Elsevier

Agricultural Water Management

Volume 271, 1 September 2022, 107748
Agricultural Water Management

Drought risk assessment in Mediterranean agricultural watersheds: A case study in Central Italy

https://doi.org/10.1016/j.agwat.2022.107748Get rights and content

Highlights

  • A methodology to perform drought risk assessment is presented for Mediterranean watersheds.

  • The assessment is validated through a rapid and complete robustness evaluation.

  • Southernmost municipalities of Tuscany have the highest drought risk.

  • Tailored adaptation strategies are proposed for different clusters of municipalities.

Abstract

Mediterranean watersheds are expected to face increased and more severe drought events due to climate change. Urgent action is needed to shift from a reactive approach to a proactive one, for which drought risk assessment is fundamental. Nevertheless, the current methodology to calculate composite risk indicators is still debated, undermining the overall robustness and validity of drought risk assessments. Furthermore, the diversity of socio-ecological contexts, spatiotemporal scales, and data availability hamper the homogenization of assessment methods. We present a complete drought risk assessment methodology, applied to the agricultural systems of five Italian coastal watersheds, introducing a simple robustness evaluation method to validate the assessment tool and archetype analysis to link the outputs with adaptation strategies. Forty-two (42) indicators were included to represent hazard, exposure, and vulnerability. Past and future drought hazards were estimated considering multiple types of droughts with data from public observatories.The results show highest hazard for the southern part of the study region, highest exposure in the coastal and high-value wine producers’ municipalities, while vulnerability patterns are less clear. Major adaptation efforts should target specific watersheds of the Grosseto province, which has the highest projected drought risk. Archetype analysis was then used to suggest possible context-specific adaptation strategies. The proposed methodology, with the integration of a robustness evaluation and the archetype analysis, represents an advancement toward shared and homogeneous guidelines in state-of-the-art drought risk assessments.

Introduction

The occurrence and severity of droughts - defined as “periods of abnormally dry weather long enough to cause a serious hydrological imbalance” (IPCC, 2012) – are expected to increase in the future due to climate change (Dai, 2011; IPCC, 2022). The Italian peninsula, located in the heart of the Mediterranean region, will be a hotspot for climate change (MedECC, 2020, Spano et al., 2020, Zollo et al., 2016), facing more frequent and intense drought events (Cammalleri et al., 2020, Caporali et al., 2021, Castellari et al., 2014, OECD, 2021). Droughts, characterized in terms of frequency, severity, duration, and extent, have typically been considered a natural phenomenon triggered by lack of precipitation, (Zargar et al., 2011). Different types of droughts have been identified, including (a) the meteorological drought – “a lack of precipitation over a region for a period of time”, (b) the agricultural drought – “a period with declining soil moisture and consequent crop failure”, (c) the hydrological drought – “a period with inadequate surface and subsurface water resources for established water uses”, and (d) the socioeconomic drought –“a failure of water resources systems to meet water demands” (Mishra and Singh, 2010). In the past, droughts were considered to propagate from meteorological to agricultural, hydrological and socioeconomic droughts. This approach has been recently questioned for its over-simplification since it fails to account for (i) feedback and trade-offs between social and physical processes, (ii) direct effects of human-induced climate change and (iii) long-term environmental impacts of drought (AghaKouchak et al., 2021, Crausbay et al., 2017, Di Baldassarre et al., 2021, Van Loon et al., 2016). As a result, the terms “ecological drought” and “human-induced hydrological drought” were recently introduced (Crausbay et al., 2017, Van Loon et al., 2016). A further step will need to be done to improve our understanding of “anthropogenic drought”, a multidimensional and multiscale phenomenon that should be intended as a “process” rather than a “product” (AghaKouchak et al., 2021).

Globally, it is estimated that drought damages account for a fifth of the total damages caused by natural hazards (World Meteorological Organization (WMO) and Global Water Partnership (GWP), 2017). In Europe, the annual economic losses caused by droughts are estimated to be around € 9 billion (€ 1.4 billion for Italy), mostly related to the agricultural sector (Cammalleri et al., 2020), with significant spatial variability between different regions (García-León et al., 2021). The entity of the losses, along with the expected increase due to climate change, boosted the interest of researchers and decision-makers in this topic (Hagenlocher et al., 2019). To better deal with droughts, recent studies call for a shift from the so-called “reactive” approach, taken in emergency situations and considered technically and economically inefficient, towards a “proactive” approach, including appropriate measures developed with the involvement of multiple stakeholders (Carrão et al., 2016, Murthy et al., 2015, Vogt et al., 2018). In fact, preparation and mitigation costs are by far lower compared to the relief costs, which significantly surge in case of inaction (Vogt et al., 2018; World Meteorological Organization (WMO) and Global Water Partnership (GWP), 2017). To better prepare for droughts, vulnerability and risk assessments are considered of major importance for developing sound and effective strategies (World Meteorological Organization (WMO) and Global Water Partnership (GWP), 2014; World Bank, 2019).

