Elsevier

Earth-Science Reviews

Volume 210, November 2020, 103345
Earth-Science Reviews

Invited Review
The triple complexity of drought risk analysis and its visualisation via mapping: a review across scales and sectors

https://doi.org/10.1016/j.earscirev.2020.103345Get rights and content

Abstract

Drought risk quantifies the likelihood of environmental and socio-economic systems to suffer negative consequences due to drought. The analysis of drought risk is understood as the initial step to foster drought risk management at different scales and thematic foci. To communicate drought risk to stakeholders, a visualisation via risk maps is of preferable usage. Even though drought risk has been subject to investigations worldwide, a systematic review of current drought risk mapping practice is lacking.

This work reviews recent practices of drought risk analysis with the aim of detangling issues of scale and focus of application, applied data, approaches to analyse risk, and drought risk mapping as generic visualisation technique. The review highlights a general lack of national and regional studies, especially in Central America as well as northern and south-eastern Asia. A majority of studies focuses specific topics, especially agriculture, but also general drought risk analyses are popular. From a methodological point of view, drought risk is based on different kinds of information: hazard, vulnerability and impact information; and combined in statistical or conceptual models. Hereby, a lack of statistical evidence for predictor selection as well as of verification of final results became evident. The majority of studies bases analysis on predictors on expert knowledge, whereas a proven suitability of the predictor to analyse the asset at risk remains questionable. The analysis of existing studies suggests that there is only little agreement on what a drought risk analysis entails. The variety of research aims and thematic foci results in a wide spectrum of drought hazard predictors and vulnerability factors applied as well as ways to analyse drought risk. Nevertheless, common paradigms of data application and combination practice were identified and resulted in the overall recommendation on considering the components of hazard predictors, vulnerability and impact information in the frame of a statistical model. Moreover, information on common practice of applied methodological approaches and data with regard to the location, spatial resolution and thematic focus of the study is provided. Furthermore, efforts to map drought risk differ widely with respect to purpose, intelligibility, and implementation details such as categorical visualisation. In order to increase intelligibility of prospective drought risk maps, basic standards have been recommended.

Introduction

Drought is omnipresent and affects a larger proportion of the world-wide population than any other natural disaster (UN-ISDR, 2009). Neither the beginning nor the end of drought can precisely be defined, it develops creepingly, may occur detached from climate’s seasonality and may persist lastingly (Wilhite and Vanyarkho, 2000, Smakhtin and Schipper, 2008; Sheffield and Wood, 2012). Drought is commonly communicated as a relative concept by means of environmental variables which characterise the physical aspects of drought as a hazard, e.g. as a deficit over time or deviation from normal (Smakhtin and Schipper, 2008; Eriyagama et al., 2009; Zargar et al., 2011, Logar and van den Bergh, 2013, Lloyd-Hughes, 2014). Hence, drought is a stochastic but complex natural phenomenon resisting scientific agreement about a precise definition (Mishra and Singh, 2010). The fundamental categorisation of drought as meteorological, agricultural, hydrological drought by Wilhite and Glantz (1985) was intended by a connection of physical aspects of drought and its implication for a specific sector. The categorisation scheme was refined by the American Meteorological Society (1997) by adding a human perspective introducing socio-economic drought. The definition of drought as a cascading hazard has thus been complemented by the impacts which can occur at all stages of the hazard. The diversification of definitions for drought did not stop at this stage and e.g. vegetation drought (Helldén and Eklundh, 1988), hydrological drought typologies (Van Loon and van Lanen, 2012), groundwater droughts (Mishra and Singh, 2010), or flash droughts (Christian et al., 2019) were introduced. Today’s understanding of drought types is defined by the physical aspects of the propagating hazard (e.g. soil moisture drought) rather than the potentially impacted sector (e.g. agricultural drought) (UNDRR, 2019). Drought impacts are particularly emphasised as consequences of drought the hazards (Van Loon et al., 2016V). Overall, existing drought definitions are governed by three core motives (Karavitis, 1999; Herceg, 2012):

a) The disciplinary conventions of investigation

b) The geographical, hydrological, geological, historical and cultural traits of a given locale

c) The difficulty to modify existing drought terminology according to updated techniques and practices.

