Ecological Assessment Based on Remote Sensing Ecological Index: A Case Study of the “Three-Lake” Basin in Yuxi City, Yunnan Province, China

Abstract

With continuous urbanization, human activities have left considerable impacts on the ecology. Therefore, it is necessary to perform timely and objective monitoring and evaluation of the ecology. With the basin of three highland lakes (Fuxian Lake, Xinyun Lake, and Qilu Lake) in Yunnan Province as the study case, four indices, i.e., the Normalized Difference Vegetation Index (NDVI), the Wet Index (WET), the Normalized Differential Build-Up And Bare Soil Index (NDBSI), and the Land Surface Temperature (LST), which indicate, respectively, greenness, humidity, dryness, and heat of the study area, were extracted. On the basis of five sets of terrestrial images of the areas around the three lakes from 2001 to 2021, principal component analysis (PCA) was performed on these four indices; the more informative principal component contribution was selected as the weight to establish a remote sensing ecological index (RSEI) evaluation model to evaluate the ecological environment quality of the study area; the Mann–Kendall test combined with Sen’s slope (Sen + MK) and the Hurst exponent were employed to explore the ecological conditions and development trends of the “three-lake” basin. The results showed that the ecological quality of the study area improved and then deteriorated from 2001 to 2021. The ecological quality classes in the study area were fair, medium, and good. The ecological quality has been greatly improved, but poor ecological quality was still observed in some regions such as Chengjiang. Eighty-eight percent of the study area witnessed a stable trend in the ecological quality over the 20 years; in 2021, the area of built-up land with fair and poor ecological quality reached 140.97 km², which occupies 68.1% of the total area under the same land use type. Analysis shows that urban area expansion and human activities have exacerbated ecological problems of towns and built-up land in the study area. In the selected indicators, both greenness and humidity are positive indicators to ecological quality, and the R² value of the two in 5-year regression was both greater than 0.99, which validated the reliability of the selected model indicators. The research findings are expected to provide a basis for scientific ecological planning and restoration of lake basins.

1. Introduction

Social and technological advances have pushed forward material civilization, but meanwhile triggered ecological and environmental problems [1]. As intensified human activities following rapid social and economic development pose an increasing burden on natural resources and the environment globally [2], ecological improvement draws unprecedented attention and grows into a priority in development initiatives [3]. Ecological assessment is to analyze the potential and constraints of regional natural resources and environment and evaluate the ecological adaptability (or sensitivity) of resources and environment in a hierarchical manner according to regional socio-economic development goals and resource utilization requirements. Ecological assessment has been a hot research topic since as early as the 1960s [4,5,6,7], and the selected indicators have shifted from single indicators to combinations of indicators in relation to topography and geomorphology [8], land cover [9], soil erosion [10], natural landscape [11], human activities [12], and meteorology [13] for the evaluation of specific areas. There were also works that explored ecological quality from such perspectives as hydrology [14], biodiversity and disaster protection [15], urban heat [16], and ecological footprint [17], where the selected evaluation indexes are indicators which are complex and highly representative of the study area.

There have been many previous works on the connotation theory, index system, and methodological models for ecological environment evaluation, and various advanced technical methods have been widely used in ecological environment research. There are many models for ecological quality evaluation: Reza et al. [18] proposed the regional index of ecological integrity (RIEI) model based on landscape fragmentation and other factors; Orfanidis et al. [19] proposed the ecological evaluation index (EEI) model to evaluate the ecology of transitional and near-shore waters; Ma et al. [20] constructed the marine ecosystem carrying capacity evaluation (MECC) model to evaluate the impact of human activities and economic development on marine ecosystems. The methods of ecological evaluation are inexhaustible, and the evaluation indices selected vary from one study to another. These indices used in previous studies can explain a certain aspect of the ecological characteristics of an ecosystem; however, complex ecosystems are affected by multiple factors, and a single ecological index often fails to provide a comprehensive and accurate description of the ecological condition of an ecosystem and, hence, cannot support objective evaluation of ecological changes. In particular, for complex urban ecosystems consisting of different land use types such as urban areas, hills, farmlands, woodlands, and wetlands, a comprehensive index is necessary for comprehensive and objective evaluation of ecological quality [21]. In 2013, Xu et al. [22] proposed a remote sensing ecological index (RSEI) model based on remote sensing images, which is calculated by principal component analysis (PCA). This model solves the problem of subjectivity incurred by manual assignment of weights and produces visualized results which provide a quick, quantitative, and objective solution to evaluation of regional ecological quality. This method has been widely used since it was proposed [23,24,25], and its practicality as well as accuracy has been verified [26,27,28]. Thus, this model, i.e., the RSEI model, is employed in our work to assess the ecological quality in the “three-lake” basin in Yuxi City.

The “three lakes” in Yuxi refer to Fuxian Lake, Xinyun Lake, and Qilu Lake. The rich water resources and favorable exploitation conditions in the basin of these three lakes provide a strong support for Yuxi’s social, economic, and cultural development. Now, as Yuxi enters a stage of rapid urbanization, the interplay of natural, demographic, economic, and policy factors has resulted in many ecological and environmental problems, such as serious rock desertification around the three lakes, slab mudslides, and Yuanjiang landslides. Ecological management of the “three-lake” basin is a complex and systematic initiative, but few previous studies have probed into the ecological quality of the highland lakes in Yunnan Province, especially in Yuxi City. In this study, the ecological quality of the “three-lake” basin in Yuxi is studied to reveal the evolutionary characteristics of ecological quality over the past 20 years. Scientific monitoring and evaluation of ecological quality is crucial for social and economic sustainability in the region [29].

