Measurement and Spatio-Temporal Evolution Analysis of Green Water Efficiency in Shaanxi Province Based on the SBM-Malmquist Model

Abstract

Improving water resource green efficiency is an important approach to alleviating the contradiction between water supply and demand. This paper takes Shaanxi Province as the study area and constructs a panel dataset using data from 10 prefecture-level cities in Shaanxi Province from 2013 to 2023. First, the SBM model and Malmquist index are used to calculate and analyze the green efficiency of water resources in Shaanxi Province. Second, the Tobit model is used to test the factors influencing the green efficiency of water resources in Shaanxi Province. The results show the following: (1) During the period from 2013 to 2023, Shaanxi Province’s water resource green efficiency was generally poor, but it showed an overall upward trend with significant regional differences. (2) The average Malmquist index for water resource green efficiency in Shaanxi Province from 2013 to 2023 was 1.176, and there was a noticeable lag in the conversion between technological innovation and its practical application in Shaanxi Province. (3) The proportion of the secondary industry and per capita water resources had a significant impact on water resource green efficiency in Shaanxi Province.

1. Introduction

As a vital resource, water sustains life, enables production, and drives economic development. With global environmental shifts intensifying, sustainable water resource management has become increasingly critical. Nearly half the world’s population currently experiences acute water scarcity [1]. This grim situation requires the international community to urgently formulate and implement scientific and efficient water resource management strategies and response measures to effectively resolve multiple overlapping water security challenges. The rapid progress of industrialization and urban development has intensified environmental degradation and water scarcity challenges in China [2,3]. Enhancing water use efficiency serves as a key approach to mitigating this issue effectively. In 2011 and 2012, the State Council of China issued two important documents, The policy documents Decision on Accelerating Water Resource Reform and Development and Opinions on Implementing the Strictest Water Resource Management System explicitly advocate for advancing water conservation initiatives and enhancing hydrological resource utilization effectiveness. Consequently, while ensuring rigorous ecological conservation, China’s policy priorities focus on two key objectives: (1) optimizing comprehensive water productivity and (2) accelerating the transition toward sustainable, low-emission development models [4]. Shaanxi Province is located in the inland region of northwest China and faces severe water scarcity. As a northwestern inland province confronting acute hydrological constraints, Shaanxi experiences escalating water demands driven by demographic expansion and urban development. Recent years have presented both substantial obstacles and transformative prospects for regional water governance. Strategic policy frameworks including the “14th Five-Year Water Resources Development Blueprint” and Provincial Water Network Infrastructure Plan establish critical pathways for optimizing water resource allocation, with profound implications for both ecological stewardship and socioeconomic progress. Enhancing water productivity emerges as a fundamental prerequisite for achieving long-term hydrological sustainability and balanced regional growth.

Contemporary academic inquiry into water use productivity has evolved from simplistic single-metric assessments toward holistic total-factor productivity frameworks. Prior research on water use efficiency has predominantly focused on economic output, often neglecting a systematic integration of water resources with complementary production inputs [5,6]. As academic research has deepened, scholars have come to recognize that water use efficiency involves the interaction of multiple production factors rather than a single isolated factor. To achieve high-quality economic, social, and eco-environmental development under green, low-carbon, and circular principles, researchers have proposed the concept of water resource green efficiency. This metric reflects the ratio of input factors (e.g., water resources) to their resulting economic, social, and ecological outputs [7,8,9]. Consequently, precise identification of critical determinants influencing ecological water use performance—while simultaneously pursuing economic growth and environmental conservation objectives—can minimize ecological externalities and facilitate optimized hydrological governance. This dual-focused approach ultimately contributes to sustainable socioeconomic advancement [10].

Currently, valuable research results have been achieved in the study of water resource green efficiency. In terms of evaluation methods, the main methods for evaluating water resource green efficiency are as follows: Ratio Analysis Method [11,12], Indicator System Method [13], Stochastic Frontier Method [14], Shadow Pricing Method [15], Data Envelope Analysis [16,17,18], etc. Early studies frequently employed data envelopment analysis (DEA) for water resource green efficiency assessments due to its distinctive methodological benefits [19,20]. However, traditional data envelopment analysis cannot take into account situations with undesirable outputs. Contemporary research has witnessed a growing adoption of slack-based measure (SBM) approaches in water efficiency analysis [21,22,23,24,25]. Incorporating slack variables, the SBM framework demonstrates enhanced adaptability when processing incomplete datasets and accounting for measurement uncertainties.

Regarding the selection of measurement indicators, early calculations of water resource green efficiency mostly only considered economic output. For example, Zhang et al. [26] employed three core urbanization metrics—demographic transition, economic development, and spatial equilibrium—as primary determinants for assessing irrigation water use efficiency across the Yellow River Basin during the 2007–2017 period. With increasing emphasis on ecological conservation, researchers have expanded existing frameworks by incorporating environmental externalities into their analyses. mainly using indicators such as sewage discharge and greywater footprint. The study by Wang et al. [27] assessed ecological water use performance across 11 provincial administrative regions within the Yangtze River Economic Zone, integrating both freshwater withdrawal metrics and effluent discharge indicators into its methodological approach. Cao et al. [28] conducted a regional-scale quantification of grain production water footprints across China’s irrigated croplands, systematically examining the crop–water nexus.

