The Impact of Digital Technology on Water Resources Management: Evidence from China

Abstract

Digital technology is gradually emerging as a new driving force in the field of water resources management. In this paper, we conduct a thorough analysis of panel data from 30 provinces in China spanning from 2013 to 2022. Utilizing the fixed-effects model, the mediation effect model, a panel threshold model, and a coupling coordination degree model, this study empirically examines the impact of digital technology on water resources management. The findings are as follows: (1) The direct impact of digital technology on water resources management is significantly positive at the 1% level, with notable regional variations. (2) Digital technology improves water management through green innovation. (3) In the process of digital technology promoting water resources management, green innovation exhibits a threshold effect, with an estimated threshold value of 1.840. (4) During the sample period, the national coupling coordination degree of digital technology and water resources management was barely coordinated, showing the following characteristics: Eastern China > Western China > Central China. These research conclusions will offer valuable insights and directions for advancing sustainable water resources management strategies and fostering the deep integration of digital technology and water resources management.

1. Introduction

Water, a precious and indispensable natural resource on earth, serves as the cornerstone of economic and social development and is pivotal in maintaining ecological health [1]. Integrated water resources management strives to sustainably and effectively manage land, water, and other related resources more effectively in a sustainable way [2]. Nevertheless, there is a formidable challenge posed by safeguarding natural ecosystems while fulfilling fundamental human needs and navigating the myriad conflicts surrounding these resources. The United Nations Water Scarcity report estimates that by 2025, 1.8 billion people will live in water-scarce regions, highlighting the urgent need for innovative water resources management. As the largest developing country in the world, China is also one of the countries with severe shortages of water resources per capita, a situation wherein unreasonable utilization of water resources and water pollution are widespread. Additionally, climate change is exacerbating water scarcity [3]. Currently, research on water resources management primarily focuses on three aspects. At the macro level, scholars have conducted comprehensive evaluations of water resources carrying capacity and water management performance [4,5]. From the spatio-temporal perspective, scholars have further explored the balance distribution of water resources and the phenomenon of decoupling, stating that the distribution of water resources in China shows a trend of a transition from “highly balanced” to “highly unbalanced” from east to west [6]. At the micro-level, traditional water resources management approaches encounter difficulties in real-time data collection and analysis and smart decision-making. However, artificial intelligence and big data play pivotal roles in water quality monitoring and water supply-and-demand forecasting [7]. Evidence shows that innovative water resources management requires collaborative efforts regarding technology [8], funding [9], policies and regulations [10], and other aspects.

In today’s world, the process of globalization is accelerating rapidly, and a digital information technology revolution led by a new round of technological and industrial revolution is sweeping across the globe [11]. Also, research on digital technology is increasingly abundant. The connotations and evolution of digital technology indicate a collection and paradigm of various intelligent and innovative technologies used for connectivity, communication, and automation [12]. The German government introduced the Future Cision for the Development of Digital Technology in 2013, known as the “paradigm” of Industry 4.0. Nowadays, the scope of digital technology has expanded from data-based technologies such as cloud computing and artificial intelligence to technologies with physical manifestations, gradually forming a comprehensive technological application system [13]. According to the Digital China Development Report (2023), in 2022, China filed nearly 32,000 PCT international patent applications in the information sector, accounting for 37% of the global total [14]. Against the backdrop of the digital transformation across industries, digital-related topics such as digitalization, the digital economy, and digital finance have become popular [15,16,17]. For example, Costa et al. used a combination of qualitative and quantitative methods to conclude that digital technology offers significant advantages in improving crop yields and reducing environmental footprints [18]. Additionally, the relationship between digital technology and natural resources has attracted the attention of many scholars. Digital technology can play an active role in improving production efficiency, reducing the demand for natural resources, decreasing the use of non-renewable resources, and enhancing the effectiveness of environmental monitoring [19]. However, the long-term effects of digital transformation on environmental sustainability remain uncertain, with this process being described as a potential “accelerator”. In the initial stages, digitalization directly increases resource and energy consumption without necessarily improving energy efficiency [20]. Furthermore, the abundance of natural resources also influences the pace of digital transformation. Studies have shown that regions with abundant resources tend to have relatively weaker digital transformation efforts [21].

Digital technologies centered on artificial intelligence, big data, cloud computing, and 5G base stations are leading the way in making profound changes to the water resources management models of China, offering new ideas for reshaping the governance forms and patterns of water resources [22]. The application of digital technology has significantly enhanced the capability to acquire, process, and analyze water resources information, making it possible to refine the management of water resources [23]. What is more, digital technology facilitates information exchange and cooperation between water resources management organizations, enterprises, and the public, breaking down information silos and fostering a favorable multi-stakeholder governance landscape. In this process, the information and knowledge spillover effects of green technologies are fully unleashed [24]. However, despite the promising application prospects of digital technology in water resources management, issues such as its effectiveness, underlying mechanisms, and impact differences at various stages remain to be explored in depth. Therefore, an in-depth study of the relationship between digital technology and water resources management is of great significance for achieving sustainable and efficient comprehensive water resources management and helping humans cope with the water resources crisis.