2. Materials and Methods

A detailed workflow was established in this study (Figure 1). First, the Landsat 5 TM images in 2001, 2006, and 2011 and Landsat 8 OLI/TIRS images in 2015 and 2021 of the study area were collected, and the remote sensing ecological index (RSEI) at a 30-m resolution was calculated for each year; second, the spatial and temporal changes in the ecological quality of the “three-lake” basin were analyzed; last, the change trend of ecological quality and trend persistence were analyzed using the Hurst exponent and the Mann–Kendall trend test combined with Sen’s slope (Sen + MK method), and the impact of land use changes on ecological quality was analyzed.

2.1. Study Area

Yuxi City is located in the central part of Yunnan Province (Figure 2) and the western part of the Yunnan-Guizhou Plateau, with a terrain that falls from the northwest to the southeast, a complex topography including mountains, valleys, plateaus, and basins, and a vegetation coverage of 58%. With the “three-lake” basin of Yuxi as the study case, the SWAT model was employed in this study to classify the sub-basins and select representative areas for analysis. Ten districts and counties were involved in the study area, namely Yiliang, Chenggong, Jining, Chengjiang, Hongta, Jiangchuan, Huaning, Tonghai, Jianshui, and Mile. The study area has the highest elevation at nearly 2799 m, covering a wide spectrum of landscapes such as cities, villages, fields, lakes, and mountains; Fuxian Lake, Qilu Lake, and Xingyun Lake are renowned as the “three pearls” of Yuxi City. In September 2021, Yuxi City held a press conference in which the municipal government released the order for “lake renovation”, with a vision to achieve economic development by protecting the three lakes and make the “three-lake” basin a demonstration zone for the practicing the theory that “lucid waters and lush mountains are invaluable assets”. The ecology of the “three-lake” basin is closely related to the ecology of its surrounding areas, and increased human intervention that accompanies social and economic advances inflicts damage on the ecology. Therefore, it is necessary to understand the dynamic changes of the ecology in the “three-lake” basin so that we can determine effective measures to improve the environment there.

2.2. Data Resources and Image Processing

The remote sensing images of the study area were obtained from the U.S. Geological Survey, and the Landsat-5 TM and Landsat-8 OLI images (path 129, row 43) of the study area in 2001, 2006, 2011, 2015, and 2021 were selected. The cloudiness of the selected images is less than 1% (Table 1), the data resolution is 30 m, and the time for image acquisition is mainly from February to April. The data were acquired and processed via operations such as radiometric calibration, atmospheric correction, and cropping; the land use data were obtained from GlobeLand30.

2.3. Methodology

2.3.1. Single Factor Calculation

The RSEI model has four factors: greenness, humidity, dryness, and heat. In this study, the normalized difference vegetation index (NDVI) is chosen to indicate greenness; humidity is represented by the humidity component WET in the tasseled-cap transform, and the coefficients differ as the sensors vary. Dryness is represented by the normalized differential build-up and bare soil index (NDBSI), which is constructed based on the bare soil index (SI) and building index (IBI); the land surface temperature (LST) is used to represent heat. Table 2 lists the specific equations for each index. In all calculation equations, ρ_R, ρ_G, and ρ_B denote red, green, and blue bands, respectively; ρ_NIR, ρ_SWIR1, and ρ_SWIR2 refer to surface reflectance of the corresponding NIR, SWIR1, and SWIR2 spectral bands, respectively, of Landsat TM and OLI images; A, B, C, D, E, and F are all constants and vary from sensor to sensor. In the calculation equations for LST, ε represents the surface specific emissivity, B(T_S) is the blackbody thermal radiation brightness, L_d is the atmospheric downward radiation, L_up is the atmospheric upward radiation, $\tau$ is the atmospheric transmittance in the thermal infrared band, and K₁ and K₂ are the constants preset before the satellite launch [30].

2.3.2. Construction of the Remote Sensing Ecological Index

In this study, the principal component transform technique was used to construct the remote sensing ecological index (RSEI). Principal component analysis (PCA) is a multidimensional data compression technique that compresses information from multiple variables into a few characteristic components through linear transformation and sequential vertical rotation of the coordinate axes [31]. Since the four indicators (wetness, greenness, dryness, and heat) differ in their quantitative values, in this paper, they were normalized using the extreme value normalization method, and then PCA was performed to rotate the spatial axes of the feature spectrum to eliminate the correlation between these indicators and focus on the first 1–2 principal components. In the first principal component PC1, the loadings of ecologically favorable greenness and humidity have the same sign, while the loadings of ecologically unfavorable dryness and heat have opposite signs to those of greenness and humidity [32]. The separation of indicators into ecologically favorable indicators and ecologically unfavorable ones showed that PC1 had a clear ecological significance and a principal component matrix of these indicators was obtained. Another advantage of using PCA is that the weights of each indicator are not determined manually but automatically and objectively based on the contribution of the indicator to each principal component, thereby avoiding the risk of subjective assignment of weights that might lead to biased results [33]. In order to avoid inconsistency of scale and the influence of errors such as noises, the data within the confidence interval of 2–98% were selected for normalization of the PCA results, and the result of PC1 was selected and processed to obtain the remote sensing ecological index RSEI₀ which was normalized to obtain the normalized remote sensing ecological index RSEI. Thus, a larger value of RSEI indicates better ecological quality, whereas a lower value suggests poorer ecological quality.

