Ecological restoration process of watershed land space with intelligent IoT technology

The Internet of Things (IoT) spans many areas of our lives and is continuously changing our world (Zhang et al. 2021). The application of IoT technology in watershed ecological protection and restoration can realize the sharing of watershed resources and the collaboration of watershed activities, and provide a structured and complete intelligent database for wise control, accurate warning, scientific utilization, and protection of watershed resources (Bakshi et al. 2015). The rapid advancement and healthy development of watershed ecological protection and restoration are receiving increasing attention (Cantrell et al. 2017). The water ecological IoT is a crucial technological application for monitoring water ecosystems. It involves collecting ecological environmental information around a body of water, transmitting it through the IoT, and using big data cleaning, storage, and analysis to assess the health of the water environment (Cao et al. 2018). This technology provides a comprehensive network for understanding the rules, mechanisms, and ecological security of watershed ecosystems under natural and anthropogenic disturbances. It is also a valuable tool for supporting policymaking and scientific research on watershed ecology.

Water plays an important role in the regional socioeconomic development. China has a relative lack of water resources, especially in the Huaihe River Basin. The Huaihe River Basin covers an area of approximately 10,000,000, spanning five provinces and cities in Hubei, Henan, Anhui, Jiangsu, and Shandong, and includes four secondary water resources: the upper reaches of the Huai River above Wangjiaqian, the middle reaches of the Huai River from Wangjiaba to Hongze Lake, the lower reaches of the Huai River below Hongze Lake, and the Zishusi River (Chang et al. 2015). Because the Huaihe River Basin only accounts for the total annual average water resources of the country, and the population density in the basin is the highest among the large basins, the pressure on water resources is twice the national average (Chen et al. 2020). Water stress in the basin is aggravated by the uneven spatial and temporal distribution of water resources, the high level of water resource development and utilization, and the serious pollution of water bodies.

The Huaihe River Basin is located in the north-south climate zone in China, and there is a lack of natural baseflow in the north of the mainstream; there are many tributaries in the basin, and there are several lakes and hanging places that are typical river basin water ecosystems with the coexistence of rivers and lakes; water pollution in the basin is serious, and there are many water facilities on the rivers, such as water trips, which have a great impact on the water ecological environment (Chen 2019). Most of the water ecosystems in the Huaihe River Basin are over-exploited without any control, and the health of the water ecosystem has different degrees of problems. During the ‘11th five-year plan’ period, the national and governmental efforts at all levels and the technical support of ‘water special projects’ have improved the water quality of the Huaihe River Basin to a certain extent, and some of the ditches and rivers have the conditions for water ecological restoration, but the water ecological environment of the basin has not yet been fundamentally solved. The shortage of water resources and environmental pollution in the basin have led to poor self-purification capacity of water bodies and reduced water ecological functions and aquatic biodiversity, which to a certain extent restrict the economic development of the Huaihe River Basin (Corbett & Mellouli 2017). Therefore, implementing degraded water ecosystem restoration, improving water quality, providing water resource utilization, and enhancing water ecosystem service functions are the only ways to solve regional water environment problems.

The ecological restoration of watershed land space has emerged as a critical concern in the face of climate change and increase in natural disasters. Conventional methods, such as mechanical dredging, manual removal of invasive plants, and chemical treatments, have proven ineffective or insufficient, and in recent years, there has been a growing interest in incorporating intelligent IoT technology into the ecological restoration process of watershed land space. The use of IoT technology has the potential to mitigate the negative impact of conventional methods on ecosystems, thus contributing to more sustainable and effective restoration practices. Despite the potential benefits of integrating IoT technology into ecological restoration, there is currently a research gap regarding the performance and drawbacks of these strategies. This research gap highlights the need for further investigation of the effectiveness, feasibility, and potential negative impacts of technological interventions on ecological restoration. In this study, we aim to explore the use of intelligent IoT technology in the ecological restoration process of watershed land spaces. By evaluating the existing methods and highlighting their performance and drawbacks, we aim to bridge the research gap and contribute to a better understanding of the potential of the IoT in sensitive ecological environments.