In summary, academia has made substantial achievements in the research on digital technology and water resources management, but few studies have analyzed the interactive relationship between digital technology and water resources management. Therefore, based on the panel data of 30 provinces in China (excluding Tibet) from 2013 to 2022, we constructed an evaluation index system for digital technology and water resources management in China and employed the entropy weight–TOPSIS method to conduct measurements. Subsequently, we utilized the fixed-effects model, mediation effect model, and panel threshold model to uncover the influence effects between digital technology and water resources management. Finally, a coupling coordination degree model is applied to further analyze the coordination relationship between digital technology and water resources management. This paper aims to provide theoretical support and empirical evidence for digital technology to promote water resources management. Compared with existing research, the unique main contributions of this paper are as follows: (1) This paper empirically analyses the impact of digital technology on water resources management. (2) With green innovation as the mechanism variable, this paper analyzes the mechanism of green innovation in mediating the impact of digital technology on water resources management and further explores its threshold effect. (3) We carried out an analysis of the coupling coordination degree between digital technology and water resources management, which will help deepen the understanding of how digital technology empowers water resources management and provide valuable insights for the effective application of digital technology and the sustainable management of water resources.

Hypothesis 1. 

Digital technology can significantly enhance water resources management.

Hypothesis 2. 

Digital technology can enhance the level of water resources management by promoting green innovation.

Hypothesis 3. 

Digital technology presents a green innovation threshold for the development of water resources management.

2. Materials and Methods

2.1. Model Formulation

2.1.1. Entropy Weight–TOPSIS

The entropy weight–TOPSIS method integrates the advantages of both the entropy weight method and the TOPSIS method. It uses the entropy weight method to calculate the weights of each indicator, effectively excluding the interference of subjective biases. Building upon this foundation, it draws upon the core concept of the TOPSIS method, in which evaluation benchmarks are constructed based on the optimal and worst solutions within the decision space. By calculating the relative distances between each evaluation alternative and these two extreme solutions, it quantifies the degree of closeness to the ideal solution, thereby facilitating precise ranking of the alternatives [25]. The corresponding calculation equations are as follows:

$$\mathrm{Positive} \mathrm{indicators}: y_{jn}={\displaystyle \frac{x_{jn}-m\left(x_{jn}\right)}{M\left(x_{jn}\right)-m\left(x_{jn}\right)}} (1<j<i,1<n<m)$$

(1)

$$\mathrm{Negative} \mathrm{indicators}: y_{jn}={\displaystyle \frac{m\left(x_{jn}\right)-x_{jn}}{M\left(x_{jn}\right)-m\left(x_{jn}\right)}} (1<j<i,1<n<m)$$

(2)

$$E_n=\left[-{\displaystyle \frac{1}{ln\left(j\right)}}\right]{\sum }_{j=1}^i\left[{\displaystyle \frac{y_{jn}}{{\sum }_{j=1}^i{y}_{jn}}}\times ln\left({\displaystyle \frac{y_{jn}}{{\sum }_{j=1}^i{y}_{jn}}}\right)\right]$$

(3)

$$W_n={\displaystyle \frac{1-H_n}{{\sum }_{n=1}^m\left(1-H_n\right)}}$$

(4)

$$R=r_{jn}=w_n\times {\displaystyle \frac{y_{jn}}{{\sum }_{j=1}^i{y}_{jn}}}$$

(5)

$$R^{+}=\left[r_1^{+},r_2^{+}{,...,r}_m^{+}\right]$$

(6)

$$R^{-}=\left[r_1^{-},r_2^{-}{,...,r}_m^{-}\right]$$

(7)

$$D^{+}=\sqrt{{\sum }_{n=1}^m{\left(R-R^{+}\right)}^2}$$

(8)

$$D^{-}=\sqrt{{\sum }_{n=1}^m{\left(R-R^{-}\right)}^2}$$

(9)

$$P={\displaystyle \frac{D^{-}}{D^{+}+D^{-}}}\left(0<C<1\right)$$

(10)

In the above equations, M(x_jn) represents the maximum value in the sample, and m(x_jn) represents the minimum value. E_n represents the entropy value of each indicator, and W_n represents the weight of each index. D represents the Euclidean distance. P represents the comprehensive index.

2.1.2. Benchmark Regression Model

To explore the direct impact of digital technology on water resources management, by referring to the practices of Wang et al. [26], the following benchmark regression model was constructed.

$${WRM}_{it}={\alpha }_0+{\alpha }_1{DT}_{it}+\gamma {control}_{it}+{\mu }_i+{\nu }_i+{\epsilon }_{it}$$

(11)

In Equation (11), $i$ represents the province, and $\mathrm{t}$ represents the time period. WRM stands for the dependent variable, water resources management, while DT represents the core explanatory variable, digital technology. ${\alpha }_0$, ${\alpha }_1$, and $\gamma$ represent the estimated coefficients. ${\mu }_i$ is used to control for individual fixed effects, and ${\nu }_i$ is used to control for time fixed effects. ${\epsilon }_{it}$ represents the random disturbance term.

