Elsevier

Journal of Hydrology

Volume 602, November 2021, 126731
Journal of Hydrology

Linking temperature to catastrophe damages from hydrologic and meteorological extremes

https://doi.org/10.1016/j.jhydrol.2021.126731Get rights and content

Highlights

  • The magnitude of storm and flood related catastrophes is expected to increase.

  • We find a positive association between economic losses and local temperature.

  • Positive economic loss-temperature associations are consistent with climate change.

Abstract

Increases in the magnitude of storm and flood related catastrophes due to climate change are predicted to increase associated economic losses. There exists, however, conflicting evidence for greater economic losses despite well acknowledged increases in the severity of observed extreme events in recent decades. Here, using a worldwide catastrophe insurance database from 1970 to 2015, we link the catastrophe economic loss from extreme storms and floods to local temperature. We find a statistically significant positive association between economic losses expressed as a proportion of GDP and local temperature. The association between economic losses and temperature is greater as the event becomes more extreme, with the signal muted for flooding as compared to storms. Although local associations of economic loss with temperature cannot be directly linked to rising global temperatures as a result of climate change, the positive economic loss-temperature associations are consistent with observed extreme precipitation-temperature associations, and hence pertinent to the advancement of understanding future natural catastrophes.

Introduction

Storm related climate catastrophes result in major loss of life and widespread economic damage, with developing countries particularly exposed as they lack the financial and material resources to mitigate impacts (Razavi et al., 2020). Overall, the economic losses from these catastrophes are expected to be exacerbated by climatic change (Arent et al., 2014, IPCC, 2012). This expectation is supported by evidence from dynamic downscaling of individual weather events and statistical downscaling of global climate models, both of which predict increased insurance losses can be expected as we head into the significantly warmer temperatures over the rest of this century (Held et al., 2013).

The contribution of climate change to economic loss potential from storm events is largely based on the assumption of increasing storm intensity driven via the Clausius-Clapeyron relation. Assuming constant relative humidity, this physical relationship dictates that as temperatures increase, so does the moisture holding capacity of the atmosphere, and hence extreme precipitation intensity should increase as a result (Trenberth, 2011, Trenberth et al., 2003). This in turn is predicted to lead to increased flood risk (Trenberth et al., 2003, Westra et al., 2014), particularly in urban settings (Fadhel et al., 2018, Hettiarachchi et al., 2018, Miller and Hutchins, 2017). Increases in precipitation intensity are likely to be coupled with changes in storm temporal pattern (Wasko and Sharma, 2015) as well as duration (Emmanuel et al., 2012, Webster, 2005) and areal extent (Prein et al., 2017). Storms may also invigorate due to increased latent heat release at higher (absolute) humidity intensifying upward motions (Lenderink and van Meijgaard, 2008, Trenberth et al., 2003), or change due to larger atmospheric changes such at the expansion of the Hadley Cell (Grise et al., 2018, Mathew and Kumar, 2019, Seidel et al., 2008, Staten et al., 2018) increasing the frequency of extreme precipitation events (Myhre et al., 2019). All these factors are predicted to lead to increased storm severity with climate change (Seneviratne et al., 2012).

Observed increases in extreme precipitation, flooding, and storm severity are abound in literature (Do et al., 2017, Donat et al., 2013, Emanuel, 2005, Martinez‐Villalobos and Neelin, 2018, O’Gorman, 2015, Slater et al., 2021, Sun et al., 2021, Wasko and Nathan, 2019, Westra et al., 2013a). While historical changes in climate extremes have been attributed to anthropogenic climate change (Diffenbaugh et al., 2017, Kay et al., 2011, Min et al., 2011, Pall et al., 2011), the linking of economic losses to climatic change is more difficult (Bouwer, 2011, Hoeppe, 2016). Changes in economic losses have been associated with local climate variability (Pielke and Landsea, 1999, Welker and Faust, 2013), and the economic cost of hurricane Harvey has recently been attributed to climate change (Frame et al., 2020), but debate remains on whether temporal trends in economic loss exist (Paprotny et al., 2018). Although a summary of historical studies finds more studies observe increases rather than decreases in economic loss (Bouwer, 2011), after appropriate normalisation for increases in economic growth, studies generally find little temporal trend and hence association with climate change (Arent et al., 2014). For example, normalised total annual losses from Hurricanes in the United States (Pielke et al., 2008, Weinkle et al., 2018) and Latin America and The Caribbean (Pielke et al., 2003) show no temporal trend. Likewise no climate change signal was identified in normalised annual aggregations of flood losses across Europe (Barredo, 2009).