(9)$RSE{I}_0=1-[PC1(f(NDVI,WET,NDBSI,LST))]$

(10)$RSEI=(RSE{I}_0-RSE{I}_0{}_{{}_{MIN}})/(RSE{I}_{0_{MAX}}-RSE{I}_{0_{MIN}})$

2.3.3. Change Trend of Ecological Quality

Sen + MK trend analysis: linear regression is a classical method to describe changes in long time series [34], but the time series are required to conform to a normal distribution and be susceptible to noise interference [35]. The calculation equations are as follows:

(11)$\beta =mean(\frac{x_j-x_i}{j-i}),\forall j>i$

(12)$S={\displaystyle \sum _{i=1}^{n-1}{\displaystyle \sum _{j=i+1}^{n}sgn(x_j-x_i)}}$

(13)$sgn(x_j-x_i)=\{\begin{array}{ccc}1& & x_j-x_i>0\\ \mathit{0}& & x_j-x_i=0\\ -1& & x_j-x_i<0\end{array}$

Hurst exponent: the Hurst exponent (H) reflects the autocorrelation of the time series as well as the hidden long-term trend in the series [36,37]. The number H > 0.5 implies that the future change is consistent with the past, indicating persistence, and the closer H is to 1, the stronger is the persistence; 0 < H < 0.5 implies that the future change is the opposite of the past, indicating anti-persistence, and the closer H is to 0, the stronger is the anti-persistence [38]. The calculation equations are shown below:

(14)$R_a=\max (X_t{}_{,a})-\min (X_t{}_{,a})$

(15)$\log (\frac{1}{A}{\displaystyle \sum _{a=1}^A(R_a/S_a)})=\log (K)+H\times \log (n)$

where S_a denotes the sample standard deviation of the a-th subinterval.

3. Results

3.1. Spatiotemporal Changes of Single Indicators

The spatial and temporal variations of each ecological quality indicator for the study area were analyzed from 2001 to 2021. The Mann–Kendall test with Sen’s slope (Sen + MK method) was used for multi-year trend analysis of each indicator, and the result is shown in Table 3. As the table shows, the area that witnessed a sharp decrease of the green index (NDVI) over the years was 11.87 km², accounting for about 0.5% of the total of the study area that passed the 95% significance test, and the area with slightly decreasing NDVI was 262.01 km², accounting for about 10.3% of the total area that passed the significance test, which is mainly observed in the urban part of the city near the “three-lake” basin. Due to the considerable reduction of vegetation coverage caused by urban expansion, the overall multi-year vegetation index in the study area registered a downward trend. The area of land where the humidity index remained stable or unchanged was 1736.56 km², accounting for 68.5% of the total area in the study area passing the significance test, and the overall humidity was stable despite a slight decrease. As mentioned before, the dryness index (NDBSI) was obtained by combining the bare soil index (SI) and the building index (IBI). As Figure 3 shows, a decrease in dryness was observed mainly in regions where the greenness index increased, and its change trend was opposite to that of the greenness index. The heat index (LST) decreased around Fuxian Lake and some parts of Jiangchuan and Tonghai but increased in Chengjiang and northwest of the study area; the area that marked an increased LST was 251.55 km², whereas the area witnessing a decreased LST was 201.46 km²; the LST index showed a stable trend over the 20 years. All indicators remained stable or unchanged over the 20 years of study, except that the greenness index (NDVI) showed significant increase in part, and the dryness index (NDBSI) partly presented significant decreases.

3.2. Temporal and Spatial Changes of Ecological Quality

As shown in Table 4, the eigenvalues of the first principal component (PC1) in 2001, 2006, 2011, 2015, and 2021 were 0.18, 0.21, 0.19, 0.19, and 0.19, respectively; its five-year eigenvalue contribution rates were 73.2%, 81.9%, 76.7%, 75.9%, and 74.8%, respectively; and its cumulative contribution rate was 70.0% or above, which means PC1 collected most of the information of all four indicators, and this is consistent with the actual situation of the study area. In general, NDVI and WET are positively correlated with ecological well-being (greater greenness and humidity correspond to higher vegetation coverage), whereas NDBSI and LST are negatively correlated with ecological well-being (stronger dryness and heat indicate a drier land surface) [39]. In PC1, the values of NDVI and WET remained positive, and the values of NDBSI and LST remained negative throughout the 20 years, but this pattern was not observed in other principal components. Therefore, PC1 was considered to be suitable for constructing remote sensing ecological indices and for evaluating the ecological quality of the “three-lake” basin.

Principal component analysis and further data processing reveal that the RSEI of the “three-lake” basin in Yuxi increased first and then decreased over the 20 years of study. To quantify the spatial and temporal distribution of ecological quality in the study area and better evaluate the environment of the “three-lake” basin, the authors referred to the “Technical Specification for Ecological Environmental Status Evaluation” promulgated by the Ministry of Ecology and Environment of the People’s Republic of China in 2015. Then in view of the actual ecological quality characteristics of the study area, the RSEI values in 2000, 2010, and 2020 were divided into five levels with 0.2 as the dividing interval, where 0.00 ≤ RSEI < 0.20 indicates poor ecological quality (I), 0.20 ≤ RSEI < 0.40 indicates average ecological quality (II), 0.40 ≤ RSEI < 0.60 indicates medium ecological quality (III), 0.60 ≤ RSEI < 0.80 indicates good ecological quality (IV), and 0.80 ≤ RSEI ≤ 1.00 indicates excellent ecological quality (V). As Figure 4 shows, the RSEI of the study area is mainly between 0.20 and 0.80, and the area with poor ecological quality (Level I) increased first and then decreased, showing an overall decreasing trend from 242.85 km² in 2001 to 196.06 km² in 2021. In 2006, the ecological quality in the study area was relatively poor, and the total area of regions with fair ecological quality and below (Levels I and II) was 1028.14 km², accounting for 38.4% of the total area of the study area. As Figure 5 shows, regions of poor and fair ecological quality (Levels I and II) were mainly observed in the northeastern part of Chengjiang and the western part of Huaning and Jiangchuan, and since 2011, the ecological condition of the eastern part of Fuxian Lake has improved. In 2021, the area with good ecological quality (Level IV) throughout the study area was 372.73 km², an increase of 159.94 km² from 2001, but regions with weak ecology were still observed in Chengjiang, most of which were regions subject to frequent human activities near urban areas.