In summary, research on water resource green efficiency has yielded some results, but most studies have focused on large-scale levels such as interprovincial or urban agglomeration scales, with few studies targeting a specific province and its prefecture-level cities; most scholars only consider economic benefits and water pollution factors at the output level, with only a few scholars incorporating social development into the evaluation system. Current research on determinants of ecological water use performance predominantly employs cross-regional comparative methodologies, while largely neglecting temporal variations in critical factor effectiveness. To bridge current research gaps, this investigation pursues three core aims: (1) evaluating provincial and municipal ecological water use performance across Shaanxi, (2) implementing a tripartite regional division (northern basin, central plain, and southern highlands) for granular spatial–temporal analysis, and (3) systematically examining efficiency evolution patterns. Regarding methodology, this research enhances measurement precision by expanding conventional economic-environmental indicator systems through inclusion of social productivity metrics. Furthermore, it identifies principal determinants of hydrological sustainability improvement across Shaanxi, while examining temporal variations in their relative influence—thereby establishing evidence-based policy recommendations for optimizing provincial water use performance. This research establishes a comprehensive analytical framework comprising the following:

(1) Input metrics: labor force participation, aggregate fixed capital formation, hydrological resource utilization, and energy usage. (2) Output metrics: economic output value, composite social progress indicators, urban green infrastructure per capita, and effluent volumes. The slack-based measure (SBM) methodology is employed to quantify provincial water sustainability performance during 2013–2023, evaluating its geographical and temporal evolution patterns. Subsequently, Malmquist productivity analysis is conducted, with decomposition into efficiency change (ECc) and technological change (BPCc) components, enabling rigorous examination of dynamic efficiency determinants. In the final analytical phase, censored regression modeling (Tobit) was implemented to identify principal drivers influencing ecological water use performance. Complementarily, geographically, and temporally weighted regression (GTWR) elucidated spatio-temporal heterogeneity in factor effects across Shaanxi’s sub-regions, generating evidence-based recommendations for provincial water governance optimization. The methodological architecture is visualized in Figure 1.

2. Study Area and Dataset Description

2.1. Research Area

Geographically positioned in northwestern China, Shaanxi occupies a central section of the Yellow River’s middle reaches [29]. Its geographical location and regional divisions are shown in Figure 2. Based on its natural geographical characteristics, Shaanxi Province can be further divided into three regions: the Guanzhong Region, the Shaanbei Region, and the Shaanan Region [30]. The Guanzhong Region includes five prefecture-level cities: Xi’an, Tongchuan, Baoji, Xianyang, and Weinan. It is densely populated and serves as the economic core of Shaanxi Province, with a prominent imbalance between water supply and demand. The Shaanbei Region comprises two prefecture-level cities: Yan’an and Yulin. It relies on advantageous industries such as energy and chemicals, resulting in significant water demand. The Shaanan Region includes three prefecture-level cities: Hanzhong, Ankang, and Shangluo. While its economic development is relatively slow, it enjoys relatively favorable water resource conditions.

2.2. Data Collection and Sources

The data for this study primarily comes from the “Shaanxi Province Statistical Yearbook” (2013–2023) (https://tjj.shaanxi.gov.cn/tjsj/ndsj/tjnj/ (accessed on 15 June 2025)), “The Shaanxi Province Water Resources Bulletin” (2013–2023) (https://slt.shaanxi.gov.cn/zfxxgk/fdzdgknr/zdgz/szygb/ (accessed on 15 June 2025)). Municipal statistical yearbooks and annual socioeconomic development reports (2013–2023) served as primary data sources, with linear interpolation applied to address data gaps.

3. Research Methods

3.1. Water Resource Green Efficiency Calculation Model Based on SBM

(1)

Selection of indicators

Drawing on the research results of previous studies [31,32,33,34,35], this paper selects four input indicators and four output indicators. Comprehensive statistical outputs are presented in

Table 1. Indicator system for slack-based measure (SBM) model evaluation.

Name	Description	Indicator Type
Labor force I1	Workforce metrics were derived from year-end urban employment figures across all municipal jurisdictions.	input indicators
Total fixed asset investment I2	Measure using total social asset investment with 2013 as the base year.
Total water consumption I3	Measurements are based on the total water consumption of each city.
Energy consumption I4	Measure using the total energy consumption of industrial enterprises above a certain size in urban areas.
GDP O1	Select cities based on their gross domestic product at the end of 2013 as the base period.	output indicators
Per capita park green space area O2	Measured by park green space area/year-end permanent population.
Social development index O3	Select factors that reflect people’s sense of well-being, such as culture, transportation, healthcare, income inequality, population growth, education, and population structure.
Sewage discharge O4	Measurements are based on the annual industrial wastewater discharge volume of each city.

.

(2)

Model calculation

The SBM model proposed by Tone et al. [36] directly handles input and output relaxation variables to eliminate biases and effects caused by radial and angular selection differences. The specific formula is as follows:

$$\rho =min{\displaystyle \frac{1-\frac{1}{N}{\displaystyle {\sum }_{n=1}^N\frac{s_n^x}{x_{k^{\prime }n}^{t^{\prime }}}}}{1+\frac{1}{M+I}({\displaystyle {\sum }_{m=1}^M\frac{s_m^y}{y_{k^{\prime }m}^{t^{\prime }}+{\displaystyle {\sum }_{i=1}^I\frac{s_i^b}{y_{k^{\prime }i}^{t^{\prime }}}}})}}}$$

(1)

$$x_{k^{\prime }n}^{t^{\prime }}={\displaystyle {\sum }_{t=1}^T{\displaystyle {\sum }_{k=1,k\ne k^{\prime }}^K{\lambda }_k^t{x}_{kn}^t}+s_n^x,n=1,2....N}$$