Global normalised economic loss aggregations across various natural disaster types also display no statistically significant upward trend (Barthel and Neumayer, 2012, Neumayer and Barthel, 2011), with continental Australian studies showing similar results (Crompton and McAneney, 2008, McAneney et al., 2019). As global temperatures have been monotonically increasing and can be substituted for a temporal trend, there is likewise no link found between increasing global temperatures and increasing global economic losses (Miller et al., 2008). The two primary arguments for differing trends in economic loss are the method of normalisation of economic loss (Barthel and Neumayer, 2012, Kron et al., 2019) and the absence of consideration of changes in vulnerability (Mechler and Bouwer, 2015). For example, climate change signals may be obscured by increasing defences and resilience to climatic catastrophes (Nicholls, 2011) meaning increases in economic loss are not as large as they would have been had vulnerability-reducing measures not been implemented (Mechler and Bouwer, 2015). Here we suggest another possibility. As global warming does not always affect the global temperature uniformly (Neukom et al., 2019) it may be that higher global temperatures will not be associated with greater economic losses, but local temperatures will.

There is significant evidence that local temperature influences storm severity (Lenderink and van Meijgaard, 2008, Peleg et al., 2018). Changes in both tropical cyclones and local scale convective events have been correlated to increases in temperature (Lenderink and Attema, 2015, Webster, 2005). Increased lightning strikes have been linked to higher temperatures (Molnar et al., 2015) and a greater number of insurance claims (Mills, 2005). Indeed, local temperature sensitivities have been shown to be the primary variable explaining rainfall variability across various temporal scales (Wasko and Sharma, 2017a, Westra et al., 2013b) with higher temperatures linked to greater flood event magnitude (Wasko and Sharma, 2017b). Local temperature increases have been successfully used as a predictor in non-stationary analysis of flood risk (Condon et al., 2015, Towler et al., 2010) and extreme rainfall (Agilan and Umamahesh, 2015, Mondal and Mujumdar, 2015), even being linked to indicators of water quality (Guo et al., 2021) and hence health outcomes (Jhajharia et al., 2013).

As discussed above, storm extremes are linked to higher temperatures from the increased moisture carrying capacity of a warmer atmosphere as specified by the Clausius Clapeyron relationship (Roderick et al., 2019, Trenberth, 2011). Assuming constant relative humidity, this increase in moisture carrying capacity can lead to larger downpours especially over short periods of time where local moisture contributes substantially (Fowler et al., 2021). While this physical link to temperature becomes weaker when the relationship is extended to other derived variables such as floods (Sharma et al., 2018, Tramblay et al., 2019, Wasko, 2021) and economic loss (Nicholls, 2011), a positive link with extreme precipitation is well established. Hence, here we seek to investigate, can a link between temperature and the economic loss resulting from extreme hydrologic and meteorological events be identified?

We analyse natural disasters on a per event basis, rather than a total or annual economic loss, to capture the event severity (in terms of economic loss). The normalised temporal trend on a per catastrophe basis is first analysed to see if economic losses have increased due to an increase in event severity. Next, because global temperature anomalies are not necessarily spatially uniform, the sensitivity of normalised economic loss to local temperature as a driver of storm intensification is investigated. The influence of storm severity and the originating storm mechanism on the relationship of normalised economic loss to local temperature are also investigated. Finally, the implications in the context of climate change are discussed.