The area of the study area with medium ecological quality (Level III) has decreased over the past 20 years (Figure 6), as the area of RSEI Level III decreased by 81.88 km² between 2001 and 2011 (Table 5), which mainly transferred into areas of RSEI Level IV (303.42 km²) and RSEI Level I (107.63 km²). Most areas with good ecological quality remained in good ecological quality, without much transfer; half of the regions with poor ecological quality transferred into regions with better ecological quality. Thus, it can be concluded that the measures taken by the municipal government of Yuxi to address environmental problems in the “three-lake” basin over the 20 years are effective. However, there was still a shift of area from high ecological quality to low ecological quality, which mainly resulted from ecological degradation caused by extensive human activities and serious rock desertification, mainly in Huaning.

3.3. Trend Persistence of Ecological Quality Changes

The ecological quality of the study area over a span of 20 years was analyzed using the Sen + MK method (Table 6). The results showed that the ecological quality of the study area was stable over the 20 years, with ecological quality in 61.7% of the total area of the study area (1655.20 km²) remaining unchanged; ecological degradation occurred mainly in the northern part of Chengjiang and parts of Tonghai and Huaning around the “three-lake” basin (Figure 7). Analysis of changes in each index revealed that the increase of LST and the weakening of NDVI in the area might be caused by human activities. The improvement in ecological conditions was mainly observed in Chengjiang and Tonghai: 480.23 km² showed slight improvement, accounting for 17.9% of the total area of the study area, and significant improvement was observed in 145.32 km² of land, accounting for 5.4% of the study area.

Analyses showed that the Hurst exponent in the study area ranged from 0.06 to 0.99, with a mean value of 0.70; the area with a Hurst exponent less than 0.50 was 298.71 km², accounting for 13.0% of the total area; the area with a Hurst exponent greater than 0.50 was 1990.36 km², accounting for 87.0% of the total of the studied area. The Hurst exponent is divided into four levels: H < 0.35 indicates strong anti-persistence, which accounts for 60.17 km² (2.6%) of the study area; 0.35 ≤ H < 0.50 indicates weak anti-persistence, which accounts for 238.55 km² (10.4%) of the study area, mainly distributed in Jiangchuan and Tonghai; 0.50 ≤ H < 0.65 means weak positive persistence; and 0.65 ≤ H indicates strong positive persistence. Figure 8 (left) shows the spatial distribution of the Hurst exponent in the studied basin. The ecological trend analysis result was coupled with the Hurst-index distribution map to evaluate the trend persistence of ecological quality of the study area (Figure 8, right), and the specific statistics are shown in Table 7. Overall, the trend of ecological quality changes in 88.0% of the study area was relatively stable; regions that witnessed continued ecological degradation took up 113.91 km² (5.0%) of the study area, mainly distributed in Chengjiang, Jiangchuan, and the urban part of Tonghai, with few distributed in other counties and districts; regions that showed continuous ecological improvement took up 142.43 km² (6.2%) of the study area, mainly distributed in the eastern and southern parts of the “three-lake” basin, especially in the southern part of Tonghai.

4. Discussion

Ecological protection within the watershed of nine plateau lakes in Yunnan Province has been the focus of work of the provincial government; Yuxi City accommodates three highland lakes—Fuxian Lake, Xinyun Lake, and Qilu Lake—with a capacity of 20.95 billion m³, accounting for 69.3% of the nine highland lakes in Yunnan Province. Since the 1980s, developed countries in Europe and the United States have been accentuating ecological protection in their water resource management policies, paying increasing attention to the state of ecological resources and ecological quality of watersheds. In China, however, it was as late as the start of the 21st century that ecological protection began to draw governmental attention [40]. Studying the surrounding ecological conditions of the plateau lakes in Yuxi could facilitate investigations of the quality of the water environment in the region.

In this study, the ecological quality of the surrounding area of the three lakes in Yuxi from 2001 to 2021 was analyzed, and the development trend was investigated. The land use pattern of the study area in 2001, 2011, and 2021 was coupled with the ecological index distribution pattern to explore the relationship between land use types and ecological quality. As Table 8 shows, the area of arable land gradually decreased (from 1228.73 km² to 744.94 km²), whereas the woodland area gradually increased (from 764.70 km² to 1278.60 km²), which mirrored the effectiveness of the government’s “grain-for-green” initiative. As the expansion of woodland alleviated the impact of greenhouse gas emissions and promoted species diversity, most regions in the study area showed an RSEI between 0.40 and 1.00. This means the land use type of woodland played a contributing role in the ecological well-being of the area under study. Districts such as Jiangchuan, Chengjiang, and Huaning—where the lakes (water bodies) took up large areas, and which had a low level of vegetation coverage—witnessed different degrees of soil erosion, thus were subject to poor ecological quality.

The current land-use situation in the urban sections in the study area is rather complicated: there is a waste of land resources to varied degrees; blind expansion of towns and villages to the periphery has encroached on agricultural land, while irrational use of built-up land in urban areas remains, marked by a low utilization rate of land in many urban sections; human activities have caused damage to the ecosystem, so the ecological quality of built-up land is poor.

To further characterize the impact of each indicator on the ecological quality, an ecological quality model was established using a 30 m × 30 m grid to sample the NDVI, WET, LST, NDBSI, and RSEI of each period in the map of the study area; a total of 500 sampling points were collected for each image, and the four indicators were regressed with the RSEI for analysis. The modeling result showed that the four indicators were not excluded in the stepwise regression of each year, and the models all passed the 1% significance test, indicating that the selected indicators are all key indicators of ecological environment quality. Analysis of the regression coefficients of each indicator shows that the coefficients of NDVI and WET are positive, which proves that NDVI and WET play a role in promoting the ecological environment quality. On the contrary, the coefficients of NDBSI and LST are negative, which proves that these factors play a suppressive role in ecological quality. This can be understood more intuitively by the three-dimensional scatter plot in Figure 9. The conclusion is also consistent with those reached by some previous works [41,42], so the model factors are correctly selected and the results are representative.