(2)

$$y_{k^{\prime }m}^{t^{\prime }}={\displaystyle {\sum }_{t=1}^T{\displaystyle {\sum }_{k=1,k\ne k^{\prime }}^K{\lambda }_k^t{y}_{km}^t}-s_m^y,m=1,2....M}$$

(3)

$$b_{k^{\prime }i}^{t^{\prime }}={\displaystyle {\sum }_{t=1}^T{\displaystyle {\sum }_{k=1,k\ne k^{\prime }}^K{\lambda }_k^t{x}_{ki}^t}+s_i^b,i=1,2....I}$$

(4)

$${\lambda }_k^t\ge 0,s_n^x\ge 0,s_m^y\ge 0,s_i^b\ge 0,k=1,2...K$$

(5)

where $\rho$ is the water resource green efficiency value; $x_{k^{\prime }n}^{t^{\prime }}$, $y_{k^{\prime }m}^{t^{\prime }}$, and $b_{k^{\prime }i}^{t^{\prime }}$ are the input, desirable output, and undesirable output of the kth region in the tth year; k is the number of research objects (k = 8 in this study); $s_n^x$, $s_m^y$ and $s_i^b$ are the slack variables of the nth input, mth desirable output, and ith undesirable output; ${\lambda }_k^t$ is the weight vector; $x_{kn}^t$, $y_{km}^t$, and $b_{ki}^t$ are the input, desirable output, and undesirable output variables for the k^′th region in the t^′th year; $x_{k^{\prime }n}^{t^{\prime }}$, $y_{k^{\prime }m}^{t^{\prime }}$, and $b_{k^{\prime }i}^{t^{\prime }}$ are the input, desirable output, and undesirable output variables for the $k^{\prime }$th region in the $t^{\prime }$th year; $s_n^x$, $s_m^y$, and $s_i^b$ are the slack variables for the nth input variable, the mth desirable output, and the ith undesirable output, respectively; N, M, and I are the number of input variables, expected output variables, and non-expected output variables, respectively; T is the number of years in the study period (T = 11 in this study). When 0 < $\rho$ < 1, there is efficiency loss in the decision-making unit, which can be improved by adjusting the inputs and outputs to enhance the green utilization efficiency of water resources; when $\rho$ ≥ 1 and $s_n^x$ = $s_m^y$ = $s_i^b$ = 0, the decision-making unit efficiency reaches an effective value, the green utilization efficiency of water resources is effective, and the allocation of water resources is reasonable.

3.2. Water Resource Green Efficiency Analysis Model Based on the Malmquist Index

The global Malmquist productivity index was initially developed by Pastor and Lovell [37], which incorporates all observation periods and observation units into a global production technology set, specifically expressed as follows:

$$M_C^G(x^t,y^t,x^{t+1},y^{t+1})={\displaystyle \frac{D_C^G(x^{t+1},y^{t+1})}{D_C^G(x^t,y^t)}}$$

(6)

In the formula, $M_C^G$ represents the global Malmquist index. When $M_C^G$ > 1, it indicates an increase in total factor productivity; when $M_C^G$ < 1, it indicates a decrease in total factor productivity. $D_C^G(x^{t+1},y^{t+1})$ and $D_C^G(x^t,y^t)$ represent the global efficiency values for periods t and t + 1, respectively. The Malmquist index can be used to decompose the total factor productivity change index into an efficiency change index (EC_c) and a technological progress index (BPC_c):

$$\begin{array}{l}{M}_C^G(x^t,y^t,x^{t+1},y^{t+1})={\displaystyle \frac{D_c^{t+1}(x^{t+1},y^{t+1})}{D_c^t(x^t,y^t)}}\times \left\{{\displaystyle \frac{D_c^G(x^{t+1},y^{t+1})}{D_c^{t+1}(x^{t+1},y^{t+1})}}\right.\times \left.{\displaystyle \frac{D_c^t(x^t,y^t)}{D_c^G(x^t,y^t)}}\right\}\\ ={\displaystyle \frac{T{E}_c^{t+1}(x^{t+1},y^{t+1})}{T{E}_c^t(x^t,y^t)}}\times \left\{{\displaystyle \frac{D_c^G(x^{t+1},\frac{y^{t+1}}{D_c^{t+1}(x^{t+1},y^{t+1})})}{D_c^G(x^t,\frac{y^t}{D_c^t(x^t,y^t)})}}\right\}\\ =E{C}_c\times \left\{{\displaystyle \frac{BP{G}_c^{G,t+1}(x^{t+1},y^{t+1})}{BP{G}_c^{G,t}(x^t,y^t)}}\right\}\\ =E{C}_c\times B{P}_c\end{array}$$

(7)

3.3. An Analytical Framework for Assessing Determinants of Water Resource Sustainability Using Tobit Regression

Based on existing research findings [38], this study categorizes the principal determinants influencing ecological water use performance in Shaanxi Province into four key dimensions: (1) economic development indicators, (2) natural resource availability, (3) water consumption patterns, and (4) sectoral composition. Subsequently, censored regression modeling (Tobit) was implemented to assess the partial effects of these variables on provincial hydrological sustainability metrics. The analytical results demonstrate the following:

This study employs GDP per capita to measure regional economic development, incorporates annual precipitation and per capita water availability to evaluate natural water endowment, utilizes per capita water consumption to assess utilization patterns, and adopts the share of secondary industry output to quantify industrial structure characteristics.

$${y_i}^{*}=x_i\beta +{\epsilon }_i,{\epsilon }_i\sim N(0,{\sigma }^2)$$

(8)

$$y_i=\left\{\begin{array}{c}{y}^{*}=x_i\beta +{\sigma }^2,y^{*}>0;\\ 0,y^{*}\le 0\end{array}\right.$$

(9)

In the equation, ${y_i}^{*}$ is the original dependent variable; $x_i$ is the independent variable; $\beta$ is the regression parameter; and the error term ${\epsilon }_i$ is independent and follows a normal distribution.