Commonly, drought hazard is quantified by indicators of drought severity, frequency, intensity, and duration. Many drought hazard indicators exist to describe meteorological, agricultural, and hydrological droughts (Kchouk et al., 2022, Mishra and Singh, 2010, Zargar et al., 2011), which might refer directly to physical variables such as precipitation, evapotranspiration, soil moisture or streamflow, or can infer drought from vegetation health. Despite the availability of multiple hazard indicators, drought risk assessments are generally carried out with the use of one or a few hazard indicators. However, considering hazard indicators of meteorological, agricultural, and hydrological droughts can better embrace the complexity of drought hazards (Sun et al., 2012, World Bank, 2019). Furthermore, the most common approach to represent drought hazard is by using historical data; nevertheless, many authors (e.g. Hagenlocher et al., 2019; Vogt et al., 2018) claim the importance of including predictions of future droughts, but at the cost of higher uncertainty (Mysiak et al., 2018).

Two major limitations to the validity and practical use of drought risk assessments are commonly shared: (1) only 11% of them conduct some form of validation and, (2) they generally miss a link with possible adaptation strategies (Hagenlocher et al., 2019). The robustness of the methodology applied can be evaluated with uncertainty or sensitivity analyses (OECD, 2008), sometimes referred to as internal validation (Carrão et al., 2016, Fontaine and Steinemann, 2009), or by comparing results with external information, referred to also as external validation. Finding suitable data for external validation might be impossible in some regions of the world and when dealing with small scales. Additionally, the validation with external datasets is complicated by the fact that composite indicators aim to represent complex past and future dynamics. Hence, the validity of composite indicators is generally evaluated by performing uncertainty and sensitivity analyses (OECD, 2008). Even if examples exist, a shared, simple but at the same time robust methodology for internal validation is still missing.

The main objective of this study is to propose a detailed and integrated drought risk assessment of Mediterranean agricultural systems and present an application to the municipalities located in coastal watersheds of Central and Southern Tuscany (Central Italy). These areas are susceptible to drought especially during the summer months, due to the concurrent high water demands for domestic and agricultural uses. Key innovations introduced in this paper are (1) a complete robustness evaluation of the assessment by applying alternative methodologies in crucial steps of the drought risk assessment; (2) the use of archetype analysis to streamline the identification of exposure and vulnerability patterns and propose a link with possible adaptation strategies.

Section snippets

Methodology

The methodology of this study draws on the guidelines introduced by OECD (2008) and the approaches of other drought risk assessments (Hagenlocher et al., 2018, Meza et al., 2020), with the addition of a robustness evaluation and archetype analysis for a more integrated assessment. The eight operational steps are:

  • 1)

    Conceptual framework definition;

  • 2)

    Study area definition;

  • 3)

    Identification of indicators;

  • 4)

    Data acquisition and pre-processing;

  • 5)

    Multicollinearity analysis;

  • 6)

    Normalization and weighted aggregation;

  • 7)

Multicollinearity analysis

Several significant multicollinearities were detected (p-value<0.05). As a result, the indicator “Share of horticulture and fruticulture” was excluded from exposure indicators since it was highly correlated with “Value of Agricultural Products” and “Volume of water used for irrigation”. For the vulnerability indicators, “Climate interference” and “Water deficit” were excluded since their variability was explained by other indicators, namely “Soil erosion” and “Soil fertility”. Similarly, “Share

Past and future drought hazard

Interesting results were found in the estimation of drought hazard. Correlations between the total severity, duration, and frequency calculated with SPI3, SPI6, SPI12, and VHI in the 58 municipalities are poor, implying that the use of only one of them would have resulted in a different pattern of drought hazard. The non-standardized indicators of future hazard showed very high correlations when considering the same indicator in the short-, medium-, and long-term future, while much worse

Conclusion

A complete drought risk assessment was conducted for 58 municipalities belonging to five coastal watersheds in Tuscany (Central Italy). The proposed approach allowed to produce a policy-relevant drought risk assessment, even though adjustments could further improve the methodology. The inclusion of multiple drought hazard indicators provided a more comprehensive analysis of drought risk. The use of future projections to account for climate change impacts also confirmed the patterns of past

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

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.

References (83)

  • M. Märker et al.

    Assessment of land degradation susceptibility by scenario analysis: A case study in Southern Tuscany, Italy

    Geomorphology

    (2008)
  • A.K. Mishra et al.

    A review of drought concepts

    J. Hydrol.

    (2010)
  • C.S. Murthy et al.

    Geospatial analysis of agricultural drought vulnerability using a composite index based on exposure, sensitivity and adaptive capacity

    Int. J. Disaster Risk Reduct.

    (2015)
  • M. Napoli et al.

    Determining potential rainwater harvesting sites using a continuous runoff potential accounting procedure and GIS techniques in central Italy

    Agric. Water Manag.

    (2014)
  • A. Nauditt et al.

    Evaluating tropical drought risk by combining open access gridded vulnerability and hazard data products

    Sci. Total Environ.