5. Conclusions

In this study, the ecological quality in the basin of three highland lakes in Yunnan, i.e., Fuxian Lake, Xinyun Lake, and Qilu Lake, was evaluated using Landsat images and remote sensing ecological indices in 2001, 2006, 2011, 2015, and 2021; meanwhile, the ecological quality of the study area over the 20 years was analyzed spatially and temporally using Sen + MK trend analysis and the Hurst index to explore the ecological status and development trend of the study area.

• Analysis of ecological quality indicators based on Landsat images showed that all the ecological quality indicators presented a stable trend over the 20 years overall, but some regions witnessed a significant increase in the greenness index and a sharp decrease in the dryness index over the 20 years of study.

• Calculation of the remote sensed ecological index (RSEI) by principal component analysis showed that the RSEI in the study area remained between 0.20 and 0.80 over the 20 years as of 2001; the area with poor ecological quality first increased but decreased, in general, from 242.85 km² in 2001 to 196.06 km² in 2021. However, there were still some areas with weak ecological quality in Chengjiang, most of which are located near urban areas due to the deterioration of ecological environment quality caused by urban expansion.

• The Sen + MK method and the Hurst exponent were used to analyze the multi-year trend and persistence of ecological quality changes in the study area. The results showed that the overall trend of ecological quality changes in the “three-lake” basin was stable; the declining RSEI was mainly found in urban regions of Chengjiang, Jiangchuan, and Tonghai, whereas continuous ecological improvement was mainly observed in the eastern and southern parts of the “three-lake” basin, especially in the southern part of Tonghai.

• In the future, the “three-lake” basin should accelerate the construction of ecological wetlands, for example, by returning fields to lakes; the areas around the lakes within the basin should be reasonably planned for greening; and efforts should be made to balance human activities and ecological quality. Regions with poor ecological quality currently should protect woodlands, grasslands, and forests, and take engineering and comprehensive measures to address environmental problems such as soil erosion and stone desertification to improve the regional ecological quality. Regions with a medium level of ecological quality and above should continue to strengthen the organic link between forest, grassland, and other types of landscapes, strictly control the intensity of land development, reduce ecological risks, and manage different land use types in the basin in a systematic manner.

Author Contributions

Study design, Y.S. and J.L.; data collection, Y.Y. and W.Z.; data analysis, Y.S. and J.L.; figures, W.Z.; writing—original draft preparation, Y.S.; writing—review and editing, Y.Y. and W.Z. All authors have read and agreed to the published version of the manuscript.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

All data in this study were correctly referenced. The remote sensing images of the study area were obtained from the U.S. Geological Survey (https://www.usgs.gov/ accessed on 12 March 2022). The data were acquired and processed via operations such as radiometric calibration, atmospheric correction, and cropping; the land use data were obtained from GlobeLand30 (http://www.globallandcover.com/ accessed on 15 March 2022).

Conflicts of Interest

The authors declare no conflict of interest.

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Figures and Tables

Figure 1. Research workflow of the present work.

Figure 2. Overview of the study area.

Figure 3. Diagrams for change trends of single indicators throughout the years of study.

Figure 4. Temporal changes of RSEI in the study area.

Figure 5. Spatial changes of RSEI in the study area over the years of study.

Figure 6. Histogram of the percentage of RSEI grade area by year.

Figure 7. Maps of ecological quality change trend over the years of study.

Figure 8. Hurst index distribution (left) and trend persistence of ecological quality changes (right).

Figure 9. 3D Scattered feature map.

Table 1

Image data list.

Sensor	Path	Row	Data Date	Cloud Cover
TM	129	43	2 March 2001	0.06
	129	43	1 April 2006	0.26
	129	43	26 February 2011	0.02
OLI	129	43	9 March 2015	0.01
	129	43	5 February 2021	0.61

Table 2

Calculation equations of single indicators of the RSEI model.

Index	Equation
NDVI	$$NDVI=({\rho }_{NIR}-{\rho }_R)/({\rho }_{NIR}+{\rho }_R)$$	(1)
WET	$WET=A{\rho }_B+B{\rho }_G+C{\rho }_R+D{\rho }_{NIR}+E{\rho }_{SWIR1}+F{\rho }_{SWIR2}$	(2)
NDBSI	$SI=\left[\left({\rho }_{SWIR1}+{\rho }_R\right)-\left({\rho }_{NIR}+{\rho }_B\right)\right]/\left[\left({\rho }_{SWIR1}+{\rho }_R\right)+\left({\rho }_{NIR}+{\rho }_B\right)\right]$	(3)
	$IBI=\frac{2{\rho }_{SWIR1}/\left({\rho }_{SWIR1}+{\rho }_{NIR}\right)-\left({\rho }_{NIR}/\left({\rho }_{NIR}+{\rho }_R\right)+{\rho }_G/\left({\rho }_G+{\rho }_B\right)\right)}{2{\rho }_{SWIR1}/\left({\rho }_{SWIR1}+{\rho }_{NIR}\right)+\left({\rho }_{NIR}/\left({\rho }_{NIR}+{\rho }_R\right)+{\rho }_G/\left({\rho }_G+{\rho }_B\right)\right)}$	(4)
	$NDBSI=(IBI+SI)/2$	(5)
LST	$LST=[\epsilon B(T_S)+\left(1-\epsilon \right)Ld]\tau +Lu$	(6)
	$B(T_S)=[L-L_{UP}-\tau (1-\epsilon )Ld]/\tau \epsilon$	(7)
	$T_S=K_2/ln\left(K_1/B\left(T_S\right)+1\right)$	(8)

Table 3

Areas of changes in single indicators over the years of study.

Indicators	Severe Decrease/(km2)	Slight Decrease/(km2)	Stable/(km2)	Slight Increase/(km2)	Severe Increase/(km2)
NDVI	11.87	262.01	1442.97	693.79	130.82
WET	35.90	441.93	1736.56	284.41	35.77
NDBSI	172.69	616.27	1508.13	208.71	8.81
LST	23.56	177.90	2228.06	227.04	24.51

Table 4

Principal component analysis of each index.

