3.1. Spatiotemporal Change of Land Use Types
It can be seen from the spatial distribution of landscape types in CZTCA (
Figure 2), that forest is the dominant landscape type in the study area, and cultivated land is the second largest landscape type, followed by construction land and water. In terms of spatial distribution, the forest type was mainly distributed in the southeastern and southern mountainous and hilly areas of the study area; cropland was mainly distributed along relatively flat rivers and intermountain basins. Construction land was mainly distributed in the downtown areas of the three cities of CZT. Its spatial expansion showed a trend of radiation from the urban areas to the surrounding areas. The spatial distribution of water was mainly represented by the distribution of rivers, lakes, and reservoirs. The distribution of grassland and unused land was relatively scattered.
As shown in
Table 1, the changes in the area of each landscape type were mainly reflected in the reduction in forest and cropland and the continuous increase in construction land. From 2000 to 2020, the forest area decreased from 4633.08 km
2 to 4150.95 km
2, with an annual change rate of 0.52%; the cropland area decreased by an average of 0.79% per year, with a total decrease of 505.21 km
2. The construction land has the most significant growth rate among all landscape types, with an annual growth rate of 10.26%, and its area has increased from 474.79 km
2 in 2000 to 1448.66 km
2 in 2020. As shown in
Table 2, the transfer characteristics of forest, cropland, and construction land in the study area were significant from 2000 to 2020. The conversion areas of forest and cropland to construction land were 301.90 km
2 and 725.23 km
2 respectively. The results reflected the rapid development of the integration of the CZTCA at the expense of forest land and cultivated land.
The spatial distribution maps of land use types in the five periods were superimposed and analyzed in ArcGIS10.8, and the land use transfer matrix of each period was constructed. In order to display the analysis results of the land use transfer matrix more intuitively, the Sankey diagram of landscape transfer was constructed with the Origin software to visualize the transfer of each landscape type in the study area, as shown in
Figure 3.
In
Figure 3, the color bandwidth is the proportion of landscape types. It can be seen from the figure that, in the past 20 years, the landscape types of forest land and cultivated land in CZTCA have been mainly transferred out, and construction land has been mainly transferred in. Grassland, water bodies, and the amount of change in the transfer of unused land area are relatively small. The conversion between forest and cultivated land landscape types in the study area was reasonably frequent. With the implementation of ecological construction protection measures such as developed land protection and returning farmland to forest, the transfer between the two gradually decreased. Overall, with the acceleration of the integration process of the study area, the continuous expansion of construction land has led to the encroachment of urban ecological land, intensifying the contradiction between land use and environmental protection.
3.2. Analysis of Land Use Dynamics
It can be seen from the change in the dynamic degree of single land use (
Table 3) that the changes in the active degree of land use types were quite different in the past 20 years. In 2000–2005, the dynamic degree of construction land, unused land, and grassland changed rapidly, while the active degree of other land use types changed relatively slowly. From 2005 to 2011, the change rates of forest, cropland, construction land, and water were faster than those from 2000 to 2005. In 2011–2016, except for forest and grassland, the rate of change of other land use types decreased. In 2016–2020, the change rate of cropland, water, and unused land dynamic showed an upward trend compared to the previous period. Construction land had the most significant dynamic change in the past 20 years, reaching a maximum value of 9% between 2005 and 2011.
The comprehensive land use dynamic degree can reflect the overall dynamic change trend for regional land use types. In this study, the comprehensive dynamic degrees of land use in the study area during 2000–2005, 2005–2011, 2011–2016, and 2016–2020 were 2.21%, 2.82%, 2.67%, and 2.70%, respectively. Generally, it showed a trend of first rising followed by a slow decline and an overall rising trend (
Figure 4). The comprehensive land use dynamic degree in the study area was the highest between 2005 and 2011, with a value of 2.82%, indicating that the most frequent conversion among land use types occurred during this period.
3.3. Dynamic Changes of Eco-Environmental Quality
The normalized NDVI, WET, LST, and NDBSI were subjected to principal component analysis by calling the principal component analysis code in the GEE platform, and the results were exhibited in
Table 4. As shown in
Table 4, the contribution rate of the first principal component (PC1) in 2000, 2005, 2011, 2016, and 2020 were 72.78%, 72.31%, 76.82%, 84.22%, and 87.03%, respectively, with an average of more than 78%, indicating that most of the characteristic information of the four ecological indicators are integrated into PC1. Therefore, it is reasonable to characterize the quality of the regional eco-environment based on the data extracted from the first principal component. The NDVI and WET in PC1 were positive indicators, indicating that they positively affected the eco-environmental quality. The LST and NDBSI were negative indicators that harmed the eco-environmental quality. These results were in line with reality.
