Local and non-local drivers of consumption-based water use in China during 2007–2015: Perspective of metacoupling

Highlights

• National WU decreased under the policy intervention in China.

• Migration effect is identified to indicate the non-local factors affected WU trends.

• The largest per capital WUs were found in low developed provinces.

• WUs in less developed areas were more driven by capital investment than consumption.

• Decline in WU intensity partially leads to the decoupling of WU and GDP.

Abstract

Increasingly growing consumption-based water use (WU) combined with climate change have exacerbated water stress globally and regionally, yet little is known about how the WU change is affected by metacoupled processes which involve human-nature interactions across space; within and across adjacent and distant places. This study aims to unveil the spatio-temporal pattern of China's WUs during 2007–2015 and its underlying local and non-local drivers. Results show that China's total WU exhibited an upward trend from 386.7 billion m³; in 2007 to 431.2 billion m³ in 2012 but dropped to 412.6 billion m³ by 2015. Widespread and continuous water use efficiency improvement contributed most to offsetting the increase in WU driven by the rising affluence and growing population in the context of rapid urbanization and industrialization. Economic structure drove a relatively large WU reduction (responsible for −23.7% of the WU change during 2007–2015), in line with China's ongoing transform from a capital investment-driven economy to a consumption-driven one and decoupling economic growth from environmental pressure. The population share representing the non-local factor of migration effect was large enough to be seen clearly in the changing WUs across China: the WUs of coastal areas ascended while inland areas descended, which was in accordance with migration patterns. Our findings could make a valuable contribution to decision-making in identifying hotspot areas, charting systematic courses for sustainable water use, and combining demand-oriented and supply-oriented measures.

Introduction

As one of the critical determinants of achieving the 2030 Agenda, water portrays a crucial role in many of the United Nations Sustainable Development Goals (UN SDGs) (Taka et al., 2021). Global initiatives including International Decade for Action ‘Water for Life’ 2005–2015 underline the importance of ensuring the availability and sustainable management of water and sanitation for all by 2030 (FAO, 2020; Hoekstra et al., 2017). Water stress, exacerbated by climate change, population growth, economic development, food production, and diet shifts, has been recognized as one of the top five global risks threatening economy and ecosystems (World Economic Forum, 2020). According to UN-Water (2021), two-thirds of the world's population are estimated to be living in water-stressed areas by 2025 if current consumption patterns continue, with many developing countries facing the most severe situation (WBCSD, 2014).

Water demand, or the total amount of freshwater required for services or producing goods and providing services, is usually depicted by consumption-based water use (WU) and its associated indicator virtual water (VW, defined as part of the water consumed in one area supported by other areas through virtual water trade) (D'Odorico et al., 2019; Gohari et al., 2013). WU and VW reveal the geographical mismatch between water suppliers and water demanders. Water-intensive sectors such as agriculture are critical in reducing WU and better managing water resources (Tuninetti et al., 2020; Zhao et al., 2019).

Over the past few decades, rapid urbanization and growing population have exacerbated water-related problems worldwide including water shortages, degraded water quality, floods, and energy shortage. As one of the world's most water-stressed and most rapidly urbanizing countries, China has developed diverse water conservancy measures to avoid the long-term water crisis since the 1980s (Giordano, 2007; Zhou et al., 2020). These policy measures have contributed to China's transition towards a more resource-efficient and circular economy in achieving social development and technological progress. In this context, the high dynamism and heterogeneity of China's consumption-based water use (WU) links widely and deeply to, multiple socioeconomic, political, and environmental factors (e.g., economic development, industrialization, and urbanization) that operate concurrently and that vary spatially (Bryan et al., 2018; Liu, 2016, Liu, 2017, Liu et al., 2018, Liu et al., 2019, Liu et al., 2020, 2017).

Previous literature reports extensive research on water consumption and drivers of its evolution in China, using correlation analysis, IPAT framework, STIRPAT model, decomposition analysis methods, among others (Fan et al., 2019; Zhao et al., 2014; Zhi et al., 2015; Zhou et al., 2020). Analytic dimensions derive usually from size, structure, intensity, or the dimension of population, affluence, and technology, or their combinations. Zhao et al. (2014) explored the influences of population, affluence, urbanization level, and diet structure on agriculture products-related water footprint change by applying an extended STIRPAT model. Zhi et al. (2015) developed the IO-IPAT-SDA approach to decompose WU drivers and evaluate their impacts on WU changes by different final users. Zhou et al. (2020) identified the deceleration of China's WU and quantified the relative contributions of 14 socioeconomic drivers using a log-mean Divisia index (LMDI) model. Fan et al. (2019) improved the traditional SDA model to identify the driving factors of different types of final demand. Most studies have focused on the whole of China or a single region within China and have rarely considered inter-regional interactions when identifying the factors influencing dynamic evolution in WU.

A small body of literature has targeted WU change from the perspective of regional relevance. For instance, given that global population growth and the associated water demand increase is partly sustained by virtual water trade (D'Odorico et al., 2019), Liu et al. (2019) examined the effects of production fragmentation on virtual water trade and concluded that the value chain-related trade increased national water use most. Liu (2016) reckoned that for an area with a high degree of external dependence like Beijing, economic and technical conditions of the importing region would have a great influence on its water footprint together with the product import volume. Hence, Liu (2016) disassembled the water consumption of an area into two parts (internal and external water footprint), analyzed their driving factors respectively, and defined them as internal and external driving factors. Gao et al. (2020) considered the spatial heterogeneity of the final demand when analyzing China's production WU, inferred the regional coordination situation based on result of the WU distribution between regions and sectors, and then put forth policy suggestions. However, these studies failed to quantify independent variables that explicitly characterized potential impacts from other regions or establish a direct mapping relationship between the independent variables and the total water consumption by regions. From this perspective, little is known about how the WU change would be affected by meta-coupled processes including human-nature interactions within and across adjacent and distant places. Local WU drivers have been intensively analyzed involving consumption level, population scale, WU efficiency, and local final demand (Fan et al., 2019; Zhao et al., 2014; Zhi et al., 2015; Zhou et al., 2020), yet local and non-local factors have almost never been explicitly considered at the same time.

To fill this research gap, we seek to answer the following question: how strongly do different drivers influence WUs both at local level and beyond regional boundaries? Our working hypotheses is that local and non-local factors simultaneously modulate such dynamics, acting differently across provinces. Our indicators for evaluating WU heterogeneity involve total WU, per capital WU, and WU per GDP. Our accounting method for WU volume is environmentally-extended input-output analysis. Local and non-local factors are selected based on previous studies and the metacoupling framework (Liu, 2017), and their relative contributions to WU change are quantified at both provincial and sectoral levels using index decomposition analysis. The findings of this study are conducive to identifying hotspots and priority steps for water saving in China, serving also as a reference for other countries in achieving sustainable water resource management.