In order to further explore how key influencing factors affect the green efficiency of water resources in Shaanxi Province over time, this paper uses a geotemporally weighted regression (GTWR) model for calculation, as follows:

$$Y_i={\beta }_0(u_i,v_i,t_i)+{\displaystyle {\sum }_k}{\beta }_k(u_i,v_i,t_i)X_{ik}+{\epsilon }_i$$

(10)

In the equation, $Y_i$ is the dependent variable, i.e., water resource green efficiency; $(u_i,v_i,t_i)$ are the spatio-temporal coordinates of the ith observation; ${\beta }_0(u_i,v_i,t_i)$ is the constant term of the regression; ${\beta }_k(u_i,v_i,t_i)$ is the regression coefficient of the kth variable at the ith observation point; $X_{ik}$ is the value of the kth variable at the ith observation point; ${\epsilon }_i$ is the random error of the ith observation point.

4. Findings and Interpretation

4.1. Quantitative Assessment and Spatio-Temporal Dynamics of Ecological Water Use Performance

Employing identical input metrics, this study classifies water use productivity into three distinct categories based on output characteristics: economic water productivity, environmental water performance, and integrated water sustainability efficiency. Economic efficiency aims to maximize economic output, with output indicator O₁ as the primary focus. Environmental efficiency considers environmental impacts on top of economic efficiency, with desired output indicators being O₁ and O₂, and undesirable output indicators being O₄. Green efficiency pursues economic and environmental efficiency while also considering social and human sustainable development. This paper uses O₁, O₂, and O₃ as desired output indicators and O₄ as an undesirable output indicator. The trends in the values of economic efficiency, environmental efficiency, and Figure 3 presents the ecological water use performance metrics for Shaanxi Province during the 2013–2023 period.

Temporally, Shaanxi’s ecological water use performance remained suboptimal throughout the 2013–2023 period, yet it demonstrated consistent improvement, rising from 0.460 to 0.843. This decade-long progression exhibited three distinct phases: The initial phase (2013–2014) displayed an inverse efficiency hierarchy where economic water productivity exceeded composite sustainability metrics, which in turn surpassed environmental water performance. Notably, the integrated sustainability index reached its nadir in 2014, potentially reflecting policy implementation challenges following China’s 2011 Central Document No. 1. Although provincial authorities adopted water conservation technologies in response to the stringent management framework, their continued emphasis on economic expansion over ecological considerations contributed to deteriorating sustainability indicators. Subsequent improvement characterized the intermediate phase (2014–2018), where a reordered efficiency pattern emerged: economic output metrics > environmental performance indicators > comprehensive sustainability measures. This shift coincided with gradual policy internalization and balanced development approaches. This was primarily because this period marked a phase of vigorous economic development in Shaanxi Province, with GDP growing from 1768.994 billion yuan in 2014 to 2443.832 billion yuan in 2018, an increase of 1.38 times. Additionally, after 2014, the national government strongly promoted the concept of green development. The Shaanxi Provincial Government, while pursuing economic development, also placed greater emphasis on green development and improved water resource management methods, leading to water resource environmental efficiency surpassing water resource green efficiency in 2015. The third phase spans from 2018 to 2023. During this phase, water resource green efficiency grew slowly but showed an overall positive trend. After 2020, water resource green efficiency > water resource environmental efficiency > water resource economic efficiency. This is primarily because the Shaanxi Provincial Government has placed particular emphasis on water resource utilization issues in recent years, continuously striving to elevate water resource utilization to new heights. However, due to rapid economic development and the continuous improvement of urbanization levels, water resource utilization pressures have become increasingly evident, and further improvements in water resource green efficiency still face certain challenges.

This study conducted a comprehensive assessment of water use productivity across Shaanxi’s ten municipal districts (2013–2023), evaluating economic output efficiency, ecological conservation effectiveness, and integrated sustainability performance. As illustrated in Figure 4, none of these metrics achieved optimal levels during the observation period, with pronounced geographical variations emerging. Northern Shaanxi exhibited superior economic water productivity, attributable to its strategic position as China’s energy and chemical production hub. The predominance of extractive industries (coal, petroleum, and natural gas) in regional GDP reflects capital-intensive operations with elevated energy demands per unit output, yet yielding substantial resource-to-economic conversion rates. Conversely, southern districts demonstrated exceptional environmental conservation outcomes, benefiting from advantageous hydrological conditions along the Qinling Mountains’ southern foothills. The region’s industries are primarily green agriculture and ecological tourism, with minimal industrial pollution, low energy consumption intensity, and relatively low non-desired wastewater emissions. Compared to the northern and southern regions of Shaanxi, the Guanzhong region has lower water resource green efficiency. This is primarily because the Guanzhong region is the industrial core of Shaanxi Province, with a concentration of water-intensive industries, leading to a mismatch between water consumption and economic output. Additionally, the Guanzhong urban agglomeration has a dense population, with domestic sewage and industrial wastewater overlapping, and wastewater treatment facilities may be insufficient or outdated, resulting in non-desired outputs (wastewater emissions) significantly lowering efficiency values.

