A Framework To Estimate Loss And Damage (L&D) To Floods In a Changing Climate, Using Capital-Based Approach


 Although there exist many mitigation and adaptation responses for disaster management, with specific reference to floods, still there are some residual impacts which arise from insufficient mitigation and inadequate adaptation, known as loss and damage (L&D). Literature has identified familiar methodologies, namely Damage and Loss Assessment, Post Disaster Needs Assessment, Catastrophe Simulation, Hazus-MH, and econometric models to quantify “loss” and “damage” separately, but, given that the starting point of L&D, being the livelihood vulnerability context of the individuals, there is a need for comprehensive approach to quantify residual “loss and damage” on the livelihood assets. This study has adopted a bottom-up approach, Sustainable Livelihood Framework and revised to provide the quantitative estimates of loss and damage of floods to the five types of capital assets (human, social, physical, financial, natural), at the individual level. A panel regression model can be used to determine the losses inflicted to each capital asset, which can be further used to derive first and second order loss and damage estimates for identifying room(s) for flood recovery interventions. A set of individual specific and community specific characteristics play an important role in determining the influence of floods on each indicator of each type of capital asset. However, this model can find differential impacts of different types of disaster in different regions across different individuals, depending on their exposure.


77
A household's potential for L&D depends on his/her mitigation efforts, livelihood context, in forms of capital 78 (social, economic, natural, human, and political), his/her vulnerability profile, and a household's coping and 79 adaptive capacity (Warner & Geest, 2013). It occurs due to potential four reasons (Warner & Geest, 2013):     3. Erosive coping measures: Households who are exposed to areas prone to climate risks often employ 90 coping measures which are "erosive" in nature, if the adopted proactive adaptation measures to avoid the 91 adverse effects are not successful. They are erosive, in the sense, that they may prove advantageous in the 92 short term by avoiding the adverse effects, but they eventually pose detrimental effects to their future  4. Measures not adopted: There are also many households, who in-spite of being vulnerable to the adverse 96 effects of sudden or slow onset events, do not employ adaptation or coping measures, due tomainly: lack of 97 knowledge required for adaptation or are devoid of skills; too expensive measures for the households to 98 adopt, as they are constrained by financial resources and means (Monnereau & Abraham, 2013); or do not 99 feel it to be their task to adopt measures to avoid effects; or do not prioritize adaptation or coping.
3 Being an uncertain event, it is assumed that the flood loss is also a random variable with a normal distribution, as 101 shown in Fig. 3. The mean and variance of this distribution of flood losses are the expected value, and the cost of 102 risk taken by the residents in floodplains (Thampapillai & Musgrave, 1985). To minimize the likelihood of 103 occurrence of floods, and their consequences, individuals are endowed with a wide portfolio of adaptation and 104 mitigation measures. Some of these measures are structural (or hard) and non-structural (or soft) measures, where 105 the former includes reservoirs, levees, flood walls, channel improvements, sea dikes, storm-surge barriers, and 106 floodways; and the latter includes land-use planning of flood plains, warning and evacuation, wet proofing, dry 108 Dua et al., 2020). Although these measures can reduce the expected value of flood losses from μ0 to μ1 (μ1< μ0), and 109 variance from σ 0 2 to σ 1 2 (σ 1 2 < σ 0 2 ), there are still some residual flood losses, which is termed as L&D, the estimation 110 of which is pertinent, to identify the potential room for recovery interventions.

Catastrophic Simulation (CatSim) 186
The government or public authority may face a dearth of funds for financing post disaster relief and recovery, 187 known as financial vulnerability, which mainly, has two components (Mechler et al., 2006): (a) asset risks from 188 natural hazards (measured by the hazard frequency and intensity, exposure and sensitivity); (b) financial resilience 189 or preparedness, i.e. the financial ability of the authority to cope with the disaster losses, with its available resources.

Financial Capital 300
The direct tangible impacts of floods are the loss in income and wages, whereas, the indirect impacts include the 301 reduced consumption of goods and services. We can model the impact of floods on financial capital, using the 302 following random effects regression equation 4, of individual i, commune j, and year t: 303 ln(FC ijt ) = β 0 + β 1 X ijt + β 2 C jt + β 3 ln(FC ijt−1 ) + β 4 F jt + β 5 D jt + β 6 F jt X ijt + β 7 F jt C jt + θ i + δ j + ϵ ijt …. (4)

304
Here, the dependent variable is the financial capital, proxied by per capita income, per capita expenditure, per capita 305 savings rate, and private credit per capita. X ijt is a set of individual specific explanatory variables, like age of

322
, where NC ijt is the natural capital, proxied by availability of agricultural land, water resources, trees and forest 323 products. X ijt is a vector of explanatory variables, including share of cropland area, ownership of land and boat, 324 income, etc. The commune characteristics, given by C jt entail access to irrigation or water facility, access to road 325 network, soil conditions, number of woodland areas or rainforests, social capital, etc.

326
In all the equations, from (1) to (5), jt is a flood dummy, i.e. whether affected households or not; and D jt is a vector

Conclusion 401
With an increasing frequency and the intensity of the floods, due to climate change, the increasing flood risks have 402 become an increasing concern, as it has profound adverse effects on the economy. Although, mitigation and 403 adaptation are the major two climate actions taken to tackle the problem, there is always a residual impact, which is 404 known as the loss and damage. Section 2 introduces us to the concept of loss and damage, and discusses that L&D 405 occurs mainly due to four major reasons-insufficient adaptation or coping measures, irrecoverable costs, erosive 406 coping measures, and measures not being adopted. This section also takes us along the journey of how this concept 407 of L&D has evolved, through various COPs, followed by an understanding of its relevance in context of floods.

Limitations and future scope of the study 430
This framework can be used to conduct L&D assessment to various sudden onset meteorological, hydrological, and 431 climatological disasters; and slow onset events, in the changing climate. However, this methodology has limitations.

432
Firstly, such a framework, with the interaction between the disaster and the explanatory variables in determining the 433 losses to the capital assets are contextual in nature, i.e. differential impact of different types of disaster in different 10 regions may be observed. Hence, the policies must be tailored according to the type of disaster and the region or 435 community contexts. Secondly, this framework could not incorporate the role of climate-risk perception of 436 individuals, which plays a pivotal role. Thirdly, the depth of flooding is not taken into account as a proxy for flood, 437 in the framework. Finally, it could not distinguish whether the adaptation measures were based on top-down or 438 bottom-up planning, in view of reducing disparity between adaptation policy needs at national and local level.

Conflict of Interest/Competing interests 440
The authors declare that they have no known competing financial interests or personal relationships that could have 441 appeared to impact the work reported in this paper.