Section snippets

Data

In this study we use the Sigma catastrophe database for the years 1970 to 2015 (SwissRe, 2016). Sigma represents a comprehensive database of significant catastrophe events collated from newspapers, direct and reinsurance periodicals, specialist publications and reports from (re)insurers. This data set provides information on 5024 natural catastrophes for which various statistics are reported including fatalities, economic loss, and insured loss. Here, we focus on the economic loss. The economic

Normalisation of individual catastrophes

We analyse economic loss on a per catastrophe basis to isolate possible increases in storm severity. The effect of normalisation on economic loss is discussed in detail in Neumayer and Barthel (2011). For spatial pooling, for a given catastrophe year (t), normalising by the country wealth is recommended (Neumayer and Barthel, 2011):yi=NormalisedEconomicLossi=EconomicLossitWealtht

This results in a sample of normalised economic loss yi where i = 1..n and n is the total number of events (as

Temporal trend

A time series of the normalised economic loss per catastrophe per year is presented in Fig. 2a. For simplicity, from here on, economic loss refers to the economic loss standardised by GDP. A surface of Kendall's tau statistic for monotonic trend is presented in Fig. 2b. Statistical significance is presented in Fig. 2c. The trend test is performed on a moving window of years, with a minimum of 10 years required for testing. Economic loss per catastrophe is seen to increase only if the data

Discussion

Historical trends in economic loss are often investigated, and the trend attributed to climatic change due to the historically increasing greenhouse gasses increasing temperatures and changing atmospheric circulations. Hence it is common practice, and entirely valid, to substitute historical global temperature anomalies for the temporal trend as both are strongly correlated. But this practice is not necessarily informative of changes to damage due to changed storm intensity, as storm intensity

Conclusions

Previous studies that have attempted to link economic loss with temperature anomalies at a global scale did not find a historical trend in economic loss. Here, we investigated the sensitivity (association) of catastrophe loss with local temperature on a per catastrophe basis. Consistent with existing literature, we found little evidence for increased economic loss in time due to storms increasing in severity. However, we did find higher local temperatures are associated with greater economic

CRediT authorship contribution statement

Conrad Wasko: Methodology, Formal analysis, Writing - original draft, Writing - review & editing. Ashish Sharma: Conceptualization, Methodology, Writing - review & editing. Alexander Pui: Conceptualization, Resources, Writing - review & editing.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

Conrad Wasko receives funding from the University of Melbourne McKenzie Postdoctoral Fellowship scheme. The authors acknowledge funding from the Australian Research Council (DE210100479, DP200101326).

References (103)

  • A.L. Kay et al.

    Attribution of Autumn/Winter 2000 flood risk in England to anthropogenic climate change: A catchment-based study

    J. Hydrol.

    (2011)
  • W. Kron et al.

    Reduction of flood risk in Europe – Reflections from a reinsurance perspective

    J. Hydrol.

    (2019)
  • S.S. Mathew et al.

    Characterization of the long-term changes in moisture, clouds and precipitation in the ascending and descending branches of the Hadley Circulation

    J. Hydrol.

    (2019)
  • J.D. Miller et al.

    The impacts of urbanisation and climate change on urban flooding and urban water quality: A review of the evidence concerning the United Kingdom

    J. Hydrol. Reg. Stud.

    (2017)
  • A. Mondal et al.

    Modeling non-stationarity in intensity, duration and frequency of extreme rainfall over India

    J. Hydrol.

    (2015)
  • E. Neumayer et al.

    Normalizing economic loss from natural disasters: A global analysis

    Glob. Environ. Chang.

    (2011)
  • C. Wasko et al.

    Influence of changes in rainfall and soil moisture on trends in flooding

    J. Hydrol.

    (2019)
  • C. Wasko et al.

    Continuous rainfall generation for a warmer climate using observed temperature sensitivities

    J. Hydrol.

    (2017)
  • S. Westra et al.

    A conditional disaggregation algorithm for generating fine time-scale rainfall data in a warmer climate

    J. Hydrol.

    (2013)
  • Arent, D.J., Tol, R.S.J., Faust, E., Hella, J.P., Kumar, S., Strzepek, K.M., Tóth, F.L., Yan, D., 2014. Key economic...
  • J.I. Barredo

    Normalised flood losses in Europe: 1970–2006

    Nat. Hazards Earth Syst. Sci.

    (2009)
  • F. Barthel et al.

    A trend analysis of normalized insured damage from natural disasters

    Clim. Change

    (2012)
  • P. Berg et al.

    Strong increase in convective precipitation in response to higher temperatures

    Nat. Geosci.