To enhance visual interpretation of regional variations in ecological water use performance across Shaanxi, we applied the Jenks natural breaks classification to categorize municipal-level efficiency scores into five tiers: low (0.000–0.142), moderately low (0.142–0.297), intermediate (0.297–0.536), moderately high (0.536–0.938), and high (0.938–1.200) performance levels. This graded representation appears in Figure 5.

Regional analysis of Shaanxi’s ecological water use performance reveals distinct geographical patterns. Southern districts consistently maintained above-average sustainability metrics, with most areas achieving moderately high to high performance tiers throughout the observation window. Specifically, Ankang municipality sustained top-tier status after 2016. This superior performance stems from the region’s advantageous hydrological conditions within the Yangtze River watershed, where ample precipitation ensures reliable water availability, confirming the positive correlation between natural water availability and sustainable usage metrics. The central Guanzhong plain displayed a capital-centric efficiency distribution. Xi’an municipality demonstrated progressive improvement (2013–2023), reaching high performance thresholds by 2020. Conversely, neighboring Xianyang and Weinan urban areas persisted in lower efficiency categories, potentially influenced by constrained resource bases, suboptimal geographic positioning, and industrial composition challenges. Northern territories, situated within the arid Loess Plateau ecoregion, faced compounded ecological constraints including pronounced soil degradation and limited surface water reserves. These challenges manifested in unstable efficiency metrics during 2014–2019, predominantly in below-average ranges. However, recent policy interventions addressing water governance optimization and equitable allocation mechanisms have yielded measurable improvements, with northern regions exhibiting consistent efficiency gains post-2019, validating the effectiveness of regulatory interventions.

4.2. Productivity Change Assessment of Ecological Water Use Performance Using Malmquist Index Methodology

To examine temporal variations in water use productivity across Shaanxi Province, this research combines static slack-based measure (SBM) evaluations with dynamic Malmquist productivity assessments, with comprehensive analytical results visualized in Figure 6.

Figure 6 illustrates that Shaanxi Province achieved a mean Malmquist productivity score of 1.176 for ecological water use performance between 2013 and 2023, corresponding to a 17.6% yearly improvement. The growth pattern displayed periodic fluctuations, peaking at 1.397 during the 2017–2018 interval before a temporary downturn in 2018–2019—a pattern corroborated by SBM model outputs. Disaggregating the index into efficiency change (EC_c) and technological change (BPC_c) components yields average values of 1.017 and 1.099, respectively, with technological advancement consistently outpacing operational efficiency gains throughout the observation window. These findings demonstrate substantial advancements in water-related technological innovation across Shaanxi Province during the observation window, with annual comparisons indicating gradual improvements in operational efficiency metrics. Nevertheless, a discernible implementation gap persists between technological development and practical deployment. This asynchronous progression between innovation capabilities and application effectiveness underscores a critical challenge in contemporary water governance: suboptimal technology diffusion mechanisms hinder the complete translation of innovative solutions into tangible productivity gains, ultimately limiting the enhancement potential for comprehensive water sustainability performance.

The Malmquist indices and their decompositions for cities in Shaanxi Province are visualized as shown in Figure 7. The Malmquist indices for water resource green efficiency across the 10 prefecture-level cities in Shaanxi Province exhibit significant spatial heterogeneity. The analysis period witnessed enhanced ecological water use performance across most urban centers in Shaanxi, with the exception of Tongchuan and Shangluo, where Malmquist productivity scores below 1 reflected gradual efficiency deterioration. Notably, while all municipalities demonstrated technological advancement (BPC_c > 1), five cities in the Guanzhong area—Tongchuan, Baoji, Xianyang, Weinan, and Shangluo—exhibited operational efficiency deficits (EC_c < 1), revealing substantial implementation gaps between innovation adoption and practical execution. Particularly striking progress occurred in northern territories, where Yulin and Yan’an registered exceptional annual efficiency gains of 22.4% and 24.2%, respectively. This is attributed to the synergistic development of technical efficiency and technological progress in these two cities. Moreover, low technical efficiency not only constrains overall efficiency improvements but may also lead to resource waste and environmental pollution. Therefore, Shaanxi Province needs to adjust and optimize its management strategies, operational practices, and policy orientations to enhance technical efficiency, ensuring that technological progress can better promote the improvement of water resource green efficiency.

4.3. Determinants of Ecological Water Use Performance

Statistical analysis software was employed to conduct censored regression modeling examining ecological water use performance and its determinants across Shaanxi Province (2013–2023), with detailed outputs presented in

Table 2. Tobit regression results for factors affecting water resource green efficiency.

Variable	Coefficient of Regression	T-Statistic	p > $\left|\mathit{t}\right|$
Per capita GDP (X1)	0.0686617	8.82	0.000
Annual precipitation (X2)	0.0014405	2.96	0.004
Water availability per capita (X3)	0.8916374	4.15	0.000
Per capita water consumption (X4)	−0.0018857	−8.03	0.000
Water use per inhabitant (X5)	−1.447362	−4.82	0.000
Constant term	0.971841	8.25	0.000

.