This study delves into the key aspects of ecological restoration in the Huaihe River Basin using algorithmically optimized frequencies and remote sensing images from Gaofen-1 and Gaofen-6 satellites. This study aims to define the scope of pollution control and ecological restoration in the Huaihe River Basin, identify rivers that can be ecologically restored, and propose a river ecological restoration threshold concept and its connotation. In addition, it establishes a river ecological restoration threshold index system, investigates river ecosystems, and classifies river ecological restoration types. All of these efforts lay a solid foundation for determining the ecological restoration threshold and conducting ecological restoration work in the future.

Since the concept of river ecosystem health was introduced in the late 1970s, research on river ecological restoration has continued and intensified. However, so far, there is no unified concept, and the main point of disagreement is whether the benchmark of health is only relative to natural rivers or needs to be determined in conjunction with social functions. The study (Guo et al. 2016) argues that healthy rivers should be able to maintain major ecological processes and have biological communities that can restore themselves to the pre-disturbance state, emphasizing that river health is a natural ecological process of rivers; the study (Han et al. 2019) argues that river ecosystem health is relative to human values and depends on social judgment, and needs to meet both natural and social functions of rivers so that health is more permanent and sustainable.

Literature (Ji et al. 2021) proposed various standard systems for the planning, objectives, evaluation, and other contents of ecological restoration to guide the standardization and the whole process of land spatial ecological restoration. The evaluation criteria of literature (Ji et al. 2021) include early natural resource asset evaluation, physical accounting of natural resources, evaluation of ecosystem service functions, and the identification of ecologically important regions. According to the principle that mountains, water, forests, fields, lakes, and grasses are a community of life, comprehensive measures should be taken to deal with the degradation of the original ecosystem and the imbalance of the overall structure and function caused by long-term and intensive development and construction, unreasonable utilization and natural disasters, as well as the process of ecological restoration, ecological improvement, ecological reconstruction, and ecological restoration of the terrestrial space ecosystem (Zhou et al. 2022, 2023). For example, the ecological restoration of coal mine subsidence mentioned in Komninos et al. (2019) not only includes the environmental restoration and ecological damage restoration of contaminated land caused by coal mine subsidence, but also includes the social restoration of unemployment and economic transformation caused thereby. The main contents of land space ecological restoration mentioned in the literature (Xiao et al. 2022) include ecological security pattern, ecological infrastructure network, and comprehensive restoration of ecological landscape and elements. The restoration project framework is shown in Table 1.

Under the guarantee of relevant standards system and security system, based on the natural resources, spatio-temporal big data platform, natural resources backbone network, Internet and IoT, ecological restoration project management platform and auxiliary decision-making analysis platform as the pillars form an information application system with the ministry, province, city, and county vertically connected and business function modules horizontally connected (Tyagi et al. 2022; Li et al. 2017). It is mainly divided into six layers: first, the perception layer, which refers to access to information (Norton et al. 2016). It adopts aerospace and aviation satellite remote sensing, cell phones, cameras, and other ground sensing devices, and network-directed crawling of relevant information to intelligently sense the status and changes of each element of the ecosystem (Yang et al. 2022). The second is the network layer, which refers to the network environment on which the platform operates. It mainly relies on the natural resource backbone network, and part of the data acquisition and sharing relies on the Internet and the IoT (Pan et al. 2018). The third is the data layer, which refers to the main form of data stored in the platform database. Among them, natural resource survey data and remote sensing data are derived from land and resource surveys, annual change survey, and all-weather remote sensing monitoring work; ecological restoration project data and field verification data are derived from the data produced and extracted during the construction and supervision of various types of ecological restoration projects. Web crawl data are natural, social, economic, and other event data and statistics related to ecological restoration collected in a regular targeted manner. The fourth is the platform layer, which refers to the specific software platform for operation. According to the user and business differences, it is divided into two platforms: ecological restoration project management and auxiliary decision analysis. Fifth is the application layer, which refers to the functional modules contained in each software platform, and its function settings should be closely integrated with business needs, mainly including ecological restoration planning and ecological restoration compensation analysis, and the whole life cycle management of various ecological restoration projects (Reif & Theel 2017). Sixth is the user layer, which refers to the service targets of the software platform and, mainly, the natural resource authorities at all levels.