The value of the RSEI decreased from 0.722 in 2000 to 0.634 in 2020, indicating that the eco-environmental quality in CZTCA had deteriorated. From 2000 to 2011, the RSEI value in the study area showed a continuous downward trend, with the most significant decrease of 8.19% between 2005 and 2011. The RSEI value decreased from 0.664 in 2016 to 0.634 in 2020, with a decline of 4.52%. In the contribution rate of four ecological indicators of PC1, the NDVI and NDBSI had more significant effects on the RSEI than other indicators, indicating that greenness and dryness were closely related to the quality of the eco-environmental quality in CZTCA.
In order to show the eco-environment quality of the study area more clearly, the RSEI was classified into five classes with an interval of 0.2 [
43,
44]: worst (0, 0.2), poor (0.2, 0.4), medium (0.4, 0.6), good (0.6, 0.8), and excellent (0.8,1), and counted the area proportion of each RSEI level in 2000, 2005, 2011, 2016, and 2020; the result was shown in
Figure 5. In 2000–2020, the RSEI level of CZTCA was mainly excellent and good, and the proportions of these two levels in 2000, 2005, 2011, 2016, and 2020 were 77.43%, 77.13%, 62.28%, 69.61%, and 60.26%, respectively. The portion of areas with the worst and poor eco-environmental quality increased from 7.34% in 2000 to 16.79% in 2020. The percentage of areas with excellent eco-environmental quality fluctuated from 2000 to 2020, reaching its lowest value in 2011 and increasing significantly between 2011 and 2016, and then leveled off.
As shown in
Figure 6, the areas with good and excellent eco-environmental quality in CZTCA were most widely distributed, with prominent spatial differentiation characteristics. The good and excellent eco-environmental quality areas were distributed in a weave pattern, mainly in the study area’s northeastern and southern mountainous and hilly regions. These areas primarily were forestland with high altitudes and were less affected by human activities. The areas with the worst and poor eco-environmental quality were mainly distributed in the urban regions of CZT, which were principally construction land and were significantly affected by human activities. From 2000 to 2020, the areas with the worst and poor eco-environment quality spread radially outwards as a whole, indicating that the expansion of construction land in CZTCA had led to the deterioration of the ecological quality in the surrounding areas. In addition, the spatial distribution of eco-environmental quality (
Figure 6) largely coincided with the spatial distribution of land use types (
Figure 2).
In order to analyze the characteristics of the ecological environment quality change in different periods, this study used the difference method to detect the dynamic changes in the eco-environmental quality in CZTCA. The percentage changes in the RSEI levels are shown in
Table 5. In the past 20 years, the ratio of areas with improved eco-environmental quality (29.88%) was more significant than those with deteriorated eco-environmental quality (13.43%) only in 2011–2016. The results were mainly because, since 2011, the construction of regional ecological civilization had gradually received attention. Generally, the eco-environmental quality in CZTCA showed a deteriorating trend from 2000 to 2020. The ratio of areas with deteriorated eco-environmental quality was 37.45%, and the percentage of areas with improved eco-environmental quality was only 19.6%.
3.4. Eco-Environmental Effects of Land Use Change
In order to further explore the relationship between land use types and ecological environment, the land use type information and remote sensing ecological index were analyzed and processed to obtain the value of the remote sensing ecological index for each land use type. The result showed differences in the eco-environmental quality under different land use types (
Figure 7), which was consistent with the analysis result obtained by Zhu [
45]. From 2000 to 2020, the eco-environmental quality of forest and cropland was good, and where forest played an essential role in maintaining a sound ecological environment in the study area. The eco-environmental quality of construction land and unused land was poor.
From the analysis results of the land use transfer matrix, the land use transfer in the study area in 2000–2020 was dominated by the conversion between cropland and forest, as well as the conversion of cropland to construction land and forest to construction land. Therefore, Spearman analysis was applied to correlate these three land use transfer patterns with the improvement/deterioration in ecological quality. The results of the correlation analysis are shown in
Table 6.
As seen in
Table 6, the conversion of cropland to forest and the conversion of construction land to the forest had a significant positive effect on the improvement of the eco-environmental quality of the study area. The result indicated that the increase in the forest was the main reason for improving eco-environmental quality in CZTCA. In 2005–2011, the conversion of forest to construction land significantly impacted the deterioration of eco-environmental quality. Combined with the preceding, it was clear that the ratio of areas with a deteriorated eco-environmental rate (30.33%) was more significant than the ratio of areas with improved eco-environmental quality (10.11%) during 2005–2011. The results showed that the expansion of construction land had exacerbated the encroachment of forest, which was the main reason for the decline in the eco-environmental quality of CZTCA. Therefore, it is of great significance to rationally adjust the land use structure and optimize the spatial distribution of land use to improve the ecological environment and promote the region’s sustainable development.