2.1.3. Mediation Effect Model

According to Hypothesis 2, digital technology may have an impact on water resources management through green innovation. By referring to the work by Irfan et al. [27] and Sun et al. [28], the benchmark regression model was constructed.

$${GI}_{it}={\alpha }_0+{\alpha }_1{DT}_{it}+\gamma {control}_{it}+{\mu }_i+{\nu }_i+{\epsilon }_{it}$$

(12)

$${WRM}_{it}={\alpha }_0+{\alpha }_1{DT}_{it}+{\alpha }_2{GI}_{it}+\gamma {control}_{it}+{\mu }_i+{\nu }_i+{\epsilon }_{it}$$

(13)

In Equations (12) and (13), GI represents green innovation, and the meanings of the other variables and parameters are consistent with those noted previously in the text.

2.1.4. Threshold Effect Model

To further examine the relationship between digital technology, green innovation, and water resources management, by referring to the practice of Cheikh and Rault [29], the single-threshold model and the double-threshold model were constructed.

$$WR{M}_{it}={\alpha }_0+{\alpha }_{1}D{T}_{it}\times I\left(th{r}_{it}\le \theta \right)+{\alpha }_{2}D{T}_{it}\times I\left(th{r}_{it}>\theta \right)+\gamma contro{l}_{it}+{\mu }_1+{\nu }_{\mathrm{i}}+{\epsilon }_i$$

(14)

$$\begin{array}{l}WR{M}_{it}={\alpha }_0+{\alpha }_{1}D{T}_{it}\times I\left(th{r}_{it}\le {\theta }_1\right)+{\alpha }_{2}D{T}_{it}\times I\left({\theta }_1\le th{r}_{it}\le {\theta }_2\right)\\ +{\alpha }_{3}D{T}_{it}\times I\left(th{r}_{it}>{\theta }_2\right)+\gamma contro{l}_{it}+{\mu }_1+{\nu }_{\mathrm{i}}+{\epsilon }_i\end{array}$$

(15)

In Equations (14) and (15), thr_it represents the threshold variable; I(·) is an indicator function that takes a value of 1 when the threshold condition inside the parentheses is met and 0 otherwise; θ represents the threshold value for the single-threshold model; θ₁ represents the first threshold value for the double-threshold model; and θ₂ represents the second threshold value for the double-threshold model.

2.1.5. Coupling Coordination Degree Model

The coupling coordination degree model includes the degree of coupling and the degree of coordination, which can be used to measure the degree of interaction between two or more systems. By referring to the practices of Tomal [30], the coupling coordination degree model was constructed.

$$C=2\sqrt{{\displaystyle \frac{U_1\times U_2}{{\left(U_1+U_2\right)}^2}}} \left(0<C<1\right)$$

(16)

$$T=\alpha U_1+\beta U_2$$

(17)

$$D=\sqrt{C\times T} \left(0<D<1\right)$$

(18)

In Equations (16)–(18), U₁ and U₂ represent the digital technology and water resources management levels, respectively; C represents the coupling degree; T represents the comprehensive coordination index of digital technology and water resources management; and α = β = 0.5, indicating digital technology and water resources management are of the same importance. D represents the evaluation result for the coupling coordination degree.

2.2. Variable Indicators and Data Source

2.2.1. Description of Indicators

The dependent variable was water resources management (WRM). Water resources management is a process aimed at balancing water use as a basis for ensuring the livelihoods of a growing population while protecting and preserving water resources to ensure the sustainability of a water supply [31]. Based on the research conducted by Bozorg-Haddad et al. [32] and Maurya et al. [33], an evaluation index system for water resources management was constructed (see

Table 1. Comprehensive index system for WRM [34].

Primary Indicators	Secondary Indicators	Unit
Water resources development and utilization	Water consumption per CNY 10,000 of GDP	Cubic meters/CNY 10,000
	Water consumption per CNY 10,000 of industrial added value	Cubic meters/CNY 10,000
	Per capita water consumption	Cubic meters per person
	Water resources development and utilization rate	%
Water environment governance	Wastewater and sewage discharge per CNY 10,000 of GDP	Cubic meters/CNY 10,000
	Urban sewage centralized treatment rate	%
	Total chemical oxygen demand discharge from wastewater	10,000 tons
	Total ammonia nitrogen emissions from wastewater	10,000 tons
Water ecological restoration	Forest coverage rate	%
	Greening coverage in built-up areas	%
	Groundwater water use rate	%

). Based on Equations (1)–(10), the level of water resources management in China during the sample period was calculated.

The independent variable was digital technology (DT). Based on the availability of data and drawing on the research by Liu et al. [35], an evaluation index system for digital technology was constructed (see

Table 2. Comprehensive index system for DT [36,37].