In 1999, Karavitis recommended an operational approach, classifying drought by means of its impacts on hydrological, environmental and socio-economic dimensions. Fifteen years later, Lloyd-Hughes (2014) claimed the prevailing impracticability of universal drought definitions and initiated a paradigm shift to define drought by its impacts, which for example Blauhut et al. (2015a) or Bachmair et al. (2015) took up on. Drought can impact all systems across a range of scales and sectors: Direct or indirect, tangible or intangible, negative or positive (Wilhite, 2001; Wilhite et al., 2007; Hannaford, 2010; Stahl et al., 2016). Consequently a variety of impact categorisation schemes exist (e.g. EC, 2007; Stölzle and Stahl, 2011; UNCCD, 2012; NDMC, 2013; Schwab, 2013; WMO--GWP, 2014, Stahl et al., 2016) which can be generalised into three main categories: economic, environmental and social impacts (Wilhite and Glantz, 1985; Kossida et al., 2009). Most often classified impact categories (or sectors) herein are agriculture, energy, industry, fishery, forestry, fire, conservation, water management, tourism and recreation, and public health (Stölzle and Stahl, 2011; Schwab, 2013; NDMC, 2013; Stahl et al., 2016; EC, 2007; UNCCD, 2012; WMO--GWP, 2014). More detailed impact information are pooled by the European Drought Impact Report Inventory database were 15 impact categories are distinguished to over 100 different impact types (Stahl et al., 2016).

The likelihood of occurrence of drought impacts is defined as drought risk, which in a general sense is a function of hazard and vulnerability (Blauhut et al., 2015a). But the precise definition of drought risk and correspondent terminologies, data and combination methods differ context-specifically (Brooks, 2003). Gonzales Tanago et al. (2016) detected that “…the diversity of scope, approach, focus, methodology, and measurement criteria hampers a common understanding of vulnerability to drought…”. In order to globally reduce disaster, the UN-International Strategy for Disaster Risk Reduction released guidance on “Drought Risk Reduction Framework and Practices” (UN-ISDR, 2009). The authors reviewed current thinking and practice, and elaborated a guiding framework to understand drought risk and its components. In 2019 the UN published the GAR report (UNDRR, 2019) with a special section on drought, presenting examples on the assessment of drought risk and its components for different scales and sectors. Nevertheless, a comparison of approaches and detailed guidance was still lacking. According to Blauhut et al. (2016), the key differentiator between drought risk analysis approaches is the use/non-use of drought impact and/or vulnerability information, herein defined as impact-, factor- or hybrid approach. Gonzales Tanago et al. (2016) and Naumann et al. (2019) define these, based on the differing foci of Disaster Risk Reduction and the Climate Change Adaption community, as outcome (impact) or conceptual (factor) approach. In a nutshell, impact/outcome approaches apply impact information in order to quantify the negative effects; factor/conceptual approaches combine intrinsic socio-economic and environmental factors that define the vulnerability in order to identify corresponding drivers; hybrid approaches apply both kinds of information (Blauhut et al., 2016; Vogt et al., 2018). The work of Hagenlocher et al. (2019) is the first systematic review on drought risk assessment elaborating the state of the art and persisting gaps with respect to drought risk reduction measures. Despite the documented boost in drought risk research, they detected numerous gaps in current drought risk research such as a lack of consistency in applied definitions between drought risk analyses, a lack of precise definitions of the components applied in drought risk analyses, a lack of result validation and lacking resultant recommendations for risk reduction. Future drought risk research needs to be fostered by a better understanding of spatial and temporal dynamics of risk, the development of a sector-, context- and scale-dependent indicator library, and the advancement of research on the relevance of individual drought vulnerability factors. Another requirement is provision of guidance on how drought risk analysis can support risk reduction.

A vital part of risk mapping is the communication of derived risk analysis to the end user addressed. The gap between science and end-users can be bridged using a spatial data infrastructure based on risk mapping (Antofie et al., 2018). Mapping is known to effectively present and communicate spatial data to stakeholders since it allows for a rapid perception of major contents (WHO, 2020). Mapping can serve as a starting point to raise awareness, realise vulnerability and to determine potential consequences (Preston et al., 2011). In the field of drought risk analysis, a visualisation of drought risk information via maps supports the analysis relevant spatial and temporal patterns as well as their interrelationships. The contents to be communicated vary specific to context, researcher’s community and stakeholder addressed. Presented information might thus be intended to: Increase public awareness, inform stakeholders or policymakers in order to foster risk management, or supply insurance companies with a quantification of potential losses. Additionally, qualitative mapping should consider stakeholder level of expertise and the heterogeneity of the audience from layman (public) to experts (scientists). Although an abundance of cartographic textbooks on best practices of designing maps exist (e.g. Krygier and Wood, 2016, Peterson, 2009, Tyner, 2014) only few reviews and little guidance for mapping risk are available (Fuchs et al., 2009; Schneider and Nocke, 2018). The World Health Organisation (WHO, 2020) published a policy brief reviewing the strength and weaknesses of recent mapping techniques in health and environment decision-making. Their report emphasises the potential of maps for highlighting localisable and diffuse issues, their capability for indicating trends over time as well as their practicability for early-warning tools. Fuchs et al. (2009) evaluated the cartographic design in flood risk mapping for a set of different stakeholders and suggested a set of aspects for an efficient design of risk maps. Schneider and Nocke (2018) showed that the choice of colour schemes for mapping (here in the realm of climate change impacts) influences the readability for the test person. However, it is also noted that “a map is primarily a means of display; it cannot predict the patterns of distribution or relationships between resources”.

