Advertisement

Natural Hazards

, Volume 99, Issue 3, pp 1197–1213 | Cite as

Climate change impacts on socioeconomic damages from weather-related events in China

  • Xiao-Chen YuanEmail author
  • Xun Sun
Original Paper

Abstract

China is vulnerable to climate change impacts, and this study investigates the potential socioeconomic damages to China from weather-related events under future climate conditions. A two-part model incorporating a hierarchical Bayesian approach is employed to explore the effects of climate on human damage (the share of affected people in a total population) and economic damage (the share of economic losses in gross domestic product). Based on these relationships, the relative changes in socioeconomic damages under representative concentration pathways (RCPs) are presented at the regional and national levels. Our results show that China would experience an increase in socioeconomic damages from rainfall-related events under RCP2.6 and RCP4.5, and the higher increments mainly appear in the central and southwestern areas. Future climate conditions may greatly increase national damages from drought events under RCP8.5. Damages in some northern and southeastern provinces could double by 2081–2090. The national damage to humans from cold-related events is almost unchanged in most climate scenarios; however, the associated economic damage has downtrends.

Keywords

Climate change Weather-related disaster Damage Bayesian 

1 Introduction

Weather-related events interact with society and cause human and economic damages (IPCC 2012). The frequency and intensity of weather extremes may increase under future climate conditions, and the associated socioeconomic damages would thus increase (Bouwer 2013; IPCC 2013, 2014; Reyer et al. 2017; Smirnov et al. 2016). To better cope with climate change, assessing the consequences of changing concentrations of greenhouse gases (GHGs) is important. Specifically, more investigations of the damages from weather-related events in vulnerable nations are essential for calculating the costs of climate change (Revesz et al. 2014; van den Bergh and Botzen 2014).

Among the countries most frequently hit by weather-related disasters, China has suffered huge losses. During 2005–2015, China had the largest number of affected people in the world at 2.27 billion (UN/ISDR 2015). Since the country covers a large area and has differing natural and social conditions, the damages varied by region. With rising temperatures, the spatial patterns of future socioeconomic consequences are of concern to policy makers. This study focuses on the changes in regional and national human and economic damages from weather-related events due to climate change.

Recent studies show that statistical approaches are useful tools to quantify the impacts of climate change on social and economic outcomes (Auffhammer et al. 2013; Hsiang 2016; Mendelsohn et al. 2012). In a direct way, climate observations related to temperature and precipitation are taken as explanatory variables to identify the causal effects of climate on the outcomes of interest. Lee et al. (2017) suggest that annual precipitation has significantly positive effects on the damages from weather-related disasters in Korea. In addition, the frequency (Lloyd et al. 2016; Yuan et al. 2016) and intensity (Fankhauser and McDermott 2014) of extreme events are normally taken as the primary determinants for damages. Socioeconomic factors reflecting adaptive capacity also have important effects. Kahn (2005) indicates that richer nations suffer fewer losses of human life due to economic development rather than fewer disasters, and further analyses show that there are nonlinear relationships between development and damages (Kellenberg and Mobarak 2008; Zhou et al. 2014). Generally, high-income areas are more likely to have strong adaptive capacities to deal with extreme events (Noy 2009; Raschky 2008; Toya and Skidmore 2007; Wu et al. 2018).

Estimating the potential costs of climate change by establishing relationships between socioeconomic damages and the associated influencing factors is a critical task. In the literature, a hierarchical Bayesian approach has been widely employed to quantify model and parameter uncertainties and to partially pool the common information from different regions while considering heterogeneity (Gelman and Hill 2007). It could eventually help quantify climate change impacts.

Global warming has caused the changes in weather-related events to differ across China (Chen and Sun 2015); however, there is a lack of studies on the potential future socioeconomic consequences of different types of events. Historical records show that human and economic damages that have resulted from rainfall-related, drought, and cold-related events in China make up a large proportion of all events; however, there is no evidence for the effects of climate on these damages. Accordingly, this study employs a two-part model incorporating the hierarchical Bayesian approach to explore the climatic factors affecting socioeconomic damages from the three types of weather events in China. Based on these identified relationships, the changes in human and economic damages under representative concentration pathways (RCPs) are presented at the regional and national levels.