Primary Indicators	Secondary Indicators	Unit
Fundamentals of digital technology	Internet penetration rate	%
	Mobile phone penetration rate	%
	Length of long-distance fiberoptic cable per unit area	Kilometers per square kilometer
	Mobile base station density	Units per square kilometer
	IPV4 address count	10,000
Digital factors inputs	The proportion of employees in the software and information technology service industry among urban unit employees	%
	R&D expenditure of industrial enterprises above the designated size	CNY 10,000
	Full-time equivalent of R&D personnel in industrial enterprises above the designated size	Person
Digital technology applications	Per capita telecommunication business volume	CNY 10,000 per person
	The proportion of information technology service revenue in GDP	%
	Digital inclusive finance development index	-

). Based on Equations (1)–(10), the level of digital technology development in China during the sample period was calculated.

To make the empirical results more scientifically reliable, we selected the following control variables: urbanization level (City), measured by the proportion of urban population to the total population; economic level (Eco), measured by the logarithm of the Gross Regional Product (GRP) per capita; population density (Pop), expressed as the ratio of the population of each province to the total area; infrastructure (Inf), expressed as the per capita urban road area and the number of urban roads per capita; population density (Pop), expressed as the ratio of the population of each province to the area of each province; and infrastructure (Inf), expressed as the area of urban roads per capita.

The mediator variable was green innovation (GI). Green innovation adheres to the principles of reducing energy consumption and promoting environmental protection, serving as an indicator of the degree of greenness in regional innovation capabilities. Technological innovations that meet the criteria of “green environmental protection and scientific quality” are recognized as green patents, which enable the output of green innovation to be visually quantified. Referring to the study by Chen et al. [38], in this paper, we select the logarithm of green patent grants to represent GI.

2.2.2. Data Source

The data used in this paper is obtained from China Statistical Yearbook [39], China Environmental Statistics Yearbook, China Science and Technology Statistics Yearbook, the China Research Data Services Platform (CNRDS), as well as provincial statistical yearbooks and bulletins. Missing values were filled in through interpolation methods.

3. Results

3.1. Data Description

Descriptive statistics of the variables are presented in

Table 3. Descriptive statistics [40].

Variables	N	Mean	SD	Min	Max
DT	300	0.185	0.142	0.038	0.589
WRM	300	0.565	0.115	0.316	0.825
GI	300	7.756	1.306	3.434	10.722
City	300	0.614	0.114	0.379	0.896
Eco	300	10.947	0.432	10.003	12.156
Pop	300	0.047	0.071	0.001	0.393
Inf	300	2.783	0.360	1.413	3.332

.

Table 3 shows that the mean value of digital technology (DT) is 0.185, the maximum value is 0.589, and the minimum value is 0.038, indicating that there is a significant difference between the development levels of digital technology in China. The minimum value of water resources management (WRM) is 0.316, and the maximum value is 0.825, with a difference of 0.509 between the minimum and maximum values, reflecting a huge disparity in the levels of water resources management among different provinces; the minimum value of green innovation (GI) is 3.434, and the maximum value is 10.722, also indicating an uneven development of green innovation among regions. The three important variables all have relatively large variation ranges, reflecting the developmental disparities among different regions in China.

3.2. Scatter Plots

In order to visualize the correlation trends between digital technology, green innovation, and water resources management, scatter plots and fitting curves were developed, as shown in Figure 1, Figure 2 and Figure 3.

From Figure 1, Figure 2 and Figure 3, it can be observed that any two variables in digital technology, water resources management, and green innovation have positive correlations with each other, thus initially confirming the theoretical logic of the hypotheses presented earlier. However, further empirical tests of the quantitative relationships between relevant variables are needed through econometric models.

3.3. Benchmark Estimation Analysis

We conducted a regression analysis on the impact of digital technology on water resources management, with the specific results shown in

Table 4. Results of the benchmark regression.

Variable	WRM
	(1)	(2)	(3)	(4)
DT	0.287 *** (4.98)	0.236 ** (3.22)	0.252 *** (4.22)	0.221 *** (2.95)
City			−0.416 *** (−2.70)	−0.506 ** (−2.00)
Eco			0.044 ** (2.06)	0.023 (0.51)
Pop			0.017 (0.09)	1.881 (0.67)
Inf			−0.018 (−0.08)	0.024 (1.01)
_cons	0.398 *** (23.39)	0.392 *** (26.84)	1.179 (1.07)	0.252 (0.56)
Province	No	Yes	No	Yes
Year	No	Yes	No	Yes
N	300	300	300	300
R2	0.097	0.186	0.177	0.230

Note: t-Statistics are in parentheses, *** p < 0.01, ** p < 0.05.

.

