We lose ground: Global assessment of land subsidence impact extent
Graphical abstract
Introduction
Land subsidence (LS), defined as the settlement of the land surface, is generated by human-induced and natural-driven processes, including natural compaction of unconsolidated deposits (Zoccarato et al., 2018), and human activities such as subsurface water mining, or extraction of oil and gas (Gambolati et al., 2005). LS is a global problem (Galloway et al., 2016; Herrera-Garcia et al., 2021; Kok and Costa, 2021), mostly studied and recognized, to different extents, in association with aquifer overexploitation (which is the focus of this paper). LS occurrence around the world is most prominent in those aquifer systems composed of loose unconsolidated materials (e.g., sands, clays, and silts) that are over-pumped (e.g., Poland, 1984; Tomás et al., 2005; Gambolati and Teatini, 2015; Bonì et al., 2015).
Climate change impacts on water availability and population growth are expected to increase competition for water, leading to extensive groundwater withdrawals. The expected overexploitation of aquifers will exacerbate current and future damage from various LS impacts. LS causes significant damages to local communities and to the environment (Yoo and Perrings, 2017; Teatini et al., 2018). As such, identifying the types of damages and quantifying them in terms of the various physical impacts and their short- and long-term economic costs would be an essential first step for preparing policies to address this problem. However, most studies on LS are indicative in the sense that they identify the driving processes, and measure the physical effects of LS in specific localities. Few are the works that assess the global impacts of LS in terms of social, environmental, and/or economic consequences.
A review of existing literature suggests that LS can cause the following impacts (e.g., Poland, 1984; Holzer and Galloway, 2005; Lixin et al., 2010; Bru et al., 2013; Erkens et al., 2016), as summarized in Dinar et al. (2020): (1) Socio-economic impacts, such as structural damages (Bru et al., 2013); (2) Environmental damages, such as malfunctioning of drainage systems (Viets et al., 1979); (3) Geological-related damages that affect underground lateral water flows (Poland, 1984); (4) Environmental damages, such as reduced performance of hydrological systems (Poland, 1984); (5) Environmental damages, such as wider expansion of flooded areas (Poland, 1984); (6) Hydrogeological damages that result in groundwater storage loss (Holzer and Galloway, 2005; Béjar-Pizarro et al., 2017); (7) Impact on adaptation ability to climate change, such as the loss of the buffer value of groundwater in years of scarcity (Erkens et al., 2016); (8) Groundwater contamination, such as seawater intrusion resulting in decrease of farmland productivity in coastal aquifer systems and decrease of fresh-water availability (Holzer and Galloway, 2005; Poland, 1984); (9) Loss of high-value transitional areas (e.g., saltmarshes) (Viets et al., 1979); and (10) Shift of land use to poorer activities (e.g., from urbanized zones to rice fields, from rice fields to fish and shellfish farms, from fish farms to wastewater ponds) (Heri et al., 2018). A summary of the literature used for the ten LS attributes and their impacts is provided in Appendix A (Table A1).
Estimates of economic damages from land subsidence are not yet widely available, and most of the published studies on this phenomenon focus on a physical quantification of subsidence and on cataloguing the damages (Borchers and Carpenter, 2014). Few works have assessed local LS damages (e.g., Jones and Larson, 1975; Warren et al., 1975; Lixin et al., 2010; Tomás et al., 2012; Sanabria et al., 2014; Yoo and Perrings, 2017; Wade et al., 2018; and Díaz et al., 2018). Selected economic damages cited in the literature range from $756 million in the Santa Clara Valley of California (Borchers and Carpenter, 2014), to $1.3 billion in the San Joaquin Valley of California between 1955 and 1972, in 2013 dollars, to $18.03 billion in the Tianjin metropolitan area in the period up to 2007 (Lixin et al., 2010). It is worth noting that, since the studies leading to these estimates use different approaches, refer to different sizes of affected regions, and span over different periods of time, one should not attempt to compare the values but rather use them as indicative only. A recent study (Kok and Costa, 2021) enumerates the various types of costs associated with LS and suggests a standardize economic framework for their cost evaluation.