2 Methodology

2.1 Study area and data description

The study area covers 30 provinces (including municipalities and autonomous regions) of China, and their name abbreviations are shown in Table S1 (in supplementary material).

The historical meteorological data for the period 1970–2014 are taken from the China Meteorological Data Service Center. The daily time series for the selected weather stations include precipitation and mean temperature. The 1970–2090 simulated monthly precipitation and mean temperature under RCPs come from five climate models (HadGEM2-ES, IPSL-CM5A-LR, MIROC-ESM-CHEM, GFDL-ESM2M, and NorESM1-M) provided by the Inter-Sectoral Impact Model Intercomparison Project. We obtained the future climate data for weather stations based on their locations. Since the simulated and observed climate variables are not identical in distribution, their data during 1970–2000 are used to correct the distributions of simulated climate variables. At each station, the bias corrections of the temperature-related and precipitation-related variables are based on a normal distribution and a log-normal distribution, respectively (Hawkins et al. 2013). Station level data within a province are averaged for provincial analysis.

The annual direct economic losses and number of people affected due to rainfall-related (rainstorm-induced flood and waterlogging), drought, and cold-related (snow, low-temperature, and frost) disasters at the provincial level are provided by the China Meteorological Disasters Yearbooks and the National Climate Center of China, and the available data during 2002–2014 are used. The provincial economic and demographic data during 2002–2014 are collected from the China Statistical Yearbooks as well as the China Socioeconomic Development Statistical Database. The economic losses and gross domestic product (GDP) in constant 2010 Chinese Yuan (CNY) are taken for our study.

2.2 Two-part hierarchical Bayesian model

A two-part model (Duan et al. 1983) is employed for empirical analysis. In the first part, a logistic regression is used for the dichotomous event of having zero or positive damage.
$${\text{logit}}[P(y_{st} = 0)] = a_{s} + b_{1} x_{1,st} + b_{2} x_{2,st} + \cdots + b_{J} x_{J,st}$$
(1)
where yst is the socioeconomic damage (i.e., the share of affected people by total population or the share of economic losses in GDP) for the sth (s = 1,2,…, S) province in year t; \({\mathbf{x}}_{st} = (x_{1,st} ,x_{2,st} , \ldots ,x_{J,st} )\) is a set of J covariates associated with the specific damage (Table 1); bj (j = 1,2,…, J) are regression coefficients; as is the intercept for a province. Being conditional on positive damage, the log-transformed yst is modeled with a normal distribution in the second part after a preliminary diagnostic evaluation.
$$\log (y_{st} )\sim{\text{N}}(\beta_{0,s} + \beta_{1,s} x_{1,st} + \beta_{2,s} x_{2,st} + \cdots + \beta_{J,s} x_{J,st} ,\sigma_{s} )$$
(2)
where the regression coefficients \({\varvec{\upbeta}}_{s} = (\beta_{0,s} ,\beta_{1,s} ,\beta_{2,s} , \ldots ,\beta_{J,s} )\) and the covariance σs need to be estimated. Here, we further assess the spread of covariate effects across provinces with a multilevel model. A multivariate normal distribution is considered for the regression coefficients βs (Yuan et al. 2016).
$${\varvec{\upbeta}}_{s} {\sim} MVN({\varvec{\upmu}}_{\beta},{\varvec{\Sigma}}_{\beta})$$
(3)
where μβ (a vector of length J + 1) is the common mean regression coefficient for all the provinces; correspondingly, Σβ is the covariance matrix. If the estimated variances of βs (diagonal of Σβ) are large, then it effectively indicates that each province is regressed independently; by contrast, small variances imply homogeneous responses to the influencing factors (Gelman and Hill 2007). We use noninformative priors for all the parameters in the equations and employ Markov chain Monte Carlo (MCMC) sampling to estimate posterior distributions. The convergence of the MCMC chain is evaluated by the potential scale reduction factor, and all the calculations are conducted using R and RStan.
Table 1

Variables included in the models for socioeconomic damages due to weather-related events

Variable

Ln (affected people/total population)

Ln (economic losses/GDP)

Rainfall-related events

Drought events

Cold-related events

Rainfall-related events

Drought events

Cold-related events

Annual precipitation

X

X

 

X

X

 

Precipitation in cold period

  

X

  

X

Annual average temperature

X

  

X

  

Average temperature in warm period

 

X

  

X

 