According to Table 4, column (1) shows the situation without introducing control variables and without controlling for individual and time effects. The estimated coefficient of digital technology is 0.287 and remained significant at the 1% level, indicating that digital technology can significantly promote water resources management, and the estimated coefficient remained significantly positive after controlling for individual and time effects. As shown in column (3), after introducing further control variables such as economic level, environmental regulation, government support, financial level, and human capital, the estimated coefficient of digital technology is 0.252 and significant at the 1% level. Column (4), based on column (3), further controls for individual and time effects. The estimation result shows that the coefficient of digital technology for water resources management is remarkably positive at the 1% level, demonstrating that digital technology has a direct and pronounced positive effect on improving water resources management. The possible reason is that the application of digital technology enhances the effectiveness of the real-time monitoring and intelligent analysis of water resources, effectively eases the contradiction between the supply and demand of water resources, enhances the scientific and efficient allocation of water resources, and has a direct and significant positive impact on the improvement of water resources management; thus, hypothesis 1 is verified.

Nowadays, the development of digital technology is vigorously advocated; thus, this conclusion is consistent with existing research findings. Digital technology has significant advantages in improving water resource allocation efficiency and precision management [41]. From the perspective of digital application effects, Cassidy et al. found that digitization can effectively reduce unprofitable water use through the implementation of optimized algorithms and early-warning systems, elevating water use efficiency to a new level [42]. In summary, digital technology can play a facilitating role in water resources management.

3.4. Endogeneity Test and Robustness Test

An instrumental variable method was employed. In the benchmark regression analysis of how digital technology empowers water resources management mentioned above, there may be bidirectional causality between digital technology and water resources management, leading to endogeneity issues. Referring to the approach employed by Yuan and Pan [43], in this paper, we selected the lagging one-period core explanatory variable of digital technology as an instrumental variable to re-validate the original model. The estimation was conducted using the two-stage least squares method (IV-2SLS), and the results are presented in column (1) of

Table 5. Results of the endogeneity test and robustness test.

Variable	(1)	(2)	(3)	(4)
	WRM	WRM	WRM	WRM
DT	0.636 *** (3.19)	0.222 *** (2.63)	0.230 *** (3.18)	0.215 *** (2.86)
Control	Yes	Yes	Yes	Yes
Kleibergen–Paap rk LM statistic	12.245 ***
Kleibergen–Paap rk Wald F statistic	75.896 ***
Province	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes
N	270	260	300	300
R2	0.028	0.301	0.254	0.227

Note: t-Statistics are in parentheses, *** p < 0.01.

.

The four municipalities directly under the control of the central government were excluded. Considering the unique economic scale, administrative level, and population sizes of municipalities directly under the control of the central government, we re-examined the data after excluding Beijing, Tianjin, Shanghai, and Chongqing, the four municipalities that may cause interference. The results are presented in column (2) of Table 5.

A control variable was added. There are many factors that affect digital technology and water resources management, and omitted variables may lead to endogeneity issues. There are significant differences in the demand for and utilization of water resources among different industries, and the degree of acceptance and application depth of digital technology also vary among industries. Therefore, industrial structure was selected as a supplementary control variable and added to the model; it is measured according to the ratio of the added value of the secondary industry to that of the tertiary industry. Financial level is one of the important indicators for measuring the stage of economic development and resource allocation capability of a region or country. The abundance of financial resources directly affects the investment scale, technological innovation capabilities, and policy implementation effects of water resources management. Therefore, the financial level, represented by the ratio of the added value of the financial industry to the GDP, was used as a supplementary control variable.

Bilateral trimming at the 1% level was carried out. To prevent potential extreme values from affecting the estimation results, we performed bilateral trimming of the data at the 1% level and re-analyzed the processed data. The re-estimated results are presented in column (4) of Table 5.

From Table 5, it can be concluded that the p-values corresponding to the Kleibergen–Paap rk LM statistic are all 0.000, and the Kleibergen–Paap rk Wald F statistic is greater than the critical value at the 10% level of the Stock–Yogo weak identification test, indicating that the selected instrumental variable is reasonable. In column (1), the coefficient estimate for digital technology is 0.636, which is significant at the 1% statistical significance level, confirming the robustness and reliability of the benchmark regression results. Additionally, after removing samples from municipalities directly under the control of the central government, adding control variables, and conducting bilateral trimming at the 1% level, the coefficients of the impact of digital technology on water resources management remain significantly positive, indicating that digital technology still has a positive effect on water resources management, fully demonstrating the robustness of the conclusion of Hypothesis 1.

3.5. Mechanism Effect Test

To reveal the transmission mechanism of digital technology with respect to water resources management, green innovation was introduced into Equations (12) and (13) for testing. The results are presented in

Table 6. Results of the mediation effect test.

Variable	(1) GI	(2) WRM
DT	0.181 ** (2.48)	0.171 *** (2.75)
GI		0.244 *** (4.62)
City	−0.520 *** (−3.58)	−0.664 *** (−4.64)
Eco	0.024 (0.60)	0.018 (0.47)
Pop	1.261 (0.71)	1.511 (0.89)
Inf	0.022 (1.09)	0.014 (0.72)
_cons	0.275 (0.69)	−0.066 (−0.17)
Control	Yes	Yes
Province	Yes	Yes
Year	Yes	Yes
R2	0.864	0.874
N	300	300
Sobel test		0.044 **
Bootstrap test	[0.010, 0.110]

Note: t-Statistics are in parentheses, *** p < 0.01, ** p < 0.05.