Years	Indicators	PC1	PC2	PC3	PC4
2001	NDVI	0.23	0.10	−0.07	−0.03
	WET	0.20	−0.02	−0.03	0.08
	NDBSI	−0.27	−0.12	0.04	−0.03
	LST	−0.10	0.11	0.10	0.01
	Eigenvalues	0.18	0.04	0.02	0.01
	Eigenvalue contribution rate	73.22	15.78	7.57	3.43
	Cumulative contribution rate	73.22	89.00	96.57	100.00
2006	NDVI	0.23	0.08	−0.08	−0.02
	WET	0.27	0.10	−0.04	0.02
	NDBSI	−0.26	−0.01	−0.03	0.04
	LST	−0.15	0.09	0.11	0.00
	Eigenvalues	0.21	0.02	0.02	0.00
	Eigenvalue contribution rate	81.95	9.34	7.84	0.87
	Cumulative contribution rate	81.95	91.29	99.13	100.00
2011	NDVI	0.17	0.12	−0.07	−0.05
	WET	0.31	−0.04	0.09	−0.02
	NDBSI	−0.21	−0.06	−0.08	0.06
	LST	−0.04	0.15	0.04	0.06
	Eigenvalues	0.19	0.03	0.02	0.00
	Eigenvalue contribution rate	76.73	13.10	8.68	1.49
	Cumulative contribution rate	76.73	89.83	98.51	100.00
2015	NDVI	0.22	−0.09	0.08	0.04
	WET	0.02	0.07	−0.04	0.09
	NDBSI	−0.31	−0.12	0.01	0.02
	LST	−0.15	−0.11	−0.11	0.00
	Eigenvalues	0.19	0.03	0.03	0.00
	Eigenvalue contribution rate	75.88	13.09	10.41	0.62
	Cumulative contribution rate	75.88	88.97	99.38	100.00
2021	NDVI	0.21	−0.09	0.09	0.04
	WET	0.01	0.07	−0.04	0.09
	NDBSI	−0.33	−0.11	0.02	0.02
	LST	−0.13	−0.12	−0.10	0.00
	Eigenvalues	0.19	0.03	0.03	0.00
	Eigenvalue contribution rate	74.77	12.63	11.62	0.98
	Cumulative contribution rate	74.77	87.40	99.02	100.00

Table 5

Ecological quality transfer matrix in 2001–2021 (Unit: km²).

Time Period	RSEI Level	I	II	III	IV	V	Total (Transferred Out)
2001–2011	I	103.92	105.26	25.47	6.67	1.53	242.85
	II	79.04	374.77	247.62	58.72	7.54	767.69
	III	14.97	107.63	360.25	303.42	31.52	817.79
	IV	4.21	19.53	86.15	353.43	176.31	639.63
	V	0.98	3.42	16.41	60.58	131.40	212.79
	Total (transferred in)	203.12	610.61	735.90	782.82	348.30	2680.74
2011–2021	I	85.00	95.38	19.39	3.06	0.28	203.11
	II	71.82	351.71	159.21	25.81	2.05	610.6
	III	26.51	188.25	332.52	174.26	14.37	735.91
	IV	10.46	75.94	192.06	363.94	140.42	782.82
	V	2.26	14.55	34.67	81.20	215.61	348.29
	Total (transferred in)	196.05	725.83	737.85	648.27	372.73	2680.74

Table 6

Spatial trend of ecological quality changes over the years of study.

Spatial Trend	Pixels	Area/(km2)	Percentage
Severe Degradation	47,552	42.80	1.6%
Slight Degradation	396,886	357.20	13.3%
Stable	1,839,109	1655.20	61.7%
Slight Improvement	533,584	480.23	17.9%
Significant Improvement	161,467	145.32	5.4%

Table 7

Persistence trend of ecological quality changes in the study area.

Spatial Trend	Pixels	Area/(km2)	Percentage
Continuous Degradation	126,562	113.91	5.0%
Anti-persistent Degradation	12,810	11.53	0.5%
Stable	2,237,178	2013.46	88.0%
Anti-continuous Improvement	8508	7.66	0.3%
Continuous Improvement	158,255	142.43	6.2%

Table 8

Relationship between land use types and RSEI (Unit: km²).

Years	Land	RSEI
		I	II	III	IV	V	Total
2001	Grass	74.43	257.02	198.1	54.87	4.49	588.91
	Arable Land	157.42	410.19	351.92	212.29	96.91	1228.73
	Shrubland	0.83	15.21	18.36	9.96	1.36	45.72
	Building Land	5.52	27.40	6.17	3.24	1.16	43.49
	Woodland	4.15	55.24	236.88	359.71	108.72	764.70
	Total	242.35	765.06	811.43	640.07	212.64	2671.55
2011	Grass	61.49	187.79	205.58	127.83	23.1	605.79
	Arable Land	126.88	325.47	355.95	295.93	94.72	1198.95
	Shrubland	1.85	13.28	16.14	14.23	3.86	49.36
	Building Land	4.95	28.56	6.01	3.36	0.89	43.77
	Woodland	7.52	54.19	150.14	338.62	223.21	773.68
	Total	202.69	609.29	733.82	779.97	345.78	2671.55
2021	Grass	81.29	205.02	93.96	13.31	0.94	394.52
	Arable Land	89.87	283.92	229.03	114.68	27.44	744.94
	Shrubland	0.74	5.26	13.48	20.86	6.05	46.39
	Building Land	13.86	127.11	47.23	16.08	2.82	207.10
	Woodland	9.68	102.90	351.03	479.99	335.00	1278.60
	Total	195.44	724.21	734.73	644.92	372.25	2671.55

References

1. Pei, J.; Zhong, K.; Li, J.; Xu, J.; Wang, X. ECNN: Evaluating a cluster-neural network model for city innovation capability. Neural Comput. Appl.; 2022; 34, pp. 12331-12343. [DOI: https://dx.doi.org/10.1007/s00521-021-06471-z]