Average temperature in cold period

  

X

  

X

Ln (GDP per capita)

X

X

X

X

X

X

Linear time trend

X

X

X

X

X

X

Dependent variables are affected people/total population and economic losses/GDP (both log transformed)

This study investigates the effects of climate on two kinds of damages from weather-related events. The share of affected people in a total population refers to human damage, while the share of economic losses in GDP represents economic damage. In previous studies, the number of extreme events identified from climate data is normally taken as an influencing factor. Instead, this study directly introduces climate variables into the model to reveal their effects. The selected variables are shown in Table 1, and precipitation and temperature are considered in the models. For damages from rainfall-related events, the annual precipitation that positively correlates with the number of rainfall-related events (Alexander et al. 2006) over the year is used and the annual average temperature is controlled. For damages from drought events, the annual precipitation that correlates with the number of drought events (China Meteorological Administration 2006) is taken as an influencing factor. Moreover, the average temperature in warm period that positively correlates with the number of heat events is also considered. For damages from cold-related events, the average temperature in cold period that negatively correlates with the number of cold-related events (Scherer and Diffenbaugh 2014) is chosen and the precipitation in cold period is controlled. Here, warm period indicates April to September, and cold period indicates January to March and October to December. In addition, GDP per capita that reflects adaptive capacity and a linear time trend are included in the models.

3 Results and discussion

3.1 Empirical model estimates

The estimated posterior distributions of the common mean regression coefficients for socioeconomic damages from different events are shown in Tables 2, 3, and 4, and those for each province are displayed in Figures S1-S6 (in supplementary material). A parameter whose 90% interval of the posterior distribution does not overlap with 0 is regarded to have a significant effect. In Table 2, we find that annual precipitation and average temperature have different impacts on the damages from rainfall-related events. The statistically significant coefficients imply that a unit increase in annual precipitation would result in the 0.20–0.23% and 0.26–0.29% increments in human and economic damages, respectively. However, the annual average temperature may have negative effects. For the damages from drought events, the average temperature in warm period is considered instead. Table 3 indicates that socioeconomic damages would increase with average temperature in warm period but would decrease with annual precipitation. The determinants of the damages from cold-related events are selected as the precipitation and average temperature in cold period. In general, more precipitation and a lower temperature might lead to higher human and economic damages. However, their coefficients are statistically insignificant according to Table 4.
Table 2

Estimates for socioeconomic damages from rainfall-related events

Variable

Ln (affected people/total population)

Ln (economic losses/GDP)

(1)

(2)

(3)

(4)

(5)

(6)

Annual precipitation

0.0023

[0.0016, 0.0031]

0.0020

[0.0015, 0.0027]

0.0022

[0.0016, 0.0029]

0.0029

[0.0023, 0.0036]

0.0026

[0.0020, 0.0032]

0.0026

[0.0020, 0.0032]

Annual average temperature

− 0.0821

[− 0.1568, − 0.0199]

− 0.0609

[− 0.1235, − 0.0027]

− 0.0602

[− 0.1309, − 0.0055]

− 0.1803

[− 0.2524, − 0.1098]

− 0.1575

[− 0.2128, − 0.1021]

− 0.1577

[− 0.2135, − 0.1087]

Ln (GDP per capita)

 

− 1.0499

[− 1.6718, − 0.3553]

0.5628

[− 0.7667, 1.9098]

 

− 1.5618

[− 2.1250, − 0.9636]

0.3025

[− 1.2053, 1.8185]

Ln (GDP per capita)2

  

− 0.2973

[− 0.5275, − 0.0683]

  

− 0.3342

[− 0.6021, − 0.0827]

A time trend is included in all the models. The medians are presented with the 5–95th uncertainty ranges in square brackets. Temperature is measured in °C and precipitation in millimeters

Table 3

Estimates for socioeconomic damages from drought events

Variable

Ln (affected people/total population)

Ln (economic losses/GDP)

(7)

(8)

(9)

(10)

(11)

(12)

Annual precipitation

− 0.0035

[− 0.0046, − 0.0026]

− 0.0032

[− 0.0040, − 0.0023]

− 0.0030

[− 0.0039, − 0.0022]

− 0.0040

[− 0.0051, − 0.0029]

− 0.0038

[− 0.0048, − 0.0028]