.

In Table 6, column (1) shows that digital technology has a significant positive impact on green innovation. Column (2) presents the regression results with green innovation as the mediating variable. The estimated coefficient of green innovation is 0.244, passing the 1% significance test, indicating that green innovation can significantly promote the development of water resources management. At this point, the estimated coefficient of digital technology remains significantly positive, suggesting that digital technology can indirectly affect water resources management through green innovation. Furthermore, the p-value of the Sobel test is less than 0.05, and the 95% confidence interval of the Bootstrap test does not include 0, both of which further verify the existence of the green innovation transmission mechanism. Therefore, the results indicate that digital technology promotes the development of water resources management through green innovation, and hypothesis 2 is validated.

3.6. Heterogeneity Analysis

In fact, due to differences in resource endowment and stages of development, there is significant heterogeneity in the regional distribution of water resources management levels. Therefore, we used the average level of water resources management as a benchmark to divide the 30 provinces into regions with higher and lower levels of water resources management. The specific regression results are presented in

Table 7. Results of the heterogeneity analysis of water resources management levels.

Variable	Higher Lever (4) WRM	Lower Lever (5) WRM
DT	0.270 ** (2.03)	0.125 ** (2.20)
City	−0.017 (−0.27)	−0.0224 (−1.21)
Eco	0.071 * (1.86)	−0.025 (−0.75)
Pop	0.010 ** (2.19)	0.009 (0.77)
Inf	−0.344 (−0.55)	−0.157 (−0.44)
_cons	1.603 (1.54)	2.244 *** (4.09)
Province	Yes	Yes
Year	Yes	Yes
N	80	210
R2	0.417	0.388

Note: t-Statistics are in parentheses, *** p < 0.01, ** p < 0.05, * p < 0.1.

.

Table 7 reveals that digital technology has a positive impact on regions with higher and lower levels of water resources management. It is noteworthy that the positive impact of digital technology on water resources management is greater in the regions with higher levels of water resources management. This may be due to the fact that these provinces are concentrated in the southeast and southwest regions, where water resources are more abundant than the central and northern regions. Additionally, these regions have maturer and more comprehensive water resources management systems. Furthermore, digital technology has developed rapidly, providing strong support for the digital transformation of water resources management. Consequently, the dividends of digital technology are more fully realized in these regions.

3.7. Threshold Effect Test

Considering the possible nonlinear relationship between digital technology and water resources management, we constructed a dynamic panel model with green innovation as the threshold variable. The “bootstrap” method was adopted [44]; as shown in

Table 8. Results of the threshold effect test.

Variable	Threshold	P	Crit10	Crit5	Crit1	Threshold Value
GI	Single	0.024	14.947	16.951	23.056	1.840
	Double	0.672	12.956	15.886	21.310	1.972

, after conducting repeated sampling 500 times, the results indicate that the green innovation threshold variable significantly passed the single-threshold test but failed to pass the double-threshold and triple-threshold tests. Based on this, regression models with corresponding threshold numbers were established, yielding the regression results in

Table 9. Results of the threshold regression.

Variable	Coefficient	std	P	[95% conf. Interval]
City	−0.606 ***	0.182	0.002	−0.979	0.234
Eco	0.052 *	0.030	0.086	−0.007	0.113
Pop	1.829	2.420	0.456	−3.119	6.778
Inf	0.016	0.022	0.470	−0.028	0.060
GI < 1.840	−0.025	0.121	0.839	−0.272	0.223
GI ≥ 1.840	0.282 ***	0.080	0.001	0.119	0.445
Control	Yes
N	300
R2	0.242

Note: t-Statistics are in parentheses, *** p < 0.01, * p < 0.1.

.

In Table 8 and Table 9, it can be observed from the change in the coefficient of digital technology that when the green innovation degree is lower than the threshold value of 1.840, the impact of digital technology on water resources management is not significant. However, when digital technology exceeds the 1.840 threshold, digital technology will positively contribute to the development of water resources management. It can be inferred that green innovation has a threshold effect on the influence of digital technology on water resources management, implying that as the level of green innovation continues to rise, the promotional effect of digital technology on water resources management will gradually intensify. Thus, Hypothesis 3 is validated.

3.8. Further Analysis

The test results in the above section indicate that there is a close relationship between digital technology and water resources management. Further exploring the coupling and coordination relationship between the two can accurately reveal the deep-seated issues of water resources management and promote digital technology to empower water resources management. Drawing on the equal interval division method proposed by Zhao et al. [45], the coupling coordination degree of digital technology and water resources management was divided into 10 types, as shown in

Table 10. Types of coupling coordination degrees.