In summary, drought is, in comparison to its counterpart natural hazard flood, insufficiently investigated (Kreibich et al., 2019). The current state of drought risk analysis, its components and its communication appear at very different states of research. Drought as a hazard is comparably well investigated (Hagenlocher et al., 2019) but a linkage of the hazard to drought impacts in order to add an absolute threshold value is often missing (Bachmair et al., 2015). Guidance of prevalent approaches to asses vulnerability to drought exist (Gonzalez Tanago et al., 2016; Hagenlocher et al., 2019), but the understanding of quantitatively testing current index selection criteria as well as result verification is still lacking (Blauhut et al., 2016; Gonzalez Tanago et al., 2016, Meza et al., 2019). Furthermore, the dynamics of drought impacts and vulnerability over space and time are almost unknown (Kreibich et al., 2019). Whereas drought is hardly predictable for the immediate future (Zargar et al., 2011), climate change projections analysed by the IPCC (2012) suggest that there is medium confidence that drought hazards will intensify during the twenty-first century due to reduced precipitation and increased evapotranspiration. In combination with the unpredictable development of vulnerability to drought, the likelihood of potential drought impacts in the future is unknown. Accordingly, it is essential to (I) improve the characterisation of drought risks and its components, i.e. hazard, vulnerability and drought impacts and (II) to ascertain how this risk can be communicated and used to enhance resilience to drought.

To combat the existing research gaps, this study sharpens aspects of previous studies. This paper reviews drought risk analyses with a visualisation by maps. The aim of this review is the detection of paradigms in drought risk analysis for the interaction between: Study objective, thematic focus, and data and methods applied. A specific focus is set on data selection criteria and verification schemes. Furthermore, the practical implementation of drought risk analysis into risk maps is evaluated. Finally, a synthesis of the detected pattern gives guidance for future applications.

Section snippets

Definitions and methodology applied

Drought risk as defined here is the likelihood to incur damages and economic losses during and after a drought (Vogt et al., 2018). The scientific field of risk to natural hazards analyses understands the term risk as the likelihood of negative consequences resulting from interactions between hazard and vulnerable conditions (Birkmann et al., 2013). For natural hazards such as floods or earthquakes, ‘negative consequences’ are routinely quantified by insurance companies through the estimate of

Locations and scales

By the means of the selection criteria, 84 studies were identified. All of the studies were published between 2001 and 2020 (Table 1). In figure 1, the spatial reference, localisation, and thematic focus of the studies are displayed on a global map. Seven DRAs evaluate drought risk at the global scale, six studies at international/continental scale (two at pan-European scale, one for the Mediterranean, one for Africa, two in southern Africa). Countries accentuated by greyish shading indicate

Thematic foci of drought risk analysis

Drought is a multi-facetted hazard and causes a variety of impacts (Stahl et al., 2016). Correspondingly, DRAs cannot be attributed to a single scientific discipline but to interdisciplinary research. All reviewed DRAs share comparable overall objectives: to characterise drought risk in space, to raise awareness, to inform agencies, to support decisions, or to foster drought risk research in order to reduce drought risk.