Economic analyses indicate that X1 maintains a robust positive correlation with watershed sustainability indicators, implying that provincial economic expansion facilitates enhanced water governance efficacy. Metropolitan areas with higher per capita economic output consistently display stronger composite development indices, which translate to measurable gains in aquatic ecosystem stewardship. Structural decomposition demonstrates that such economically vibrant cities leverage three key advantages—(1) geographically clustered industries, (2) advanced technical competencies, and (3) well-developed civic infrastructure—collectively generating synergistic efficiencies in water resource applications. These progressive urban systems further institutionalize circular optimization through refined allocation protocols and adaptive management regimes.

Regarding hydrological endowment factors, variables X2 and X3 both contribute positively to ecological water use performance, though per capita water availability exhibits a more substantial regression coefficient compared to yearly rainfall metrics. Shaanxi’s hydrological challenges stem from disproportionate rainfall distribution, where arid northern/central districts contrast with precipitation-rich southern zones. Crucially, despite higher rainfall volumes in southern territories, insufficient hydraulic infrastructure (including storage reservoirs and precipitation collection systems) constrains the transformation of rainfall into accessible water supplies, ultimately restricting its potential enhancement of sustainable water management metrics.

In terms of water use factors, X4 has a negative impact on water resource green efficiency. This may be because Shaanxi is located in the arid–semi-arid region of Northwest China, where per capita water resources are below the national average. An increase in per capita water consumption may be accompanied by an increase in undesirable outputs (such as wastewater discharge), thereby reducing water resource green efficiency.

Regarding sectoral composition, variable X5 demonstrates a substantial adverse effect on ecological water use performance, evidenced by its considerable regression coefficient. This likely reflects Shaanxi’s industrial profile as a western China manufacturing hub, where the secondary sector disproportionately features water-intensive, high-pollution activities including energy extraction (coal/petroleum), metallurgical processing, and construction materials production. These industries exhibit markedly greater hydrological resource demands per economic output compared to service or technology sectors, with their growth consequently depressing regional sustainable water management metrics.

This study applies geographically and temporally weighted regression (GTWR) methodology to examine temporal dynamics in determinant factors affecting aquatic ecological performance across Shaanxi Province. Figure 8 illustrates the evolving regression coefficients for three geographical divisions—northern, central, and southern zones—throughout the observation period. Initial analysis reveals diminishing economic output elasticity regarding water sustainability metrics in northern territories during the 2013–2015 phase. The 2015–2017 phase witnessed substantial enhancement in economic output sensitivity within northern Shaanxi. Subsequently, during 2017–2022, northern territories demonstrated stronger economic growth elasticity regarding aquatic sustainability metrics compared to central and southern zones. During the initial study phase (2013–2016), hydrological efficiency in the central plains demonstrated stronger sensitivity to rainfall variations compared to southern mountainous areas. During the 2016–2021 observation window, precipitation elasticity for sustainable water metrics was more pronounced in southern Shaanxi compared to northern and central zones. From 2013 to 2023, the impact coefficient of per capita water resources on water resource green efficiency in the Shaanbei and Shaanan regions showed a fluctuating downward trend, while the impact coefficient in the Guanzhong region remained relatively stable around 1. After 2017, the impact of per capita water consumption on green efficiency in the Shaanbei region was greater than in the Shaanan and Guanzhong regions, with a weakening trend in the impact on the Shaanbei region. During the period from 2013 to 2023, the impact coefficient of the secondary industry’s share on water resource green efficiency in the Shaanbei and Guanzhong regions showed an increasing trend, while the impact coefficient in the Shaanan region fluctuated around −1. Regions in Shaanxi Province should adopt tailored approaches based on the varying degrees of influence of different factors, adjusting their priorities in a timely manner to more effectively enhance water resource green efficiency.

5. Discussion

5.1. Rationality Analysis of Green Efficiency of Water Resources

Employing the Meta-frontier Super-SBM approach combined with Tobit regression, Zheng et al. [39] assessed urban water resource utilization efficiency (UWRUE) and integrated comprehensive water resource utilization capacity (CWRUC) across 53 municipalities in China’s Yellow River Basin (2003–2020). Their analysis identified several underperforming urban centers including Xi’an and Xianyang. Our computational results reveal these cities’ ecological water efficiency scores of 0.755 and 0.320, respectively, both falling below optimal operational thresholds. Cheng Z et al. [40] studied the water resource utilization efficiency of various provinces in China from 2011 to 2020 based on the Super-SBM model and found that the water resource utilization efficiency of Shaanxi Province was approximately 0.75. The results of this study show that the green efficiency of water resources in Shaanxi Province was 0.609 during the period from 2013 to 2023. The results are reasonable and reliable when compared with existing literature.

Geographically situated in China’s northwest, Shaanxi has demonstrated consistent growth in ecological water use performance, with its efficiency metrics rising from 0.460 to 0.843 over the 2013–2023 period. Shao et al. [41] identified consistent improvements in municipal water use productivity across China, with western territories achieving a mean efficiency score of 0.886 and northern urban areas attaining 0.908 in 2018. Comparative analysis reveals Shaanxi’s ecological water performance initially lagged behind central-western regional benchmarks, though progressive policy interventions have systematically reduced this disparity through enhanced hydrological governance.

An intra-provincial analysis of Shaanxi reveals that southern districts exhibited marginally superior sustainable water use performance compared to northern and central zones during the examined period, attributable to their comparatively favorable hydrological conditions. Wang and Chen [42] posit a pronounced ‘resource curse’ phenomenon in central-western China, wherein natural asset abundance constrains ecological performance growth. This study investigates ecological water use performance in Shaanxi Province, revealing insufficient empirical evidence to substantiate the ‘resource curse’ hypothesis in local hydrological systems. The dual mechanism of resource advantages and their paradoxical constraints, as conceptualized by Shao et al. [41], offers a robust theoretical framework for interpreting these findings.

