was featured in at 35.Casualties, mental health impact and population displacement and recovery time are four risk metrics that are also commonly featured in the literatures reviewed although at less than half the number of academic journal papers than economic loss.Risk metrics such as wellbeing loss and environmental loss are comparatively less represented and appear less frequently in the academic journal papers reviewed.
Moreover, even within economic loss itself the distribution between direct and indirect economic losses are not equal.Direct economic loss is noticeably featured much more frequently as a natural hazard risk metric than indirect economic loss (refer to figure 4).This is most likely due to the fact that direct economic loss is much easier to quantify and measure.On the other hand, indirect economic loss is often not tangible and not as easily identifiable, making it more difficult to calculate.Risk metrics used in practice tells a similar story.Economic loss is again the most used natural hazard risk metric in practice (refer to figure 5).Economic loss is featured prominently in both news articles and policy papers.However, for the rest of the risk metrics they are all more commonly found in news article than policy papers.For wellbeing/welfare loss not mention of it was found in policy papers.This could be due to the fact that it is much harder to define and measure than other metrics such as economic loss.Moreover, it can be seen that the number of news articles that featured casualties is close to the number that featured economic loss as a risk metric.In fact, direct economic loss and casualties are often presented together in many news articles regarding natural hazards.The number of casualties is also found to be much high in earthquake and comparatively much lower in floods and hurricane.This is most likely due to the fact that floods and hurricanes usually can be predicted and warned early so that the population can make preparations for it while earthquake often strike with little to no warning.In practice, direct economic loss is again more used than indirect economic loss within economic loss (refer to figure 6).Factors contributing to indirect economic loss such as business interruption and loss of revenue are often mentioned in practice papers but rarely quantified.This can also be explained by the fact that indirect economic loss is intangible and difficult to measure.

Comparision between direct and indirect economic losses in practice
Economic losses are not an equitable risk metric since they do not measure equal impact on all people or organisations.In fact, Individuals who are already poor or marginalised in society are frequently more adversely affected by economic losses.It can create a misleading representation of the hazard's impact since high-income individuals may incur larger monetary losses, but these losses often only represent a small proportion of their overall wealth, and they often have higher income and savings which means they are able to recover from their losses quickly.Conversely, low-income individuals may suffer much less monetary losses, but they could be significant portion of their wealth.People such as low-income individuals, small enterprises, and areas with limited infrastructure, are likely to suffer the smaller economic losses in terms of monetary loss from a natural hazard.These individuals and organizations might not have the means and abilities to recover from the financial losses brought on by the hazard, which further increase their unpredictability and vulnerability.As a result, using absolute monetary figures often overlook the difficulties and hardship suffered by more vulnerable individuals who will have a much harder time trying to recover from the hazard.
Comparatively, casualty as a natural hazard risk metric is more equitable.Casualties consider the impact of natural hazards on human life which do not disproportionately benefit wealthier individuals or regions.By measuring natural hazards in terms of injuries and deaths policy makers can more accurately assess the impact of those hazards on more vulnerable populations and take steps to mitigate the adverse effects.However, casualties as a risk metric does not take long-term impacts such as loss of livelihoods, displacement and mental health impact into account.Moreover, wealthier individuals are more likely to receive faster and better healthcare as they have the money to pay for it.
Mental health impact as a natural hazard risk metric considers the psychological well-being of individuals affected by the natural hazard.Mental health may not affect everyone equally as similar to economic loss, individuals or communities who have pre-existing vulnerabilities and limited access to mental health resources may be affected more by the hazard.However, this means that people who are more affected will be represented as so thus more help and resource can be allocated to them, making the risk metric more equitable than economic loss.
Wellbeing and welfare loss is another natural hazard risk metric that is potentially more equitable.Contrary to asset losses, which only take monetary value loss into account, well-being losses take the utility of consumption into account as it varies over the course of the recovery process (Boakye et al. 2020).The San Francisco Bay Area earthquake serves as an illustration of how well-being losses following a disaster impact the poor three times more severely than the wealthy (Hallegatte, 2021).
Environmental loss as a risk metric measures the impact of the natural hazard on the environment rather than humans.Environmental loss affects different individuals and communities differently, depending on their reliance on the natural resources and the environment.However, the relationship between different aspects and consequences of environmental loss can be complex and difficult to measure.
Population displacement is an equitable risk metric as it measures the number of people who are forced to relocate either short-term or long-term from their homes due to a natural hazard regardless of their financial standings.However, wealthier individuals or families could move to other properties elsewhere that they own while poorer people who don't have anywhere else to go are forced to stay in shelters or even on the streets.To ensure that it is equitable the risk metric should take into account the places where displaced population end up and the length of time they are displaced for.
Recover time is a risk metric that is potentially more equitable than economic loss.Recovery time can measure the time it takes for different individuals or communities to recover from the impact of the natural hazard.From this vulnerable individuals or communities which are taking much longer to recover can be identified and resources can be allocated to those that are more in need.Moreover, recovery time can measure the length of time of business closure which is more equitable as measuring revenue loss will always show larger business as losing more than small business but larger business could probably sustain itself longer during closure.
All of the risk metrics above are used by many professionals, companies, and organisations both private and governmental all over the world.They are valuable assets to policy makers to make informed decisions and thus able to develop policies that address risks posed by different natural hazards.Some examples include risk assessment, land use planning, community engagement, and recovery planning.Insurance companies also make use of these risk metrics in the pricing of their insurance premiums regarding losses due to natural hazards.Vulnerable populations sometimes do not receive they aid and resources they require due to data skewed by inequitable risk metrics.It is therefore important for policy makers to use equitable risk metrics for their decision making to ensure that policies and strategies for natural hazard risk reduction are fair and just for every member of society.