2. Li, C.; Wu, Y.; Gao, B.; Zheng, K.; Wu, Y.; Li, C. Multi-scenario simulation of ecosystem service value for optimization of land use in the Sichuan-Yunnan ecological barrier, China. Ecol. Indic.; 2021; 132, 108328. [DOI: https://dx.doi.org/10.1016/j.ecolind.2021.108328]

3. Hansen, M.; Li, H.; Svarverud, R. Ecological civilization: Interpreting the Chinese past, projecting the global future. Glob. Environ. Chang.; 2018; 53, pp. 195-203. [DOI: https://dx.doi.org/10.1016/j.gloenvcha.2018.09.014]

4. Gare, A. China and the struggle for ecological civilization. Capital. Nat. Soc.; 2012; 23, pp. 10-26. [DOI: https://dx.doi.org/10.1080/10455752.2012.722306]

5. Simpson, J.; Santo Domingo, J.; Reasoner, D. Microbial source tracking: State of the science. Environ. Sci. Technol.; 2002; 36, pp. 5279-5288. [DOI: https://dx.doi.org/10.1021/es026000b]

6. Strobel, C.; Buffum, H.; Benyi, S.; Paul, J. Environmental Monotoring and Assessment Program: Current Status of Virginian Province (US) Estuaries. Environ. Monit. Assess.; 1999; 56, pp. 1-25. [DOI: https://dx.doi.org/10.1023/A:1005911822444]

7. Ferguson, B. The Concept of Landscape Health. Acad. Press.; 1994; 40, pp. 129-137. [DOI: https://dx.doi.org/10.1006/jema.1994.1009]

8. Zhou, H. Study on ecological environmental quality as assessment index system of Xinjiang. China Environ. Sci.; 2000; 20, pp. 150-153.

9. Zhou, W.; He, B. Eco-environmental quality assessment of Ruoergai county in Sichuan province based on multi-sources remote sensing data. J. Geo-Inf. Sci.; 2014; 4, pp. 54-62.

10. Zhang, Z.; Peng, X.; Chen, X.; Wang, C. Spatial information quantitative analysis method on integrated assessment and monitoring of ecological environment. Environ. Sci.; 1999; 20, pp. 68-72.

11. Xu, P.; Zhao, D. Ecological environmental quality assessment of Hangzhou urban area based on RS and GlS. Chin. J. Appl. Ecol.; 2006; 17, pp. 1034-1038.

12. Gao, Q.; Shi, X.; Zhang, C.; Zhang, M.; Ma, H. Dynamic assessment and prediction on quality of agricultural eco-environment in county area. Trans. Chin. Soc. Agric. Eng.; 2014; 30, pp. 228-237.

13. Morris, P.; Therivel, R. Methods of Environmental Impact Assessment; University of British Columbia: Vancouver, BC, Canada, 1995.

14. Ladson, A.; White, L.; Doolan, J.; Finlayson, B.; Hart, B.; Lake, P.; Tilleard, J. Development and testing of an Index of Stream Condition for waterway management in Australia. Freshw. Biol.; 1999; 41, pp. 453-468. [DOI: https://dx.doi.org/10.1046/j.1365-2427.1999.00442.x]

15. Xie, H.; Yao, G.; Liu, G. Spatial evaluation of the ecological importance based on GIS for environmental management: A case study in Xingguo county of China. Ecol. Indic.; 2015; 51, pp. 3-12. [DOI: https://dx.doi.org/10.1016/j.ecolind.2014.08.042]

16. Guha, S.; Govil, H.; Mukherjee, S. Dynamic analysis and ecological evaluation of urban heat islands in Raipur city, India. J. Appl. Remote Sens.; 2017; 11, 036020. [DOI: https://dx.doi.org/10.1117/1.JRS.11.036020]

17. Li, X.; Tian, M.; Wang, H. Development of an ecological security evaluation method based on the ecological footprint and application to a typical steppe region in China. Ecol. Indic.; 2014; 39, pp. 153-159. [DOI: https://dx.doi.org/10.1016/j.ecolind.2013.12.014]

18. Reza, M.; Abdullah, S. Regional Index of Ecological Integrity: A need for sustainable management of natural resources. Ecol. Indic.; 2011; 11, pp. 220-229. [DOI: https://dx.doi.org/10.1016/j.ecolind.2010.08.010]

19. Orfanidis, S.; Panayotidis, P.; Stamatis, N. An insight to the ecological evaluation index (EEI). Ecol. Indic.; 2003; 3, pp. 27-33. [DOI: https://dx.doi.org/10.1016/S1470-160X(03)00008-6]

20. Ma, P.; Ye, G.; Peng, X.; Liu, J.; Qi, J.; Jia, S. Development of an index system for evaluation of ecological carrying capacity of marine ecosystems. Ocean. Coast. Manag.; 2017; 144, pp. 23-30. [DOI: https://dx.doi.org/10.1016/j.ocecoaman.2017.04.012]

21. Hang, X.; Luo, X.; Cao, Y.; Li, Y. Ecological quality assessment and the impact of urbanization based on RSEl model for Nanjing, Jiangsu Province, China. Chin. J. Appl. Ecol.; 2020; 31, pp. 219-229.

22. Xu, H. A remote sensing urban ecological index and its application. Acta Ecol. Sin.; 2013; 33, pp. 7853-7862. (In Chinese)

23. Wang, Z.; He, X. Assessments of ecological quality in Jinjiang district of Chengdu city using the FVC and RSEI models. J. Ecol. Rural Environ.; 2021; 37, pp. 492-500. [DOI: https://dx.doi.org/10.19741/j.issn.1673-4831.2020.0511]

24. Liu, S.; Yuan, Y.; Zhao, H.; Li, Q. Analysis of ecological environment changes in hydropower development zone based on RSEI: A case study in the middle and lower reaches of the Qingjiang River, China. J. Ecol. Rural Environ.; 2019; 35, pp. 1361-1368.