− 0.0036

[− 0.0046, − 0.0027]

Average temperature in warm period

0.2419

[0.1261, 0.3572]

0.2417

[0.1506, 0.3437]

0.2337

[0.1378, 0.3383]

0.2416

[0.1170, 0.4008]

0.2908

[0.1761, 0.4119]

0.2837

[0.1716, 0.4078]

Ln (GDP per capita)

 

− 1.7334

[− 2.4396, − 0.9814]

2.6697

[0.2862, 5.0884]

 

− 2.6805

[− 3.4933, − 1.8598]

0.9829

[− 1.7652, 3.7035]

Ln (GDP per capita)2

  

− 0.7409

[− 1.1441, − 0.3649]

  

− 0.6062

[− 1.0548, − 0.1732]

A time trend is included in all the models. The medians are presented with the 5–95th uncertainty ranges in square brackets. Temperature is measured in °C and precipitation in millimeters

Table 4

Estimates for socioeconomic damages from cold-related events

Variable

Ln (affected people/total population)

Ln (economic losses/GDP)

(13)

(14)

(15)

(16)

(17)

(18)

Precipitation in cold period

0.0016

[− 0.0016, 0.0048]

0.0023

[− 0.0001, 0.0048]

0.0021

[− 0.0005, 0.0047]

0.0002

[− 0.0028, 0.0032]

0.0001

[− 0.0024, 0.0027]

0.0001

[− 0.0025, 0.0026]

Average temperature in cold period

− 0.0293

[− 0.0971, 0.0312]

− 0.0602

[− 0.1207, − 0.0010]

− 0.0467

[− 0.1095, 0.0145]

− 0.0510

[− 0.1221, 0.0188]

− 0.0622

[− 0.1279, − 0.0021]

− 0.0542

[− 0.1198, 0.0095]

Ln (GDP per capita)

 

− 2.2608

[− 3.0739, − 1.4246]

0.6319

[− 2.2671, 4.0001]

 

− 1.7930

[− 2.4317, − 1.0871]

1.0711

[− 1.9053, 4.1188]

Ln (GDP per capita)2

  

− 0.4677

[− 0.9896, − 0.0106]

  

− 0.4791

[− 0.9429, − 0.0227]

A time trend is included in all the models. The medians are presented with the 5–95th uncertainty ranges in square brackets. Temperature is measured in °C and precipitation in millimeters

The estimates also indicate that socioeconomic damages would be reduced by income growth. As interpreted from the literature, economic development helps build adaptive capacity to mitigate damages. Nevertheless, there are no significant U-reshaped relationships between damages and economic development in most models.

3.2 Climate change impacts on regional socioeconomic damages

A baseline climate condition representing the 2005–2014 mean levels is assumed to detect climate change impacts on human and economic damages. The future climate condition is defined as the average level of climate models. The coefficients in the empirical models with climate variables only (i.e., columns 1 and 4 in Tables 2, 3, and 4) are used to estimate the relative changes (median values) under future and baseline climate conditions. Here, the results under RCP2.6 and RCP8.5 are presented for further comparisons.

3.2.1 Damages from rainfall-related events

The regional changes in human damage from rainfall-related events under RCPs are shown in Fig. 1. During 2041–2050, more damages are found in most provinces of south China. Specifically, Anhui, Chongqing, Hubei, and Jiangxi are predicted to have more than 25% increases under RCP2.6. By contrast, northeast and northwest China would have fewer damages. These decreases would expand to most provinces of north China under RCP8.5. By 2081–2090, the mean increment in southwest China, including Chongqing, Guizhou, Sichuan, and Yunnan, may reach 36% under RCP2.6. However, the mean decrement in northeast China, including Heilongjiang, Jilin, and Liaoning, would be as much as 33% under RCP8.5. Notably, Fujian and Zhejiang are expected to experience lower damages in the future. In general, there are more human damages under RCP2.6 than RCP8.5.
Fig. 1

Average changes in human damage from rainfall-related events under RCPs relative to the baseline climate condition

Figure 2 presents the regional changes in economic damage from rainfall-related events. Under RCP2.6, the increments mainly appear in the central and southwest China, and there would be more provinces of higher damage growth by 2081–2090. By comparison, the spatial pattern of damage change obviously varies in the two periods under RCP8.5. For most provinces, the economic damage would become lower by 2081–2090, and we find that north and southeast China might have more than a 25% decrease in economic damage. Furthermore, the mean change could be as much as − 32% for the entire country.
Fig. 2