Limitations
This report has two major limitations: 1) Due to limited time, only a finite of risk metrics and academic and practice papers can be reviewed.
2) This report also doesn't cover all types of natural hazards and only include earthquake, floods, and hurricane.
Both of those limitation means there may be other equitable risk metrics that are used in other risk assessments but have not been identified and included in this report.
There could also be many more policy papers that utilised the risk metrics discussed above but could not be found due to the lack to public access.Each of the natural hazard is only focusing on one country in this report, if more countries are explored for each hazard there could potentially be more policy papers that can be accessed and reviewed.
Economic losses both direct and indirect and casualties as a natural hazard risk metrics are also used frequently by professional assessors hired by insurance companies for their insurance pricing strategies.This is another example of risk metrics being used in practice.However, we could not review those documents as they are most often not readily available to the public.
The limitations discussed above could serve as a guide for the planning of future work.

Proposal of more equitable risk metrics
While carrying out literature review, we came across some risk metrics that could potentially be more equitable than economic loss but are rarely used.These metrics could offer more equitable insight into victims of natural hazards.
For instance, in Nofal et al. (2021) the paper calculated losses for impacted buildings in North Carolina in terms of percentage structural, contents, and total losses after Hurricane Florence in 2018.Measuring economic loss in terms of percentage of a person's assets after a natural hazard offers a more equitable perspective as it accounts for the varying financial capacities of the victims.Through focusing on the proportion of person's total wealth, we can gain more accurate understanding of the relative losses experienced by victims of the natural hazard regardless of their financial standings.Measuring economic loss as a percentage of a person's wealth can enable policymakers to organise a more equitable distribution of resources and support following a natural hazard, prioritizing those who have been most affected by it.
Another potential equitable risk metric is loss of education.Education is a important factor in life.Children losing access to education can have detrimental impact the community, disruption to community development and have long term consequences which could include reduced earning potential, reduced quality of life and lower social mobility.It is therefore important to quantify the impact of education loss.The metric can be measured in terms of damage or destruction of educational facilities leading to interruption of education for a period of time measured in days or months etc. due to natural hazards.For example, a hurricane could force a school to close to days, and a devastating earthquake could lead to complete destruction of school facilities leading to the school being closed for months, years or even permanently.All of these could potentially result in children being delayed and missing out on valuable education.

Conclusion
In conclusion, this report reviewed eight earthquake, floods, and hurricane risk metrics one by one in both academic and in practice.The result analysis found that direct economic loss is the most dominant risk metric used in both academic and practice while other metrics such as casualties, mental health impact and population displacement are comparatively much less represented in both academic and practice with the exception of casualties also being used as much as economic loss in news articles.The discussion found economic loss to be not an equitable risk metrics while the rest of the metrics are also discussed, and their level of equity discussed.The two major limitations of this report are found to be the limit in time and the exclusion of other natural hazards.Finally, two risk metrics which are rarely used but could be equitable are identified and their potential evaluated.