    (2022)
  • T.P.L. Nguyen et al.

    Perceiving to learn or learning to perceive? Understanding farmers’ perceptions and adaptation to climate uncertainties

    Agric. Syst.

    (2016)
  • L. Sharafi et al.

    Drought risk assessment: Towards drought early warning system and sustainable environment in western Iran

    Ecol. Indic.

    (2020)
  • J.H. Stagge et al.

    Modeling drought impact occurrence based on meteorological drought indices in Europe

    J. Hydrol.

    (2015)
  • L. Vergni et al.

    Spatio-temporal variability of precipitation, temperature and agricultural drought indices in Central Italy

    Agric. . Meteorol.

    (2011)
  • A. AghaKouchak et al.

    Anthropogenic drought: definition, challenges, and opportunities

    Rev. Geophys.

    (2021)
  • S. Bachmair et al.

    How well do meteorological indicators represent agricultural and forest drought across Europe?

    Environ. Res. Lett.

    (2018)
  • P. Barazzuoli et al.

    Olocenic alluvial aquifer of the River Cornia coastal plain (southern Tuscany, Italy): database design for groundwater management

    Environ. Geol.

    (1999)
  • P. Barazzuoli et al.

    A conceptual and numerical model for groundwater management: a case study on a coastal aquifer in southern Tuscany, Italy

    Hydrogeol. J.

    (2008)
  • D. Bianchi et al.

    Vineyard water stress evaluation using a multispectral index: a case study in the Chianti area

    Acta Hortic.

    (2021)
  • S. Bianchi et al.

    Hydrogeological investigations in southern Tuscany (Italy) for coastal aquifer management

    AQUA mundi

    (2011)
  • C. Cammalleri et al.

    Global warming and drought impacts in the EU

    Eur. 29956 EN, Publ. Off. Eur. Union, Luxemb.

    (2020)
  • F. Cantini et al.

    Evidence-Based Integrated Analysis of Environmental Hazards in Southern Bolivia

    Int. J. Environ. Res. Public Health

    (2019)
  • E. Caporali et al.

    A review of studies on observed precipitation trends in Italy

    Int. J. Climatol.

    (2021)
  • O.D. Cardona et al.

    Determinants of risk: Exposure and vulnerability. In: Managing the Risks of Extreme Events and Disasters to Advance Climate Change Adaptation

  • S. Castellari et al.

    Rapporto sullo stato delle conoscenze scientifiche su impatti, vulnerabilità ed adattamento ai cambiamenti climatici in Italia

    Minist. dell’Ambiente e della Tutela Del. Territ. e Del. Mare, Roma

    (2014)
  • M. Clark et al.

    Comparative analysis of environmental impacts of agricultural production systems, agricultural input efficiency, and food choice

    Environ. Res. Lett.

    (2017)
  • S.D. Crausbay et al.

    Defining Ecological Drought for the Twenty-First Century

    Bull. Am. Meteorol. Soc. 98, 2543–2550.

    (2017)
  • A. Dai

    Drought under global warming: a review

    WIREs Clim. Chang.

    (2011)
  • G. Di Baldassarre et al.

    Integrating Multiple Research Methods to Unravel the Complexity of Human‐Water Systems

    AGU Adv.

    (2021)
  • G. Di Baldassarre et al.

    Water shortages worsened by reservoir effects

    Nat. Sustain.

    (2018)
  • B. Di Lena et al.

    Analysis of drought in the region of Abruzzo (Central Italy) by the Standardized Precipitation Index

    Theor. Appl. Climatol.

    (2014)
  • C.F. Dormann et al.

    Collinearity: A review of methods to deal with it and a simulation study evaluating their performance

    Ecography (Cop. )

    (2013)
  • K. Eisenack et al.

    Design and quality criteria for archetype analysis

    Ecol. Soc. 24, art6.

    (2019)
  • M. Enenkel et al.

    Why predict climate hazards if we need to understand impacts? Putting humans back into the drought equation

    Clim. Change

    (2020)
  • M.M. Fontaine et al.

    Assessing vulnerability to natural hazards: impact-based method and application to drought in Washington State

    Nat. Hazards Rev.

    (2009)
  • R. Giordano et al.

    Integration of local and scientific knowledge to support drought impact monitoring: some hints from an Italian case study

    Nat. Hazards

    (2013)
  • Cited by (19)

    • Dynamic risk assessment of waterlogging disaster to spring peanut (Arachis hypogaea L.) in Henan Province, China

      2023, Agricultural Water Management
      Citation Excerpt :

      Based on the “Two-Factor” theory, Zhang et al. (2019) constructed a dynamic drought risk assessment model to assess the drought risk of maize in Jilin Province. Villani et al. (2022) assessed drought risk in the Mediterranean agricultural basin from three aspects: hazard, vulnerability, and exposure. Pei et al. (2019) selected four evaluation indices from four cities in Heilongjiang Province as application examples, and analyzed the temporal and spatial changes in agricultural drought risk.

    View all citing articles on Scopus
    View full text