5.2. Malmquist Index Rationality Analysis

Utilizing the Malmquist index approach, this study examined the ecological efficiency of water resource utilization in Shaanxi Province, demonstrating an average yearly increase of 17.6%. A decomposition of the Malmquist index revealed that the technical efficiency index (ECc) was lower than the technical progress index (BPCc), indicating a significant lag in the conversion between technological innovation and practical application in Shaanxi Province. The findings align with the research by Jin et al. [24], which analyzed industrial water efficiency across 30 Chinese provinces (2000–2016) and suggested that technological advancements alone may not significantly enhance water use productivity. Ozturk and Cinperi [43] also proposed that while the implementation of innovative technologies may improve water use efficiency, limited technological effectiveness remains a major constraint on optimal water resource management.

The average technical efficiency index for Shaanxi Province is 1.017, while the average technical progress index is 1.099. This may be related to Shaanxi Province’s relatively underdeveloped economy and water scarcity. Most industries in Shaanxi Province are focused on achieving rapid economic growth, which often involves high water consumption, energy use, and pollution levels. Recent policy initiatives in Shaanxi Province have prioritized water resource preservation, yielding measurable progress in the development of energy-efficient technologies and related infrastructure. However, stringent regulatory requirements and increased pollution control costs have compelled many enterprises to invest substantial resources in these areas. As a result, most enterprises continue to maintain their previous production practices to easily achieve higher profits, leading to a significant disparity between the technological progress index and the technological efficiency index, with technological innovations struggling to be translated into practical applications. Similar situations are relatively common in western China [24].

5.3. An Examination of Determinants Influencing the Ecological Efficiency of Water Resource Utilization

Regarding the determinants of ecological water use efficiency, this research reveals that sectoral composition and natural water availability exert substantial influence on sustainable water resource utilization in Shaanxi Province. Wang S et al. [16] analyzed an examination of water use productivity across China’s provincial regions (2008–2016) and demonstrated comparatively lower efficiency levels in central and western areas, where sectoral composition exerted a notably adverse effect.

This study incorporates the Human Development Index (HDI) into the calculation of Shaanxi Province’s water resource green efficiency. The mechanisms through which the HDI’s dimensions—including education, healthcare, and health—influence Shaanxi Province’s water resource green efficiency can be analyzed from three perspectives: population structure, technological innovation, and enhanced environmental awareness. First, improvements in educational levels (such as adult literacy rates and higher education enrollment rates) directly enhance the workforce’s ability to learn and apply water-saving technologies, promoting the optimization of irrigation systems and the adoption of water circulation technologies (e.g., the coverage rate of drip irrigation technology in the Xianyang region is significantly positively correlated with the average years of education per capita). Second, the improvement of the healthcare system reduces health-related losses among workers (e.g., the prevention and control of endemic diseases has increased effective working hours by 23% in the Hanzhong region), ensuring stable production capacity output for water-intensive industries. Third, improvements in social welfare drive upgrades in environmental quality demands, prompting the government to implement stricter water resource regulation policies. The underlying transmission mechanism follows a chain-like path of “HDI improvement → human capital appreciation → technological efficiency improvement → alleviation of resource misallocation,” but due to the uneven development within Shaanxi Province (the HDI difference between the Shaanbei energy region and the Guanzhong urban agglomeration reaches 0.28), this promotional effect exhibits a gradient effect, and in counties with weak institutional soft environments, a phenomenon of diminishing marginal returns may occur.

5.4. Model Applicability and Research Advantages

(1)

Applicability of the model

The slack-based measure (SBM) approach serves as a fundamental methodology for assessing water use performance and has been extensively employed in studies evaluating the efficiency of water resource management. However, the SBM model is only applicable for static studies of DMUs within specific years [44]. In order to analyze temporal variations in ecological water use efficiency across Shaanxi Province, the current research employs the Malmquist index methodology. Previous scholars have applied the Malmquist productivity index and its extensions, such as the ML index and GML index, in various efficiency evaluation studies [45,46,47]. In the final stage of analysis, the Tobit regression approach was employed to examine how different determinants influence the ecological efficiency of water utilization in Shaanxi Province. The Tobit model is primarily used to address statistical analysis issues in cases of truncated data. Given that Shaanxi Province’s water resource ecological efficiency scores range from 0 to 1, the Tobit regression model was chosen as the appropriate analytical method for this investigation. In summary, the SBM model, Malmquist productivity index analysis, and Tobit model are all applicable to this study.

(2)

Research strengths

In assessing the ecological efficiency of regional water resource utilization, establishing an appropriate indicator system requires consideration of local environmental endowments, socioeconomic factors, and study purposes. The slack-based measure (SBM) approach employed in this research demonstrates cross-regional applicability, featuring adaptable metric configurations that can evolve with emerging water governance challenges and changing development priorities, ensuring both methodological rigor and long-term relevance.

In examining the determinants of ecological water use efficiency, this study identifies key metrics across four dimensions—economic development level, natural resource availability, consumption patterns, and sectoral composition—tailored to Shaanxi Province’s regional characteristics. For different research regions, the selected influencing factors should be determined according to the actual conditions of the region. If other regions are to be studied, it is only necessary to replace the influencing factors with appropriate ones based on the characteristics of the region to ensure the specificity of the model.