Figure 2 .
Figure 2. A flowchart visualising the process of natural hazard risk assessment and where risk metrics come in Migration and Displacement triggered by Floods in theMekong Delta(Dun, 2011)FloodsThis paper aimed to ascertain whether flooding qualified as a factor in migration or displacement.Findings indicate that the effects of the Mekong Delta's recurrent floods can lead to independent household or individual relocation decisions and are a reason for government-initiated household resettlement.A climate of control: flooding, displacement and planned resettlement in the Lower Zambezi River valley, Mozambique(Munro et al., 2017)

Figure 3 .Figure 4 .
Figure 3.A chart summarising the number of academic journal papers reviewed for each risk metric

Figure 5 .
Figure 5.A chart summarising the number of practice papers found for each risk metric

Figure 6 .
Figure 6.Comparison between direct and indirect economic loss in practice

Table 1 .
1 Academic Academic journal papers that includes direct economic loss as a natural hazard risk metric a case study on various earthquake scenario that might occur in the Boston Metropolitan area.Result includes breakdown of economic loss estimate in terms of monetary value in a table.Urban seismic risk assessment by integrating direct economic loss and loss of statistical life: an empirical study in Xiamen(Zhou et al., 2020)Earthquake This paper measured earthquake risk of Xiamen in terms of direct economic loss and produced results in the form of direct economic loss maps with measurement in monetary value.
EarthquakeThis paper utilized Hazus to produce maps of "Sub-province-based loss ratio corresponding to 475 years return period" and "Sub-province based average annualized loss ratio (AALR) distribution for Turkey".Development of the OpenQuake engine, the Global Earthquake Model's open-source software for seismic risk Earthquake This paper produced earthquake risk maps of Turkey in terms of economic losses, including loss map, total economic loss exceedance curves for a portfolio of assets.

Table 2 .
Practice papers that uses direct economic loss as a natural hazard risk metric

Table 3 .
Academic journal papers that include indirect economic loss as a natural hazard risk metric (Deryugina, Kawano and Levitt, 2018)ate by using Leontief inputoutput model.By using the modified linear input-output model and dynamic input-output model, this paper explains how to quantify and evaluate the impact of chain reaction on large-scale systems in theory.The Economic Impact of Hurricane Katrina on Its Victims: Evidence from Individual Tax Returns(Deryugina, Kawano and Levitt, 2018)

Table 5 .
Academic journal papers that include casualties as a natural hazard risk metric (Gul and Guneri, 2016)Earthquake This paper estimated the earthquake risks of four regions in Istanbul in terms of casualties and produced casualties' estimation chart for each of the region.Seismic estimation of casualties and direct economic loss to Byblos city: a contribution to the '100 resilient cities' strategy.(Makhoul,EarthquakeThispaper estimated the earthquake risk of Byblos in terms of casualties and summarized the result in a table.The results are also presented in a casualties' severity map and chart. the daily variation in hurricane and tornado casualties using data from the Spatial Hazard Events and Losses Database for the United States (SHELDUS).It finds that hurricanes are more lethal on weekdays.Hurricane forecasts provide more lead time, allowing strategic behaviour in choosing protective measures.The study's findings have important policy implications for managing the impacts of hurricane.

Table 6 .
Practice papers that include casualties as a natural hazard risk metric Syria earthquake has claimed the lives of 35,418 in Turkey and the death toll is still rising.

Table 7 .
Academic journal papers that includes mental health impact as a natural hazard risk metric approximately 200 people who the age of 60 to participate in a survey to measure the value of PTSD index, that measured their levels of anxiety, depression, self-reported health, and life satisfaction in order to improve future service planning.It was found that maybe one in six people needed clinical observation because their PTSD index was higher than other observed people.But all the older had negative psychological repercussions as a result of the flood.

Table 8 .
the mental health impact of Hurricane Maria on Puerto Ricans living in Puerto Rico and Florida.The authors found that rates of PTSD were high in both places.The study also highlights the challenge faced by Puerto Ricans residing in Florida who are separated from their families and communities.Practice papers that use mental health impact as a natural hazard risk metric Turkey earthquake: How the true costs calculated?(Harper, 2023)News article EarthquakeThis article reported that after Turkey earthquake, the individuals are frightened to go back to work with poor mental state, unable to devote 100% to the

Table 9 .
Academic journal papers that includes wellbeing/welfare loss as a natural hazard risk metric

Table 11 .
Academic journal papers that include environmental loss as a natural hazard risk metric

Table 14 .
Practice papers that use population displacement as a natural risk metric

Table 16 .
al., 2007)FloodsThis paper conducted post-disaster interviews with more than 1,000 homes following the devastating flood that struck Central Europe in 2002.Several regions' flood mitigation efforts, post-disaster losses, and reconstruction efforts were researched.In order for further improve readiness and reaction during future flood disasters.the recovery process of Upper and Lower Ninth Ward in New Orleans one year after Hurricane Katrina.The authors used a survey to identify the extent of structural and flood damage, and post-storm recovery in the neighbourhood.The paper also identified impediments to recovery that may disproportionately affect the areas, including pre-existing inequalities and lack of resources.Florida after Hurricane Charley.The authors used multiple data sources including building permits, remotely sensed imagery and property appraiser data to evaluate the phases of housing recovery, incorporation of mitigation and effect of property sales.Practice papers that use recovery time as a natural hazard risk metric