To sum up, under the background of intelligent sensor technology, the land space ecological restoration and development strategies in the Jianghuai watershed area have a profound research foundation.

Then, the gray level co-occurrence matrix to make statistics on the correlation of spatial texture features is used. The statistical information table is shown in Table 2.

Table 2

Statistical table of gray level co-occurrence matrix in the Jianghuai watershed

Serial no. .	Texture features .	Calculation formula .	Characteristics and significance .
1	Mean value		It reflects the average value of pixel texture in the window, and the average value is proportional to the texture regularity.
2	Variance		It indicates the degree of gray dispersion in the window. The larger the variance, the rougher the image texture.
3	Synergy	–	It reflects the homogeneity of image pixel values in the window, mainly reflecting the uniformity of image texture in some areas.
4	Contrast ratio		It indicates the depth of the pixel groove in the window, and the texture feature also reflects the clarity of the image.
5	Dissimilarity		It is the same as contrast, but the difference is that the index weight increases linearly, showing the degree of local changes in the image.
6	Information entropy		The measurement of image information reflects the complexity of image texture in the window.
7	Second moment		Reflect the uniformity of pixel value distribution and texture thickness in the window.
8	Relevance		Reflect the similarity between row and column elements in the window and reflect the gray linear relationship in the image.

Serial no. .	Texture features .	Calculation formula .	Characteristics and significance .
1	Mean value		It reflects the average value of pixel texture in the window, and the average value is proportional to the texture regularity.
2	Variance		It indicates the degree of gray dispersion in the window. The larger the variance, the rougher the image texture.
3	Synergy	–	It reflects the homogeneity of image pixel values in the window, mainly reflecting the uniformity of image texture in some areas.
4	Contrast ratio		It indicates the depth of the pixel groove in the window, and the texture feature also reflects the clarity of the image.
5	Dissimilarity		It is the same as contrast, but the difference is that the index weight increases linearly, showing the degree of local changes in the image.
6	Information entropy		The measurement of image information reflects the complexity of image texture in the window.
7	Second moment		Reflect the uniformity of pixel value distribution and texture thickness in the window.
8	Relevance		Reflect the similarity between row and column elements in the window and reflect the gray linear relationship in the image.

View Large

To a large extent, this paper extracts the texture features of five windows for the high score one image band and the high score six new band information. The step size extraction window is 3 × 3, 5 × 5, 7 × 7, 9 × 9, 11 × 11, 13 × 13. The direction is 0°, 90°, 45°, and 135°. Common texture features are extracted. After analysis and experiment, 135° is determined as the direction, and the step size is 1.

In the river ecosystem health assessment, most scholars prefer to use the integrated index method, but some scholars also try to use mathematical statistical methods such as the neural network method (Barrera Sánchez et al. 2022) and the fuzzy evaluation method (Wu et al. 2022). Since the composite index method involves assigning weights, it is more subjective. Therefore, this study proposes to use the T − S fuzzy neural network to calculate each criterion layer as well as the integrated index to determine the health status of rivers, and the specific algorithmic process.

• (1)

  First, 100 sets of data are interpolated between [1, 5] using equal interval uniform distribution as training samples.

• (2)

  The network structure (including the settings of input nodes, implicit nodes, and output nodes) is determined according to the training data dimension and also initialized the parameters and coefficients of the affiliation function of the fuzzy neural network and performed normalization to construct a suitable fuzzy neural network.

• (3)

  Then, the constructed fuzzy neural network is trained using training samples.

• (4)

  Finally, the indexes to be evaluated are used as prediction samples, and the trained fuzzy neural network is used to predict the evaluation indexes and judge the health status of each section according to the evaluation criteria.