Average changes in economic damage from rainfall-related events under RCPs relative to the baseline climate condition

3.2.2 Damages from drought events

The regional changes in human damage from drought events under RCPs are illustrated in Fig. 3. It shows that most provinces in north China and the coastal provinces in southeast China would have increased damages in the mid and late twenty-first century under RCP2.6. Especially for Hainan Island, the share of affected people might double with the effect of climate change. The situation for the country is predicted to be more severe under RCP8.5. During 2041–2050, there are only two provinces (Guangxi and Yunnan) that have small reductions in human damage. By 2081–2090, due to the rising temperature, all the provinces would be faced with higher damages, and their mean changes could be as much as 99%.
Fig. 3

Average changes in human damage from drought events under RCPs relative to the baseline climate condition

The changes in economic damage from drought events have similar patterns (Fig. 4). The decrements exist extensively in central and southwest China under RCP2.6. Particularly, the damage for Chongqing would reduce by 44% in 2081–2090. This is consistent with the findings on human damage under RCP8.5 that predicts higher economic damage in the future. Notably, during 2081–2090, the projected damage in the southeastern provinces (Anhui, Fujian, Jiangsu, and Zhejiang) may reach more than twice as much as that under the baseline climate condition.
Fig. 4

Average changes in economic damage from drought events under RCPs relative to the baseline climate condition

3.2.3 Damages from cold-related events

Figure 5 shows the regional changes in human damage from cold-related events. Overall, a small difference between the two periods is found under RCP2.6, which indicates that the provinces of positive change mainly cover central China, and the maximum increment is approximately 10%. With higher temperatures, the damages under RCP8.5 would reduce in most areas. Especially from 2081–2090, approximately two in five provinces would have more than a 10% decrease in damage. The northwestern area, including Inner Mongolia, Qinghai, and Xinjiang, might experience a larger decline of − 18% on average.
Fig. 5

Average changes in human damage from cold-related events under RCPs relative to the baseline climate condition

However, the economic damage from cold-related events would become lower in most provinces considering climate change impacts (Fig. 6). Jilin and Yunnan are the only two provinces showing positive changes less than 1% under RCP2.6. The mean reductions for the remaining areas are 6.0% in 2041–2050 and 5.5% in 2081–2090. In addition, we find larger decrements across the country under RCP8.5. By 2081–2090, the mean change for the provinces would be − 16%. Particularly, for some western provinces such as Gansu, Qinghai, Tibet, and Xinjiang, economic damages would decrease by more than 20%.
Fig. 6

Average changes in economic damage from cold-related events under RCPs relative to the baseline climate condition

3.3 Climate change impacts on national socioeconomic damages

According to the share of provincial population or GDP out of the total in 2015, the regional human and economic damages are aggregated. The changes in national socioeconomic damages from weather-related events under future climate conditions are displayed in Fig. 7. It shows that the human damage from rainfall-related events would rise under the RCPs, and the linear trend specifically indicates an increase of 0.42% year−1 for RCP4.5. However, the economic damage from rainfall-related events would reduce under RCP6.0 and RCP8.5 with changes of − 0.25% year−1 and − 0.27% year−1, respectively. The two kinds of damages from drought events vary greatly under the RCPs. There are very slight declines for RCP2.6 but obvious increases for other RCPs. RCP8.5 predicts potential human and economic damage increases of 119.4% and 106.3% by 2090, respectively. For the socioeconomic damages from cold-related events, there are different patterns. The share of affected people would slightly increase in general, and the larger change is found under RCP4.5. By comparison, the share of economic losses is predicted to decrease in the future. We observe that economic damages would fall 0.13% year−1 under RCP8.5.
Fig. 7

Changes in socioeconomic damage from a rainfall-related, b drought, and c cold-related events under RCPs relative to the baseline climate condition

4 Conclusions

Climate change is expected to have significant impacts on human and economic damages from weather-related events. This study focuses on the damages referring to the share of affected people in the total population and the share of economic losses in GDP and investigates the climatic factors affecting the damages from rainfall-related, drought, and cold-related events in China. In addition, the changes in damages under RCPs at the regional and national levels are presented.