In a recent publication, Herrera-Garcia et al. (2021) identified 200 locations (mostly urban) in 34 countries that experienced LS during the past century. However, these authors also indicate that the LS extent is known only in one third of these locations. Given lack of direct data on damages, Herrera-Garcia et al. (2021) use what they define as the exposure to potential land subsidence (PLS) and focus on areas where the probability for potential subsidence is high. Their calculations suggest that PLS affects 8% of the global land surface, and that 2.2 million square kilometers of global land is exposed to high to very high probability for PLS, involving 1.2 billion urban inhabitants and threatening nearly US$ 8.2 trillion in GDP. This estimate on the global economic exposure could be a lower-level estimate because the authors assumed that the GDP per capita is homogenous within each country, not taking into account the geographical variations in productivity, for example, between different regions within a country, or between cities and rural areas. However, this economic estimate on the global subsidence exposure does not directly translate to subsidence impact or damages. The lack of information on the cost of damages caused by current and historical subsidence worldwide, prevents these authors from evaluating the impact of global land subsidence.
Realizing the need for a global assessment of LS impacts and the present difficulty to provide global economic quantification for those effects (Kok and Costa (2021), Herrera-Garcia et al. (2021)), in this paper we have taken an approach of quantitatively (not economically) assessing global LS impact extents and their determinants. We start with a meta-analysis and review of relevant literature on LS occurrence and physical quantification of its impacts in various sites around the world. In the absence of economic value for the LS-induced damage, we develop an index to assess the LS impact extent (LSIE), using the classification of the 10 LS impacts listed above. This assessment allows us to identify different types of impacts in different locations and is used to explain the effects of physical, regulatory, and population conditions on LSIE. Such conditions include aquifer lithology, managing institutions, social systems, existing policies, population pressure, water-level depletion from over-pumping, and several others.
From here on the paper develops as follows: Section 2 explains the principles used to develop the LSIE index. We then present in Section 3 an empirical investigation into the social, physical and institutional determinants most likely affecting land subsidence and its impact as measured by LSIE. Section 4 presents the data-collection process, the variables constructed, and the hypotheses regarding their effects on LSIE. This is followed in Section 5 by the empirical specifications of our models and the derived hypotheses. Section 6 includes results from the LSIE global distribution, and results from the statistical analysis. The results are followed by policy simulations in Section 6, with estimates of the incremental impact of policy variables on LSIE. Discussion on the policy results is provided in Section 7. In Section 8 we present our conclusions and policy implications.
Section snippets
The LS Impact Extent (LSIE) Index
Use of indicative indexes to assess environmental health status has been practiced by many national and international agencies (OECD, 2003; EEA—Gabrielsen and Bosch, 2003; EPA—Fiksel et al., 2012). Use of indexes allows comparison across states and geographical regions (OECD, 2003). As explained below, we developed an indicative index to measure LS impact extent in the locations of the dataset we compiled.
Due to the heterogeneous and partial nature of the information we extracted from all
Land subsidence extent and its causes
LS is caused by a combination of social, policy, and physical factors—stratigraphic, lithological and geomechanical characteristics of the aquifer system, and groundwater table depletion, or lowering of the piezometric head for a phreatic or confined aquifer system, respectively (Poland, 1984; Tomás et al., 2011; Gambolati and Teatini, 2015). This latter variable is controlled by the anthropogenic pressure on the aquifer system, usually represented by urban and agricultural demands, and is
Study area, data, variable construction, and general hypotheses
Technical published articles were retrieved, using search engines and publication databases, such as Jstore (www.jstor.org) and Agricola (https://www.ebsco.com/products/research-databases/agricola). We focused on technical papers in peer-reviewed journals and on books and book chapters. We searched only for English-written documents. We used the following keywords—land subsidence, groundwater, over-pumping, economic analysis, hydrology, land subsidence impacts—to search for titles, abstract
Empirical specifications and hypotheses
The model in (4) is developed using linear terms for all variables and quadratic relationships for Pop and Dep. Given that our dependent variable, LSIE, contains real values that range from 0.1 to 1.0 and between 0.028 and 0.960 for LSIE-EW and LSIE-W, respectively, we use the ordinary least squares (OLS) estimation procedure to uniquely identify the model. Since our dependent variable is continuous it is justified to employ a linear equation with quadratic terms for the continuous independent
Results
The analysis in this paper utilizes only 113 of the 119 observations in our dataset, due to missing values of depletion of groundwater in aquifers in some of the sites and due to one outlier observation (The Mekong Delta). One possible explanation for The Mekong Delta, being an outlier is that the observation of the Mekong Delta (serving 10.7 million people) spans over a very wide region with many different geological, hydrological, and social/economic conditions that could lead to unexpected
Policy simulations
Several of the variables in the investigated models provided in Table 6 could be considered for policy intervention options using the sign and value of the regressors to quantify their incremental effects. To keep the paper length, we will demonstrate the effects of policy impacts using model 1 only. The analysis includes the effects of population change (Pop), access to surface water (Suw), reduction in GW level (Dep), and indirectly the interactions between governance level and regulation
Discussion, policy implications, and limitations
In spite of its major social cost in hundreds of sites around the world, the majority of which have irreversible negative physical and economic impacts, land subsidence has not been given proper preventive attention by regulatory agencies and local water management organizations in many countries. We were able to identify and analyze land subsidence effects in 113 locations where mainly physical consequences of land subsidence have been assessed but economic damages, likely in the range of
CRediT authorship contribution statement
Ariel Dinar: Conceptualization, Formal analysis, Funding acquisition, Investigation, Methodology, Supervision, Validation, Writing – original draft, Writing – review & editing. Encarna Esteban: Conceptualization, Methodology, Validation, Writing – review & editing. Elena Calvo: Conceptualization, Methodology, Validation, Writing – review & editing. Gerardo Herrera: Conceptualization, Formal analysis, Data curation, Investigation, Methodology, Supervision, Validation, 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.