    (2013)
  • Berkeley Earth, 2015. Berkeley Earth Project. [available at...
  • G. Blöschl et al.

    Changing climate shifts timing of European floods

    Science (80-.

    (2017)
  • L.M. Bouwer

    Have Disaster Losses Increased Due to Anthropogenic Climate Change?

    Bull. Am. Meteorol. Soc.

    (2011)
  • L.E. Condon et al.

    Climate change and non-stationary flood risk for the upper Truckee River basin

    Hydrol. Earth Syst. Sci.

    (2015)
  • A.S. Daloz et al.

    Is the poleward migration of tropical cyclone maximum intensity associated with a poleward migration of tropical cyclone genesis?

    Clim. Dyn.

    (2018)
  • N.S. Diffenbaugh et al.

    Quantifying the influence of global warming on unprecedented extreme climate events

    Proc. Natl. Acad. Sci.

    (2017)
  • M.G. Donat et al.

    Updated analyses of temperature and precipitation extreme indices since the beginning of the twentieth century: The HadEX2 dataset

    J. Geophys. Res. Atmos.

    (2013)
  • K. Emanuel

    Increasing destructiveness of tropical cyclones over the past 30 years

    Nature

    (2005)
  • H.J. Fowler et al.

    Anthropogenic intensification of short-duration rainfall extremes

    Nat. Rev. Earth Environ.

    (2021)
  • D.J. Frame et al.

    The economic costs of Hurricane Harvey attributable to climate change

    Clim. Change

    (2020)
  • L. Gao et al.

    A Global Data Set for Economic Losses of Extreme Hydrological Events During 1960–2014

    Water Resour. Res.

    (2019)
  • K.M. Grise et al.

    Regional and seasonal characteristics of the recent expansion of the tropics

    J. Clim.

    (2018)
  • J.O. Haerter et al.

    Unexpected rise in extreme precipitation caused by a shift in rain type?

    Nat. Geosci.

    (2009)
  • J. Hall et al.

    Understanding flood regime changes in Europe: a state-of-the-art assessment

    Hydrol. Earth Syst. Sci.

    (2014)
  • H. Held et al.

    Projections of global warming-induced impacts on winter storm losses in the German private household sector

    Clim. Change

    (2013)
  • S. Hettiarachchi et al.

    Increase in flood risk resulting from climate change in a developed urban watershed – the role of storm temporal patterns

    Hydrol. Earth Syst. Sci.

    (2018)
  • IPCC Managing the Risks of Extreme Events and Disasters to Advance Climate Change Adaptation. A Special Report of...
  • T.J. Ivancic et al.

    Examining why trends in very heavy precipitation should not be mistaken for trends in very high river discharge

    Clim. Change

    (2015)
  • D. Jhajharia et al.

    Influence of climate on incidences of malaria in the Thar Desert, northwest India

    Int. J. Climatol.

    (2013)
  • E. Kalnay et al.

    The NCEP/NCAR 40-year reanalysis project

    Bull. Am. Meteorol. Soc.

    (1996)
  • M. Kendall

    Rank correlation methods, 2nd

    (1975)
  • M.G. Kendall

    A New Measure of Rank Correlation

    Biometrika

    (1938)
  • Koenker, R., 2021. quantreg: Quantile Regression. R package version 5.85. [available at...
  • R. Koenker et al.

    Regression Quantiles

    Econometrica

    (1978)
  • G. Lenderink et al.

    A simple scaling approach to produce climate scenarios of local precipitation extremes for the Netherlands

    Environ. Res. Lett.

    (2015)
  • G. Lenderink et al.

    Increase in hourly precipitation extremes beyond expectations from temperature changes

    Nat. Geosci.

    (2008)
  • C. Li et al.

    A Three-Parameter S-Shaped Function of Flood Return Period and Damage

    Adv. Meteorol.

    (2016)
  • Cited by (15)

    • Rethinking urban storm water management through resilience – The case for using green infrastructure in our warming world

      2022, Cities
      Citation Excerpt :

      Flooding, particularly in developed high density human inhabited spaces, ranks among the most life-threatening disasters we face on earth (Wasko et al., 2021).

    View all citing articles on Scopus
    View full text