The thematic focus of DRAs can either be of general (~27%) or specific

Drought risk analysis and time

The temporal orientation of DRA can be described by two characteristics. DRA is of dynamic or static nature, and it characterises past/todays or future drought risk conditions (Table 1). Dynamic DRAs are represented by 41% of the studies. Herein, drought risk is analysed for a single or a set of specific drought situations such as damage functions. This allows to predict drought risk for different situation, e.g. within early warning systems. Static DRAs (59% of the studies) generalise drought

Paradigms of drought risk analysis

To get to the core of DRAs, the basis of this review is the de-framing of “common” terminologies. The need for deframing is shown in Table 2 giving insights into terminology applied in the reviewed studies. The variety of research objectives, and thematic and temporal foci of investigation is also reflected in the variety of ways of identifying drought risk. Moreover, the different disciplinary backgrounds of researchers world-wide contribute to the abundance of applied terminology and

Mapping drought risk as a tool for communication

Drought risk visualisation is the stage of communication to the end users. Hagenlocher et al. (2019) identified more than 100 DRAs, but a variety of DRAs do not apply a visualisation via maps. Such analyses mostly are of specific thematic focus and based on impact information (e.g. Elagib, 2014; Gil et al., 2011; Jayanthi et al., 2014; Schindler et al., 2007), but also DRAs based on the factor approach exist (e.g. Yuan et al., 2014). In these examples, drought risk is displayed e.g. as charts

Conclusion

This study reviewed paradigms of drought risk analysis and their presentation as drought risk maps across scales and sectors. Even though drought risk analysis is understood to be the basis of drought risk management, large white spaces on the global map of national and regional DRA studies exist. More studies with higher spatial resolution, especially for regions with high drought risk are needed. Drought affects a large variety of sectors, but the majority of investigations focus on

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.

Acknowledgements

I acknowledge Alexander Ulbricht, Michael Stoelzle and Kerstin Stahl for their constructive critics to improve this work. Furthermore, I acknowledge Julia Urquijo (the inventor of the triple complexity terminology), Itziar Gonzalez Tanago and Lucia de Stefano for fruitful discussions at the initial state of this review and their insights on the understanding of vulnerability to drought from a social science perspective.

This work was supported by Ministry of Science, Research and Art

References (191)

  • M. Alamgir et al.

    Parametric assessment of seasonal drought risk to crop production in Bangladesh

    Sustainability (Switzerland)

    (2019)
  • T.E. Antofie et al.

    Mapping of risk web-platforms and risk data: collection of good practices. Improving the access and share of curated EU-wide risk data for fostering DRM

    (2018)
  • Asare-Kyei D, Renaud F G, Kloos J, Walz Y, and Rhyner J 2017 Development and validation of risk profiles of West...
  • M. Azam et al.

    Copula-based stochastic simulation for regional drought risk assessment in South Korea

    Water (Switzerland)

    (2018)
  • S. Bachmair et al.

    Exploring the link between drought indicators and impacts

    Natural Hazards and Earth System Sciences

    (2015)
  • S. Bachmair et al.

    A quantitative analysis to objectively appraise drought indicators and model drought impacts

    Hydrology and Earth System Sciences

    (2016)
  • S. Bachmair et al.

    Developing drought impact functions for drought risk management

    Natural Hazards and Earth System Sciences Discussions

    (2017 May)
  • Bayarjargal Y, Karnieli A, Bayasgalan M, Khudulmur S, Gandush C, and Tucker C 2006 A comparative study of NOAA–AVHRR...
  • J. Birkmann

    Framing vulnerability, risk and societal responses: the MOVE framework

    Natural Hazards

    (2013)
  • Blauhut V, Gudmundsson L and Stahl K 2015a Towards pan-European drought risk maps : quantifying the link between...
  • Blauhut V, Stahl K, and Kohn I 2015b The dynamics of vulnerability to drought from an impact perspective Drought:...
  • V. Blauhut et al.

    Estimating drought risk across

    Europe from reported drought impacts, drought indices, and vulnerability factors Hydrology and Earth System Sciences

    (2016)
  • Blauhut V, Stölzle M and Stahl K 2017 April Towards low flow risk maps In EGU General Assembly Conference Abstracts...
  • Bronson K and Knezevic I 2016 Big Data in food and agriculture Big Data and Society 3(1) 205395171664817...
  • Brooks N 2003 Vulnerability, risk and adaptation: A conceptual framework Tyndall Centre for Climate Change Research...
  • Bunea A-I 2019 Drought Risk Assessment by Means of Drought Hazard and Vulnerability Indices in Muntenia Region. Risks...
  • D. Boucher et al.

    Current and projected cumulative impacts of fire, drought, and insects on timber volumes across Canada

    Ecological Applications

    (2018)
  • J. Buurman et al.

    Drought risk assessment in Vietnamese communities using household survey information

    International Journal of Water Resources Development

    (2020)
  • J. Chen et al.

    Risk Assessment of Drought

    Based on IDM-VFS in the Nanpan River Basin Yunnan Province China Sustainability

    (2017)
  • Chopra P, 2006 Drought Risk Assessment Using Remote Sensing and GIS: A Case Study of Gujarat Enschede Netherlands...
  • J.I. Christian et al.