In this study, FRAGSTATS software was used to calculate the patch scale, patch-type scale, and overall landscape scale land and resource restoration indicators by analyzing the meaning of indicators and the relationship between indicators, the current characteristics of land and resource restoration, and the existing ecological landscape problems in the Jianghuai watershed area.

The vector data of land use were resampled into 30 m * 30 m grid data, and FRAGSTATS software was used to calculate the relevant indicators on the overall landscape scale. On the overall landscape scale, 12 indicators were selected, including the overall landscape area (total alkalinity (TA)) and the proportion of the largest patch in the landscape area (largest patch index (LPI)). The results of the indicators are shown in Table 3.

From the calculation results of landscape scale indicators, the proportion of the largest patch in the landscape area (LPI) has changed significantly, indicating that the dominance of dominant landscape types as well as the dominant species in the landscape gradually decrease and that the intensity and frequency of human activities' interference with the landscape continue to increase. In patch shape change, edge density (ED) and area-weighted average share index (SHAPE_AM) decreased slightly, reflecting the trend of landscape pattern toward anti-fragmentation, and the fragmentation degree of habitat improved slightly. In terms of the proximity index, the European average nearest-neighbor distance (ENN_MN) increased slightly, indicating that the distance between patches of the same type became larger, and the landscape distribution showed a more decentralized trend. The index model was reasonable and scientific. In addition, in terms of area edge index, density difference index, shape complexity index, proximity index, and aggregation and dispersion index, the results of each index on the landscape-type scale of the Jianghuai watershed in 2021 are calculated in detail, as shown in Table 4.

Table 4 shows that in terms of area edge indicators, the highest patch-type area (CA) was farmland, followed by water area and construction land, and the smallest patch-type area was grassland and other land. The highest proportion of the largest patch in the LPI is the water area, while other land and grassland are the lowest. This shows that compared with the water area, the dominance of the dominant landscape types in the landscape and the dominant species in the landscape are lower, the intensity and frequency of human activities are greater, and the smallest patch type is grassland and other land, which further confirms the interpretation of patch-type area indicators. The higher patch density (PD) is for construction land and water area, and the lower one is for forest land, grassland, and other land, while the number of patches for forest land, grassland, and other land is more, and the patches are more fragmented.

The landscape connectivity evaluation index based on graph theory quantifies the performance of the ecological network before and after optimization. According to the analysis principle of graph theory, the ecological source and ecological corridor in the study area are abstracted as the graph of nodes and connections. When the connection distance between nodes is less than the optimal distance threshold of 400 m, the link between nodes is considered to exist; otherwise, it is considered to not exist. A landscape component is a patch group that exists as a link between two ecological source patches. There is a link between patches of the same landscape component, and there is a fracture between patches of different landscape components. The landscape connectivity index represents the smoothness of the links between ecological sources in the landscape. The relevant indexes representing connectivity in graph theory are selected, which are binary connectivity index H, IIC, probabilistic connectivity index PC, and three specific connectivity indexes.

The calculation results of the landscape connectivity index of the ecological network before and after optimization are shown in Table 6. According to the performance evaluation results of the ecological network before and after optimization, the connectivity of the ecological network after optimization has been greatly improved, in which the Harry index (H) has increased by 10.92 times, the overall connectivity index (IIC) has increased by 43.45%, and the possible connectivity index (plastocyanin (PC)) has increased by 99.58%. This indicates that the ecological network connection after optimization is more stable than the network before optimization.

Then, the linear operation module is used to calculate the land use demand of the 2035 Jianghuai watershed area by using the land use data of 2000, 2005, 2010, 2015, and 2018, as shown in Table 7.