While the focus of this study is water resource green efficiency, its research approach and methodology hold certain reference value for efficiency evaluation in other fields. For instance, in regional energy green efficiency research, the SBM model can be adopted to calculate efficiency values, examine the effects of multiple determinants on performance, and develop tailored interventions to improve operational effectiveness.

5.5. Analysis of Research Limitations

In assessing the ecological performance of water utilization across Shaanxi Province, the present research incorporated socioeconomic indicators alongside conventional economic and environmental parameters to optimize measurement precision. Shaanxi Province was divided into the northern, southern, and central regions for detailed analysis of their temporal and spatial evolution. However, there are still some shortcomings: In terms of data limitations, this study selected four input indicators—labor force, total social fixed-asset investment, total water consumption, and energy consumption—and four output indicators—GDP, per capita park green space area, social development index, and wastewater discharge volume. Given gaps in historical records necessitating data interpolation for certain years, the findings may contain marginal estimation errors. The indicator system was constructed by identifying eight key variables that reflect Shaanxi’s regional particularities while drawing upon established research frameworks. In future studies, more representative indicators could be added to calculate water resource green efficiency, making the results more universally applicable.

6. Research Conclusions and Outlook

6.1. Research Conclusions

This research developed a systematic and rigorous assessment framework for evaluating ecological water use efficiency. The slack-based measure (SBM) approach was employed to quantify provincial water resource performance in Shaanxi between 2013 and 2023, with subsequent examination of its spatio-temporal patterns. For analyzing temporal efficiency variations, the Malmquist productivity index was utilized and decomposed into two components: efficiency change (ECc) and technological change (BPCc). Furthermore, Tobit regression modeling was conducted to identify key determinants influencing sustainable water utilization, offering policy-relevant insights for enhancing water management. The principal findings are summarized below:

(1)

Temporally, the 2013–2023 period witnessed suboptimal yet gradually improving performance in Shaanxi Province’s water-related economic output, ecological conservation, and integrated sustainability metrics. The composite sustainability index rose from 0.460 to 0.843 over this decade. Spatially, regional disparities emerged clearly: northern Shaanxi demonstrated superior economic productivity in water usage, southern areas excelled in ecological conservation effectiveness, while the central Guanzhong plain lagged behind in comprehensive water resource sustainability indicators.

(2)

During the 2013–2023 period, Shaanxi’s water sustainability productivity demonstrated a mean Malmquist score of 1.176, reflecting a 17.6% yearly improvement. Disaggregating this measure into efficiency change (ECc) and technological advancement (BPCc) components showed consistent superiority of technological gains over operational efficiency enhancements, suggesting substantial delays in implementing innovative solutions across the province’s water management systems.

(3)

Regression analysis reveals substantial impacts from industrial composition and water availability, with the proportion of secondary sector output and per capita water reserves exhibiting significant coefficients of −1.447 and 0.892, respectively, in Shaanxi’s water sustainability performance model.

6.2. Pathways to Improving Water Resource Green Efficiency

The Shaanbei region has relatively poor water resources and a high proportion of energy and chemical industries, which consume large amounts of water and cause significant pollution. Therefore, the Shaanbei region should actively promote the transformation of the energy and chemical industries toward high-end and low-carbon development, and develop water-saving emerging industries. Structural economic optimization can simultaneously decrease water intensity in economic output and strengthen investment potential for water conservation technologies alongside GDP growth, ultimately boosting the ecological water use efficiency of northern Shaanxi. Northern Shaanxi should further account for its mineral reserves and ecological constraints when formulating development strategies, systematically refining extraction patterns and spatial distribution to capitalize on sectoral strengths while increasing recycled mine water utilization. Regulatory agencies may establish tiered water pricing mechanisms targeting water-intensive sectors like coal extraction and fossil fuel power plants. While lowering water usage improves sustainable water management efficiency, policymakers must mitigate potential industrial output reductions through technological innovation to sustain economic productivity.

The Shaanxi South region accounts for over 70% of the province’s water resources, but issues with inefficient agricultural irrigation are quite severe. Consequently, proactive adoption of precision irrigation systems (e.g., automated drip irrigation), coupled with supportive measures including crop cultivation subsidies, should be prioritized to enhance water productivity in agricultural operations. Additionally, it is important to reasonably plan and construct water source projects such as reservoirs, mountain ponds, and water storage pits to enhance water resource storage capacity and improve the stability and reliability of water supply. Promoting watershed conservation programs and tradable water entitlement systems enables the transformation of irrigation efficiency gains into financial returns, simultaneously elevating water quality standards and stimulating economic growth in southern Shaanxi, which collectively enhances the ecological performance of regional water resource utilization.

The Guanzhong region needs to establish a multi-dimensional policy framework, focusing on three key areas: water conservation as a priority, industrial optimization, and systematic governance. In the agricultural sector, efforts should be made to comprehensively promote the upgrading of intelligent irrigation systems in high-standard farmland, combined with the promotion of water-fertilizer integration technology and dryland agriculture, to achieve simultaneous reduction in irrigation water demand and improvement in water productivity metrics, industrial operations—particularly in energy-intensive and chemical manufacturing sectors—must adopt circular economy principles under rigorous water governance frameworks. This approach not only reduces total water consumption and energy input but also significantly reduces wastewater discharge. Consequently, this approach significantly improves the ecological utilization efficiency of water resources across the Guanzhong Plain.