Temperature and precipitation have different effects on socioeconomic damages. In general, precipitation would increase the damages from rainfall- and cold-related events but decrease those from drought events. By contrast, higher temperatures may result in lower damages from rainfall- and cold-related events but higher damages from drought events. The identified relationships between climate and damages facilitate the exploration of climate change impacts.

Under RCP2.6 and RCP4.5, China would have increased socioeconomic damage from rainfall-related events, and the higher increments mainly appear in the central and southwestern areas. However, there are different trends for human and economic damages under the other two RCPs. Future climate conditions may dramatically raise national socioeconomic damages from drought events under RCP8.5. The damages in some northern and southeastern provinces could double by 2081–2090. The national human damage from cold-related events is almost unchanged in most climate scenarios. In contrast, downtrends are found for economic damage due to the extensive decrements across the country.

Our findings provide detailed information on socioeconomic damages from different weather events caused by the effects of climate change, which is essential for making more specific adaptation strategies across regions. Additionally, the estimates are important for the development of an integrated assessment model for climate change. Notably, the absolute affected people and economic losses due to weather-related events can be further obtained based on future socioeconomic development scenarios.

Notes

Acknowledgements

The authors are grateful for financial support from the National Key R&D Program of China (2016YFA0602603) and the National Natural Science Foundation of China (NSFC) (Nos. 71704009 and 71521002). For their roles in producing, coordinating, and making available the ISI-MIP model output, we acknowledge the modeling groups (HadGEM2-ES, IPSL-CM5A-LR, MIROC-ESM-CHEM, GFDL-ESM2 M, and NorESM1-M) and the ISI-MIP coordination team. We also appreciate the anonymous reviewers and the editor for their insightful and constructive comments that substantially improved the manuscript.

Supplementary material

11069_2019_3588_MOESM1_ESM.docx (2.4 mb)
Supplementary material 1 (DOCX 2496 kb)