Table 7

Land use demand of 2035 Jianghuai watershed

Land category name .	Cultivated land .	Woodland .	Grassland .	Waters .	Land used for building .	Unused land .
	16,074	555.27	561.28	1,844.2	3,600.9	7.49

Land category name .	Cultivated land .	Woodland .	Grassland .	Waters .	Land used for building .	Unused land .
	16,074	555.27	561.28	1,844.2	3,600.9	7.49

View Large

Figure 12 shows that the maximum intermediary degree in the network is 1.430, and the corresponding ecological node number is 61. The minimum value is 0.12, and the corresponding ecological node number is 79. The intermediary degree of most ecological nodes in the network is between 0.12 and 0.5, accounting for 81.01% of all nodes. There are 13 nodes whose mediality value is greater than 0.5, which indicates that under such a landscape network structure, most nodes in the Jianghuai watershed have a low mediality value and strong network heterogeneity (Xie et al. 2020).

The middle, north, and east of the study area have small intermediate degree values mainly because economic development, construction activities, and agricultural production have affected the movement of landscape ecological flow to a certain extent. The structural importance value of most nodes is greater than the ecological function importance value. There are 61 ecological nodes whose importance of ecological function is greater than that of the structure. There are 18 ecological nodes with a value less than 1. A value less than 1 indicates that the ecological background of these ecological nodes is poor. During ecological restoration, attention should be paid to these nodes and their control range, as shown in Table 8.

Table 8

Prediction of the area of ecological blind areas in the Jianghuai watershed

Ecological zoning type .	Ecological blind area .
	First-level ecological blind area .	Secondary ecological blind area .
Partition I	5.57	18.25
Zone II	34.02	36.71
Zone IV	13.59	0
Division VI	4.33	0
Division VII	0	10.09
Total	57.51	65.05

Ecological zoning type .	Ecological blind area .
	First-level ecological blind area .	Secondary ecological blind area .
Partition I	5.57	18.25
Zone II	34.02	36.71
Zone IV	13.59	0
Division VI	4.33	0
Division VII	0	10.09
Total	57.51	65.05

View Large

The use of intelligent IoT technologies is rapidly transforming the ecological restoration process of watershed land spaces. Intelligent IoT technologies provide numerous insights, predictions, and control over restoration efforts by integrating sensor devices, big data analytics, and artificial intelligence. Intelligent IoT technologies can detect and monitor environmental factors, such as temperature, humidity, soil moisture, and pH levels, which are crucial for ensuring successful restoration. These technologies also allow for real-time data acquisition and analysis, enabling quick decision-making and prompt action. Moreover, intelligent IoT technologies can predict and mitigate potential problems such as soil erosion, water loss, and invasive species. They also promote sustainability by reducing input resources, such as water and energy. Intelligent IoT technologies have proven to be effective tools for optimizing ecological restoration efforts, while improving process efficiency and cost-effectiveness.

The ecological restoration process of watershed land spaces with intelligent IoT technology aims to explore how the integration of IoT technology can aid in the ecological restoration of watershed land spaces. The motivation behind this study is to show how the use of IoT can provide a more efficient and effective approach to ecological restoration processes, reduce the time, cost, and resources required, as well as outline the proposed approach and provide insight into the benefits that can be achieved when using IoT in the ecological restoration of land spaces. The focus is on the use of sensors, data analytics, and machine learning to monitor and manage ecosystems, reduce pollution, and restore natural habitats. It is important to note that ecological restoration is a critical element in the conservation of natural resources and habitats, and the use of IoT technology provides an exciting opportunity to advance restoration efforts. The proposed article contributes to the literature and provides valuable information to scientists, researchers, and other stakeholders in the field of ecological restoration.

Ecological restoration is critical for promoting biodiversity conservation and ensuring the sustainability of the planet's natural resources. Combining this process with the power of IoT technology can expedite and enhance the restoration process in various ways. First, IoT sensors can collect and analyze data on the ecological status of the watershed land space, helping to identify areas that require restoration intervention. These data can then be used to design personalized restoration plans that address the specific needs of each location, leading to more efficient allocation of resources. Second, IoT can be leveraged to monitor the progress of restoration efforts in real time. This will help project managers to identify any issues that may arise, allowing prompt corrective actions. It can also help to measure the impact of restoration efforts over time, providing valuable insights for future planning and decision-making. Third, the use of IoT-powered drones can help to plant a large number of seeds in a shorter time, leading to faster reforestation. These drones can also monitor seed growth and vegetation health, reducing the need for manual inspection. Finally, IoT devices can help to automate critical post-restoration maintenance tasks, such as weed control, thereby ensuring the site's long-term sustainability. Combining IoT technology with ecological restoration can promote environmental sustainability and biodiversity conservation, resulting in a better future.