25. Zhang, J.; Zhou, Q.; Cao, M.; Liu, H. Spatiotemporal Change of Eco-Environmental Quality in the Oasis City and Its Correlation with Urbanization Based on RSEI: A Case Study of Urumqi, China. Sustainability; 2022; 14, 9227. [DOI: https://dx.doi.org/10.3390/su14159227]

26. Boori, M.; Choudhary, K.; Paringer, R.; Kupriyanov, A. Spatiotemporal ecological vulnerability analysis with statistical correlation based on satellite remote sensing in Samara, Russia. J. Environ. Manag.; 2021; 285, 112138. [DOI: https://dx.doi.org/10.1016/j.jenvman.2021.112138]

27. Ariken, M.; Zhang, F.; Liu, K.; Fang, C.; Kung, H. Coupling coordination analysis of urbanization and eco-environment in Yanqi Basin based on multi-source remote sensing data. Ecol. Indic.; 2020; 114, 106331. [DOI: https://dx.doi.org/10.1016/j.ecolind.2020.106331]

28. Saleh, S.; Amoushahi, S.; Gholipour, M. Spatiotemporal ecological quality assessment of metropolitan cities: A case study of central Iran. Environ. Monit. Assess.; 2021; 193, 305. [DOI: https://dx.doi.org/10.1007/s10661-021-09082-2] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/33900465]

29. Zhong, X.; Xu, Q. Monitoring and Evaluation of Ecological Environment Changes in Yuxi City Based on RSEI Model. Res. Soil Water Conserv.; 2021; 28, pp. 350-357.

30. Liu, X.; Zhou, Q.; Zhou, L.; Meng, H.; Li, M.; Peng, C. RSEI-Based Dynamic Monitoring of Ecological Quality of the Soil and Water Conservation Functional Area in the Chongqing Section of the Three Gorges Reservoir Area. Res. Soil Water Conserv.; 2021; 28, pp. 278-286.

31. Hu, X.; Xu, H. A new remote sensing index for assessing the spatial heterogeneity in urban ecological quality: A case from Fuzhou City, China. Ecol. Indic.; 2018; 89, pp. 11-21. [DOI: https://dx.doi.org/10.1016/j.ecolind.2018.02.006]

32. Xu, H.; Deng, W. Rationality Analysis of MRSEI and Its Difference with RSEI. Remote Sens. Technol. Appl.; 2022; 37, pp. 1-7.

33. Xu, H.; Wang, M.; Shi, T.; Guan, H.; Fang, C.; Lin, Z. Prediction of ecological effects of potential population and impervious surface increases using a remote sensing based ecological index (RSEI). Ecol. Indic.; 2018; 93, pp. 730-740. [DOI: https://dx.doi.org/10.1016/j.ecolind.2018.05.055]

34. Yilmaz, B. Analysis of hydrological drought trends in the gap region (southeastern Turkey) by Mann-Kendall test and innovative sen method. Appl. Ecol. Environ. Res.; 2019; 17, pp. 3325-3342. [DOI: https://dx.doi.org/10.15666/aeer/1702_33253342]

35. Chervenkov, H.; Slavov, K. Theil-Sen estimator vs. ordinary least squares–trend analysis for selected ETCCDI climate indices. Comptes Rendus Acad. Bulg. Sci.; 2019; 72, pp. 47-54. [DOI: https://dx.doi.org/10.7546/CRABS.2019.01.06]

36. Fyodorov, Y.; Khoruzhenko, B.; Simm, N. Fractional Brownian motion with Hurst index $ H= 0$ and the Gaussian unitary ensemble. Ann. Probab.; 2016; 44, pp. 2980-3031. [DOI: https://dx.doi.org/10.1214/15-AOP1039]

37. Gairing, J.; Imkeller, P.; Shevchenko, R.; Ciprian, T. Hurst index estimation in stochastic differential equations driven by fractional Brownian motion. J. Theor. Probab.; 2020; 33, pp. 1691-1714. [DOI: https://dx.doi.org/10.1007/s10959-019-00925-w]

38. Jing, J.; Xu, Y.; Wang, Y.; Dou, S.; Yin, M. Spatiotemporal Variation Characteristics of Drought and Flood in the Southwest Karst Area from 1960 to 2019. Res. Soil Water Conserv.; 2021; 28, pp. 179-186.

39. Huang, H.; Chen, W.; Zhang, Y.; Qiao, L.; Du, Y. Analysis of ecological quality in Lhasa Metropolitan Area during 1990–2017 based on remote sensing and Google Earth Engine platform. J. Geogr. Sci.; 2021; 31, pp. 265-280. [DOI: https://dx.doi.org/10.1007/s11442-021-1846-8]

40. Chen, P.; Chen, X.; Yang, K.; Yao, Z.; Liu, Y. Overview and enlightenment of comprehensive evaluation on biodiversity and ecosystem services of inland aquatic ecosystems in Japan. Environ. Pollut. Control; 2020; 42, pp. 1060-1066.

41. Xu, H.; Wang, Y.; Guan, H.; Shi, T.; Hu, X. Detecting ecological changes with a remote sensing based ecological index (RSEI) produced time series and change vector analysis. Remote Sens.; 2019; 11, 2345. [DOI: https://dx.doi.org/10.3390/rs11202345]

42. Xiong, Y.; Xu, W.; Lu, N.; Huang, S.; Wu, C.; Wang, L.; Dai, F.; Kou, W. Assessment of spatial–temporal changes of ecological environment quality based on RSEI and GEE: A case study in Erhai Lake Basin, Yunnan province, China. Ecol. Indic.; 2021; 125, 107518. [DOI: https://dx.doi.org/10.1016/j.ecolind.2021.107518]