Acknowledgements
The authors were inspired by a session on land subsidence at the Rosenberg International Forum, San Jose, California, USA, October 7-10, 2018, dedicated to sustainable groundwater management. The authors acknowledge input from the UNESCO experts to the Delphi process. Partial funding was provided by the Giannini Foundation of Agricultural Economics Minigrant Program. Dinar would like to acknowledge support from the W4190 Multistate NIFA-USDA-funded Project, “Management and Policy Challenges in
References (42)
- et al.
Mapping groundwater level and aquifer storage variations from InSAR measurements in the Madrid aquifer, Central Spain
J. Hydrol.
(2017) - et al.
Twenty-year advanced DInSAR analysis of severe land subsidence: the Alto Guadalentín Basin (Spain) case study
Eng. Geol.
(2015) - et al.
Multitechnical approach for damage assessment and reinforcement of buildings located on subsiding areas: study case of a 7-story RC building in Murcia (SE Spain)
Eng. Struct.
(2018) - et al.
Are intra- and inter-basin water transfers a sustainable policy intervention for addressing water scarcity?
Water Security
(2020) - et al.
Mapping ground subsidence induced by aquifer overexploitation using advanced differential SAR interferometry: Vega Media of the Segura River (SE Spain) case study
Remote Sens. Environ.
(2005) - et al.
Persistent scatterer interferometry subsidence data exploitation using spatial tools: the Vega Media of the Segura River Basin case study
J. Hydrol.
(2011) - et al.
A novel approach to reducing uncertainty: the group Delphi
Technol. Forecast. Soc. Chang.
(1991) An evaluation of Delphi
Technol. Forecast. Soc. Chang.
(1991)- et al.
Quantifying the contribution of multiple factors to land subsidence in the Beijing Plain, China with machine learning technology
Geomorphology
(2019) - et al.
Land subsidence from groundwater use in California
Control of deformation of buildings affected by subsidence using persistent scatterer interferometry
Struct. Infrastruct. Eng.
Land subsidence: the forgotten enigma of groundwater (over-)extraction
Global-scale assessment of groundwater depletion and related groundwater abstractions: combining hydrological modeling with information from well observations and GRACE satellites
Water Resour. Res.
Double trouble: subsidence and CO2 respiration due to 1,000 years of Dutch coastal peatlands cultivation
Hydrogeol. J.
Framework for sustainability indicators at EPA
Social decision-making and land subsidence
Environmental indicators: typology and use in reporting
Improving sustainability of programmes in strategic environmental assessment procedures: the QUAlitative Structural Approach for Ranking (QUASAR) the environmental effects
European Journal of Sustainable Development
Preface: land subsidence processes
Hydrogeol. J.
Geomechanics of subsurface water withdrawal and injection
Water Resour. Res.
Anthropogenic land subsidence
Cited by (33)
Three-dimensional Biot consolidation solution of adjacent aquitards induced by regional drawdown of the water table
2024, Computers and GeotechnicsDeclining groundwater and its impacts along Ganga riverfronts using combined Sentinel-1, GRACE, water levels, and rainfall data
2024, Science of the Total EnvironmentModeling the optimal management of land subsidence due to aquifers overexploitation
2024, Journal of Environmental ManagementPresent-day land subsidence over Semarang revealed by time series InSAR new small baseline subset technique
2023, International Journal of Applied Earth Observation and GeoinformationSentinel-1 InSAR-derived land subsidence assessment along the Texas Gulf Coast
2023, International Journal of Applied Earth Observation and Geoinformation