Interval	Type
0–0.1	Extremely imbalanced
0.1–0.2	Hyper imbalanced
0.2–0.3	Moderately imbalanced
0.3–0.4	Mildly imbalanced
0.4–0.5	On the verge of imbalanced
0.5–0.6	Barely adequate coordination
0.6–0.7	Primary coordination
0.7–0.8	Intermediate coordination
0.8–0.9	Good coordination
0.9–1.0	High-quality coordination

. Based on Equations (16)–(18), the coupling coordination degree of digital technology and water resources management in China from 2013 to 2020 was calculated. In addition, referring to the method of regional division proposed by Wang and Wang [46], China was divided into Eastern China, Central China, and Western China, as presented in the results in

Table 11. Coupling coordination degrees of digital technology and water resources management in China from 2013 to 2022.

Region	Province	2013	2014	2015	2016	2017	2018	2019	2020	2021	2022
Eastern China	Beijing	0.770	0.771	0.780	0.797	0.789	0.790	0.785	0.801	0.796	0.751
	Fujian	0.507	0.462	0.365	0.361	0.391	0.405	0.406	0.428	0.433	0.421
	Guangdong	0.443	0.450	0.442	0.436	0.446	0.463	0.488	0.474	0.481	0.499
	Hainan	0.510	0.494	0.509	0.509	0.517	0.606	0.603	0.453	0.433	0.465
	Hebei	0.582	0.542	0.535	0.553	0.511	0.496	0.495	0.466	0.452	0.478
	Jiangsu	0.783	0.778	0.776	0.751	0.751	0.739	0.739	0.697	0.689	0.707
	Liaoning	0.424	0.431	0.423	0.444	0.464	0.471	0.435	0.474	0.476	0.463
	Shandong	0.578	0.568	0.574	0.549	0.543	0.550	0.537	0.542	0.525	0.512
	Shanghai	0.684	0.711	0.720	0.699	0.696	0.693	0.700	0.740	0.762	0.741
	Tianjin	0.479	0.434	0.412	0.408	0.413	0.420	0.432	0.459	0.458	0.444
	Zhejiang	0.577	0.576	0.569	0.564	0.563	0.567	0.470	0.570	0.528	0.528
	Average	0.576	0.565	0.555	0.552	0.553	0.564	0.554	0.555	0.548	0.546
Central China	Anhui	0.530	0.525	0.518	0.516	0.517	0.520	0.536	0.534	0.475	0.492
	Heilongjiang	0.502	0.474	0.595	0.559	0.437	0.416	0.434	0.409	0.424	0.438
	Henan	0.610	0.580	0.569	0.535	0.527	0.528	0.546	0.495	0.495	0.518
	Hubei	0.599	0.575	0.537	0.527	0.531	0.517	0.531	0.495	0.495	0.525
	Hunan	0.545	0.523	0.522	0.472	0.460	0.454	0.441	0.458	0.455	0.465
	Jiangxi	0.599	0.579	0.575	0.559	0.539	0.506	0.496	0.473	0.466	0.491
	Jilin	0.508	0.483	0.475	0.483	0.490	0.444	0.407	0.434	0.437	0.429
	Shanxi	0.579	0.572	0.562	0.553	0.553	0.569	0.570	0.498	0.486	0.504
	Average	0.559	0.539	0.544	0.526	0.507	0.494	0.495	0.475	0.467	0.483
Western China	Chongqing	0.660	0.644	0.641	0.625	0.626	0.620	0.612	0.553	0.570	0.577
	Gansu	0.753	0.731	0.723	0.732	0.747	0.753	0.743	0.686	0.675	0.706
	Guangxi	0.467	0.460	0.465	0.476	0.493	0.509	0.423	0.448	0.411	0.443
	Guizhou	0.513	0.500	0.508	0.508	0.529	0.549	0.543	0.539	0.553	0.470
	Inner Mongolia	0.522	0.490	0.475	0.488	0.501	0.517	0.463	0.456	0.468	0.477
	Ningxia	0.493	0.406	0.396	0.390	0.425	0.446	0.378	0.448	0.443	0.419
	Qinghai	0.652	0.616	0.618	0.638	0.640	0.601	0.564	0.534	0.528	0.565
	Shaanxi	0.500	0.476	0.470	0.476	0.486	0.485	0.395	0.422	0.418	0.543
	Sichuan	0.580	0.573	0.575	0.565	0.570	0.599	0.609	0.678	0.583	0.540
	Xinjiang	0.499	0.506	0.495	0.505	0.527	0.455	0.425	0.459	0.435	0.435
	Yunnan	0.750	0.747	0.746	0.737	0.744	0.738	0.734	0.716	0.718	0.686
	Average	0.581	0.559	0.556	0.558	0.572	0.570	0.535	0.540	0.527	0.533
National average		0.573	0.556	0.552	0.547	0.548	0.548	0.531	0.528	0.519	0.524

. Furthermore, the trend of changes in the average coupling coordination degree of digital technology and water resources management in Eastern China, Central China, Western China, and nationwide is illustrated in Figure 4.