This article presents a novel approach to ecological restoration that integrates intelligent IoT technology. This study aims to provide insights into the strengths and weaknesses of the proposed approach, as well as how others in the field can benefit from its adoption. One of the main strengths of the proposed approach is its ability to enhance the effectiveness and efficiency of the ecological restoration processes. With the integration of intelligent IoT technology, it is possible to monitor and analyze the ecological parameters of watershed land space in real time, thus enabling timely intervention. Additionally, the proposed approach seeks to leverage the power of the IoT to promote sustainability, natural resource conservation, and environmental protection. However, one limitation of the proposed approach is the high cost of setting up the required infrastructure and technology. In addition, the integration of large amounts of data generated by IoT devices may pose challenges in data management and analysis, leading to the need for more specialized skills and knowledge. The proposed approach presents exciting opportunities for the ecological restoration of watershed land space, and other professionals in the field can benefit from its adoption by leveraging the latest technological advancements, while promoting environmental conservation. Future studies should focus on implementing the proposed approach in diverse ecological settings to evaluate its effectiveness, sustainability, and scalability.

This study was an attempt to evaluate the health of river ecosystems and explore the construction of a river ecological restoration model, mainly based on the research object of river ecosystems in the Huaihe River Basin. The experimental data of the algorithm model proposed in this study show that the land type of the Jianghuai River Basin is mainly woodland and watershed with a relatively uneven spatial distribution. A total of 85 ecological corridors that showed a pattern of more in the middle and less in the periphery were identified. A total of 267 radiation channels with an irregular tree distribution were identified. In addition, 41 ecological nodes were identified in the model, which were mainly distributed in ecological corridors, and 50 ecological breakpoints were identified in the model that concentrated in the traffic arteries of the Yangtze and Huaihe River Basins. The empirical results showed that the fitting accuracy of the model was 89%. The ecological restoration model system proposed in this study has a certain guiding significance for the ecological restoration of related rivers.

The authors would like to show sincere thanks to those techniques who have contributed to this research.

This study is supported by 1. The Key Research Project of Humanities and Social Sciences of Colleges and Universities in Anhui Province, Research on the ecological restoration mode and security mechanism of land space under the concept of ecological civilization – taking the Jianghuai watershed as an example (SK2021A0687); 2. Research Program of Institutions in Colleges and Universities of Anhui Province (Philosophy and Social Sciences), Research on the mechanism and path of land consolidation to help new rural transformation from the perspective of rural revitalization-taking the Jianghuai watershed as an example (2022AH051063); 3. Anhui Province Philosophy and Social Science Planning Project, Research on the Paradigm and Guarantee Mechanism of Ecological Restoration of Anhui Territorial Space from the Perspective of Rural Ecological Revitalization (AHSKY2022D132); 4. Scientific Research Startup Fund Project of Chuzhou University, Study on Land Spatial Ecological Restoration of Jianghuai Watershed (2022qd039); 5. Anhui Provincial Quality Engineering Course Ideological and Political Demonstration Project for Higher Education Institutions, Modern Land Management (2022 kcsz204); 6. Quality Engineering Key Project in Colleges and Universities of Anhui Province，Exploration and Practice of Deep Transformation of Geographical Sciences Serving the National Land Remediation Strategy (2021jyxm1045); 7. The Key Natural Science Research Project of Colleges and Universities in Anhui Province, Study on the optical properties and radiative forcing of atmospheric aerosols in Anhui Province (KJ2021A1078); 8. The Youth Project of the National Natural Science Foundation of China, Study on land use system differentiation characteristics of different rural development types in Jiangsu province (42001193).

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.