References

  1. Alexander LV, Zhang X, Peterson TC, Caesar J, Gleason B, Tank AMGK, Haylock M, Collins D, Trewin B, Rahimzadeh F, Tagipour A, Kumar KR, Revadekar J, Griffiths G, Vincent L, Stephenson DB, Burn J, Aguilar E, Brunet M, Taylor M, New M, Zhai P, Rusticucci M, Vazquez-Aguirre JL (2006) Global observed changes in daily climate extremes of temperature and precipitation. J Geophys Res Atmos 111:D05109Google Scholar
  2. Auffhammer M, Hsiang SM, Schlenker W, Sobel A (2013) Using weather data and climate model output in economic analyses of climate change. Rev Environ Econ Policy 7:181–198CrossRefGoogle Scholar
  3. Bouwer LM (2013) Projections of future extreme weather losses under changes in climate and exposure. Risk Anal 33:915–930CrossRefGoogle Scholar
  4. Chen HP, Sun JQ (2015) Changes in climate extreme events in China associated with warming. Int J Climatol 35:2735–2751CrossRefGoogle Scholar
  5. China Meteorological Administration (2006) Classification of meteorological drought. Standards Press of China, BeijingGoogle Scholar
  6. Duan N, Manning WG, Morris CN, Newhouse JP (1983) A comparison of alternative models for the demand for medical care. J Bus Econ Stat 1:115–126Google Scholar
  7. Fankhauser S, McDermott TKJ (2014) Understanding the adaptation deficit: why are poor countries more vulnerable to climate events than rich countries? Glob Environ Change 27:9–18CrossRefGoogle Scholar
  8. Gelman A, Hill J (2007) Data analysis using regression and multilevel/hierarchical models. Cambridge University Press, New YorkGoogle Scholar
  9. Hawkins E, Osborne TM, Ho CK, Challinor AJ (2013) Calibration and bias correction of climate projections for crop modelling: an idealised case study over Europe. Agric For Meteorol 170:19–31CrossRefGoogle Scholar
  10. Hsiang S (2016) Climate Econometrics. Annu. Rev. Resour. Econ 8:43–75CrossRefGoogle Scholar
  11. IPCC (2012) Managing the risks of extreme events and disasters to advance climate change adaptation. Cambridge University Press, Cambridge and New YorkGoogle Scholar
  12. IPCC (2013) Climate Change 2013: the phyisical science basis. Cambridge University Press, CambridgeGoogle Scholar
  13. IPCC (2014) Climate Change 2014: impacts, adaptation, and vulnerability. Part A: global and sectoral aspects. Cambridge University Press, CambridgeGoogle Scholar
  14. Kahn ME (2005) The death toll from natural disasters: the role of income, geography, and institutions. Rev Econ Stat 87:271–284CrossRefGoogle Scholar
  15. Kellenberg DK, Mobarak AM (2008) Does rising income increase or decrease damage risk from natural disasters? J Urban Econ 63:788–802CrossRefGoogle Scholar
  16. Lee M, Hong JH, Kim KY (2017) Estimating damage costs from natural disasters in Korea. Nat Hazards Rev 18:04017016CrossRefGoogle Scholar
  17. Lloyd SJ, Kovats RS, Chalabi Z, Brown S, Nicholls RJ (2016) Modelling the influences of climate change-associated sea-level rise and socioeconomic development on future storm surge mortality. Clim Change 134:441–455CrossRefGoogle Scholar
  18. Mendelsohn R, Emanuel K, Chonabayashi S, Bakkensen L (2012) The impact of climate change on global tropical cyclone damage. Nat Clim Change 2:205–209CrossRefGoogle Scholar
  19. Noy I (2009) The macroeconomic consequences of disasters. J Dev Econ 88:221–231CrossRefGoogle Scholar
  20. Raschky PA (2008) Institutions and the losses from natural disasters. Nat Hazards Earth Syst Sci 8:627–634CrossRefGoogle Scholar
  21. Revesz RL, Howard PH, Arrow K, Goulder LH, Kopp RE, Livermore MA, Oppenheimer M, Sterner T (2014) Improve economic models of climate change. Nature 508:173–175CrossRefGoogle Scholar
  22. Reyer CPO, Otto IM, Adams S, Albrecht T, Baarsch F, Cartsburg M, Coumou D, Eden A, Ludi E, Marcus R, Mengel M, Mosello B, Robinson A, Schleussner CF, Serdeczny O, Stagl J (2017) Climate change impacts in Central Asia and their implications for development. Reg Environ Change 17:1639–1650CrossRefGoogle Scholar
  23. Scherer M, Diffenbaugh NS (2014) Transient twenty-first century changes in daily-scale temperature extremes in the United States. Clim Dyn 42:1383–1404CrossRefGoogle Scholar
  24. Smirnov O, Zhang MH, Xiao TY, Orbell J, Lobben A, Gordon J (2016) The relative importance of climate change and population growth for exposure to future extreme droughts. Clim Change 138:41–53CrossRefGoogle Scholar
  25. Toya H, Skidmore M (2007) Economic development and the impacts of natural disasters. Econ Lett 94:20–25CrossRefGoogle Scholar
  26. UN/ISDR (2015) The human cost of weather related disasters 1995–2015Google Scholar
  27. van den Bergh J, Botzen WJW (2014) A lower bound to the social cost of CO2 emissions. Nat Clim Change 4:253–258CrossRefGoogle Scholar
  28. Wu JD, Han GY, Zhou HJ, Li N (2018) Economic development and declining vulnerability to climate-related disasters in China. Environ Res Lett 13:034013CrossRefGoogle Scholar
  29. Yuan XC, Sun X, Lall U, Mi ZF, He J, Wei YM (2016) China’s socioeconomic risk from extreme events in a changing climate: a hierarchical Bayesian model. Clim Change 139:169–181CrossRefGoogle Scholar
  30. Zhou Y, Li N, Wu WX, Liu HL, Wang L, Liu GX, Wu JD (2014) Socioeconomic development and the impact of natural disasters: some empirical evidences from China. Nat Hazards 74:541–554CrossRefGoogle Scholar

Copyright information

© Springer Nature B.V. 2019

Authors and Affiliations

  1. 1.School of Management and EconomicsBeijing Institute of TechnologyBeijingChina
  2. 2.Center for Energy and Environmental Policy ResearchBeijing Institute of TechnologyBeijingChina
  3. 3.Key Laboratory of Geographic Information Science (Ministry of Education)East China Normal UniversityShanghaiChina
  4. 4.Columbia Water Center, Earth InstituteColumbia UniversityNew YorkUSA

Personalised recommendations