Table 11 indicates that, nationally, the mean value of the coupling coordination degree of digital technology and water resources management was between 0.5 and 0.6, which corresponds to a state of barely adequate coordination. Furthermore, the coupling coordination degree exhibited a decreasing trend, with the mean value decreasing from 0.573 in 2013 to 0.524 in 2022, which indicates that although digital technology plays a certain role in water resources management, the overall coordination degree is not high and is still in the initial stages of coordination.

At the region level, it can be seen from Figure 4 that there are significant differences in the coupling coordination degree of digital technology and water resources management in various regions. Regarding regional disparities, for example, in 2022, the national average of the coupling coordination degree of digital technology and water resources management was 0.524, with an average of 0.546 in Eastern China, 0.483 in Central China, and 0.533 in Western China. The coupling coordination degree of digital technology and water resources management in Eastern China and Western China was higher than the national average, while in Central China, it was significantly lower than the national average. This may have been caused by the imbalance in the level of economic development, scientific and technological innovation capacity, and water resources endowment between regions.

At the provincial level, disparities are also evident. In Eastern China, Beijing boasted the highest overall coupling coordination degree of digital technology and water resources management, while the coupling coordination degree of Fujian was 0.3–0.4 in most years, corresponding to a state of mild imbalance. In Western China, the coupling coordination degree of digital technology and water resources management in Gansu and Yunnan was greater than 0.7 in most years, generally reaching a state of intermediate coordination, and the degree of integration of the provinces was more desirable. In contrast, the provinces of Central China with lower coupling coordination degrees, such as Jilin and Heilongjiang, were on the verge of imbalance, lagging behind other provinces in the central region. Therefore, there is also a significant imbalance between provinces in different regions.

4. Discussion

Traditional methods of water resources management often rely on human experience and static data analysis. So, water resources management is difficult to adapt to the rapidly changing market demand and complex external environment. The widespread application of digital technology has not only continuously transformed social production and lifestyles but also injected new impetus into enhancing the level of water resources management [47,48]. Against this backdrop, this paper proposes three hypotheses. Evaluation index systems of digital technology and water resources management were constructed based on panel data from 30 provinces in China spanning from 2013 to 2022. This paper systematically reveals the direct impact of digital technology on water resources management, providing robust empirical support for the widespread application of digital technology in the field of water resources management. Importantly, this paper introduces green innovation as a mediator and a threshold variable to further explore the influence of digital technology on water resources management. Additionally, this paper further analyzes the coupling coordination degree between digital technology and water resources management. We believe that during the process of digital transformation, not only economic effects but also ecological effects should be investigated. Therefore, in the process of water resources management, firstly, it is necessary to promote the application and upgrading of digital technology to build a smart water resources management system. Relevant departments should strengthen the penetration of digital twin and other technologies in the whole chain of water resources management, which will widen the scope of water resources management decisions from mainly focusing on the identification of water status to the integrated simulation and control of water quantity, water quality, and ecology. Secondly, a deep integration of green innovation, digital technology, and water resources management should be fostered to create a collaboratively evolving innovation ecosystem. By establishing interdisciplinary research platforms, we can jointly explore the potential applications of digital technology and green innovation in the dynamic assessment of water supply and demand, water pollution prevention, and ecological restoration, with the collaboration of multiple parties. Finally, different strategies should be used for the application of digital technology. When formulating strategies for the promotion of digital technologies, it is crucial to fully consider regional differences and tailor policies and measures to local conditions. Additionally, strengthening regional cooperation is also an effective way to enhance the level of water resources management nationwide. The task of promoting the digital transformation of water resources management is arduous and requires our long-term and unremitting efforts.

5. Conclusions

Our research findings are as follows: Firstly, in terms of the direct impact, digital technology has a significant promotional effect on water resources management, with an estimated coefficient of 0.221 and a significance level of 1%. The endogeneity and robustness tests also confirmed the reliability of this result. Secondly, from the perspective of transmission mechanisms, digital technology can improve water resources management by enhancing green innovation. Thirdly, in terms of heterogeneity analysis, the enabling role of digital technology in water resources management exhibits notable regional heterogeneity. The estimated coefficient was 0.125 in regions with lower levels of water resources management, while it was 0.270 in regions with higher levels, indicating a stronger positive effect of digital technology on water resources management in these regions. Fourthly, based on the threshold effect analysis, the impact of digital technology on water resources management is influenced by the level of green innovation. When the level of green innovation reaches 1.840, digital technology will facilitate water resources management. Fifthly, from the perspective of the coupling coordination degree of digital technology and water resources management, the overall performance is not ideal. China was generally in the stage of barely adequate coordination from 2013 to 2022. Additionally, there are significant differences in the coupling coordination degrees between regions, mainly showing a trend of Eastern China > Western China > Central China, indicating that there is still considerable room for development.

Although this study has achieved some results in terms of the impact of digital technology on water resources management, there are still limitations. Future research can incorporate policy environment and institutional factors into the analytical framework. What is more, it would be more ideal if future research could expand its scope to the prefecture-level-city level or industrial level.