    A methodology for flash drought identification: Application of flash drought frequency across the United States

    Journal of Hydrometeorology

    (2019)
  • Ciurean L R et al. 2013 Conceptual Frameworks of Vulnerability Assessments for Natural Disasters Reduction pp....
  • Cloppet E 2010 Agricultural Drought Indices in France and Europe: Strengths, Weaknesses, and Limitations In Proceedings...
  • Cook B I, Ault T R and Smerdon J E 2015 Unprecedented 21st century drought risk in the American Southwest and Central...
  • I. Dabanli

    Drought hazard, vulnerability, and risk assessment in Turkey

    Arabian Journal of Geosciences

    (2018)
  • K.S. Dayal et al.

    Spatio-temporal drought risk mapping approach and its application in the drought-prone region of south-east Queensland, Australia

    Natural Hazards

    (2018)
  • L. De Stefano et al.

    Methodological approach considering different factors influencing vulnerability—pan-European scale DROUGHT-RandSPI Technical Report No.

    (2015)
  • Cutter S L (2003) The Vulnerability of Science and the Science of Vulnerability. Annals of the Association of American...
  • Dilley M, Chen R S, Deichmann U, Lerner-Lam A L and Arnold M 2005 Natural disaster hotspots: a global risk analysis The...
  • M. Deng et al.

    Agricultural drought risk evaluation based on an optimized comprehensive index system

    Sustainability (Switzerland)

    (2018)
  • A.J. Dougill et al.

    Anticipating Vulnerability to Climate Change in Dryland Pastoral Systems : Using Dynamic Systems Models for the Kalahari

    (2010)
  • Elagib N A 2014 Development and application of a drought risk index for food crop yield in Eastern Sahel Ecological...
  • Elsler M 2017 Making the invisible visible: A social Network Analysis of stakeholders relevant for drought management...
  • Eriyagama N, Smakhtin V and Gamage N 2009 Mapping Drought Patterns and Impacts: A global Perspective (Report) IWMI...
  • European Commission 2012 Report from the Commission to the European Parliament and the Council on the Implementation of...
  • The Emergency Events Database - Université catholique de Louvain (UCL) -

    (2020)
  • E. Eze et al.

    Feasible crop insurance indexes for drought risk management in Northern Ethiopia

    International Journal of Disaster Risk Reduction

    (2020)
  • Fan G, Zhang Y, He Y and Wang K 2017 Risk Assessment of Drought in the Yangtze River Delta Based on Natural Disaster...
  • FAO. (2006). Food safety risk analysis A guide for national food safety authorities. Rome: FAO FOOD AND NUTRITION PAPER...
  • J. Frischen et al.

    Drought Risk to Agricultural Systems in Zimbabwe: A Spatial Analysis of Hazard, Exposure, and Vulnerability

    Sustainability

    (2020)
  • Cited by (27)

    • Agricultural risk assessment of compound dry and hot events in China

      2023, Agricultural Water Management
      Citation Excerpt :

      Thus, it is important to understand the risk of droughts and hot extremes in China to cope with extremes under global warming. Quantitative and qualitative risk assessments of droughts and hot extremes have been conducted in different regions including China (Blauhut, 2020; He et al., 2013; Prabnakorn et al., 2019). Though different definitions or conceptualizations exist, drought risk assessments are generally based on hazards, exposure, and vulnerability (Carrao et al., 2016; Zhao et al., 2020).

    • Do global risk assessments leave countries behind? How the selection of countries influences outcomes of drought risk assessments

      2022, Climate Risk Management
      Citation Excerpt :

      Further, the stated motivations for funding and the mandate of organisations to focus on countries most vulnerable or at risk to climate change or extreme weather events, suggest that a comparison of risk across countries should impact resource allocation. At present, drought risk assessments that can be used for comparison of risk, exposure and vulnerability across multiple countries at the country scale are limited to a handful of global assessments (Carrão et al., 2016; Meza et al., 2020), global tools/platforms (Global Drought Observatory (GDO), Water Risk Filter (WWF)); and regional assessments for Europe and Africa (see Blauhut, 2020; Hagenlocher et al., 2019 for a review of existing assessments). This means that when determining which country or countries are the most at-risk or vulnerable, organisations are mostly limited to global assessments.

    • Multiscale drought hazard assessment in the Philippines

      2021, Computers in Earth and Environmental Sciences: Artificial Intelligence and Advanced Technologies in Hazards and Risk Management
    View all citing articles on Scopus
    View full text