*3.5. Forecast of Future RSEI*

The RSEI of the study area was predicted after the implementation of the PAR to analyze its sustained impact on regional EEQ. The ARIMA model starts from the timeseries itself and forecasts future data based on past behavioral data [44]. The expert modeler was used for prediction, where the autocorrelation and partial correlation coefficients of the model were all within the confidence zone of 95%. After prediction, the mean value of RSEI in the study area was predicted to improve to 0.7355 in 2025, and to 0.8603 in 2030. Continuous improvement and optimization were maintained from thereon. The spatial distribution effects are shown in Figure 9.

**Figure 9.** Forecast of the mean values of RSEI.

## **4. Discussion**

The results show that not all areas with a large number of relocated people have significantly increased RSEI. In northern Pu'an, Qinglong, and other severely rocky desertification areas (Figure 1) where the population density is relatively large (Figure A2) and the contradiction between peasants and land is prominent, the RSEI increase is more obvious in areas with a large number of people involved in PAR (Figure 7). However, in non-rocky desertification areas such as Ceheng and Wangmo in the southeast, where the population density is relatively small, the RSEI does not have an obvious increase, although the number of people relocated is relatively large. This shows that the implementation of PAR is more effective in areas with relatively poor ecology and a prominent contradiction between peasants and land, while in areas where the ecology is better, the contradiction between peasants and land is not prominent, and the ecological effect is not obvious (Table 5).


**Table 5.** Summary table of positive and negative effects of PAR.

In terms of the effectiveness of EEQ, the improvement in RSEI in the study area contributed by PAR is mainly due to the increase in NDVI and the decrease in NDISI. After relocation, the large-scale management of forestland will become a trend, with farmers moving away from contracted land, and farmland being transformed to forest, which contributes to the substantial increase in NDVI. Whether the substantial increase in NDVI will cause single-species forest and excessive water consumption needs further consideration [45]. The increase in WET is significantly smaller than the increase in NDVI, which is further verified. The RSEI of the study area is indeed significantly improved, while the RSEI of the urban areas to which farmers are moved is reduced. Does such an increase in RSEI really improve the living environment for local humans? During the implementation of PAR, the overall increase in RSEI in the study area was partially reduced (with a decrease in the more concentrated areas of population) and the comprehensive indicators showed improvement. Whether such improvement is conducive to sustainable social development is something our follow-up study needs to consider deeply (Table 5).

Biodiversity has an important impact on ecosystem services, and the impact of human activities on biodiversity is the focus of various scholars and organizations such as the United Nations Environment Programme (UNEP) [46,47]. After PAR, the number of land managers decreased and land-use patterns changed, which shifted the trend towards largescale land management. For the purpose of easy management and economic efficiency, the planting structure of the land tends to be homogeneous after large-scale management, which leads to an impact on biodiversity and the weakening of symbiosis among plants, thus further affecting the ecosystem service function [48]. In contrast, some areas become similar to ecological reserves, where the disturbance of human activities is reduced, and biological succession proceeds in an orderly manner, which is beneficial to biodiversity. There are many influencing factors and mechanisms of internal change of biodiversity, such as human disturbance, environmental factors, management methods, etc. The impact of PAR on biodiversity needs to be further studied in future work [49,50] (Table 5).

The global trend in terms of helping rural areas is to promote traditional and sustainable farming with nature-friendly measures, rather than relocating the rural population to cities [51]. PAR is the transfer of farmers from rural to urban areas, which destroys traditional rural culture, increases their cost of living and changes their way of life. The fact that farmers move to cities to take up jobs they are not good at, and that employment training is not sufficient to help low-ability farmers relocate successfully, also greatly increases the burden on the government, in addition to the fact that the minimum cost of living for relocated households is approximately 70% higher in cities than in rural areas. PAR poses a great challenge to the sustainability of farmers [17]. The owners of the relocated rural areas are not those executing PAR, but some of the poorest farmers. The most crucial role of PAR is to alleviate the contradiction between peasants and land. The remaining farmers will have more production resources (such as renting the land of the relocated farmers), and may also obtain more employment opportunities (there will be some vacancies for forest-protection work and road-cleaning work after the relocation) and development opportunities. Relocating farmers to live together in a concentrated area will also facilitate the establishment of infrastructure such as medical care, education, training, factories, etc., to a certain extent, and it is easier to accept new knowledge, which will help farmers to achieve sustainability (Table 5).

The GEE platform was used to address the problem of the difficult acquisition of effective Landsat images under cloudy and foggy weather in Guizhou. With its powerful processing capabilities, GEE provided a foundation for the accurate analysis of the temporal and spatial patterns and evolution of EEQ. The RSEI model allows the objective analysis of EEQ. The effects of PAR and climatic factors and other ecological restoration projects on the EEQ changes could be distinguished scientifically using the residual trends. The results could then be used to establish a high and low series of coupling coordination levels between the number of relocated populations and EEQ variables, and to clarify the effective degree of correlation between PAR and EEQ enhancement. Furthermore, based on the ARIMA model, a prediction of future ecological trends concluded that the relocation of impoverished residents has a significant and sustainable driving effect on the promotion of regional EEQ. However, the RSEI model needs to use a water body mask to ensure the normalization accuracy of RSEI [30]. The Southwest Guizhou Autonomous Prefecture belongs to the ecological protection barrier in the upper reaches of the Pearl River, and 1.86% of the water area has important ecological services that unfortunately cannot be effectively reflected [41]. Due to the influence of cloudy weather in Guizhou, it is difficult to collect remote-sensing data in the same period for different years. The base data used in this paper were Landsat data from April to October of the target years. The time span was large, and there may be bias in the measurement of RSEI each year. There are other shortcomings, which provide a basis for the key research directions of the future.

PAR in karst mountains can improve local EEQ to a certain extent, but there have been some controversies. In later studies, researchers further broadened their research scope to investigate the effect of PAR on the regional carbon neutral effect [52] from the direction of carbon neutrality [53].

#### **5. Conclusions**

The residual trends method eliminated the confounding effects of other influencing factors and clarified the contribution of PAR to RSEI growth. The PAR made a contribution to RSEI improvement with a 13.07% increase in the study area during 2015–2020; this RESI was significantly higher than the increase noted during 2000 to 2015. The EEQ of the most ecologically vulnerable areas, such as Qinglong and Pu'an, was most significantly improved after the PAR. However, the EEQ of economically developed areas, such as Xingyi, showed a decreasing trend.

After the implementation of PAR, a large number of rural houses were dismantled and the land reclaimed as green regions. Additionally, many rural construction projects did not advance, contributing to an obvious decrease in the mean value of NDISI (to 17.24%). Furthermore, the core population of farmers moved from cultivated land to urban employment, leaving much farmland abandoned or available for planting trees, contributing to a sharp increase in the NDVI (to 12.28%). The residual trends model predicted that the mean values of RSEI in 2017, 2018, and 2020 were 0.5676, 0.5660, and 0.5701, respectively, under the influence of a multitude of factors, except for the PAR. The observed mean values of RSEI were 0.5942, 0.6374, and 0.6363, respectively. The improvement in RSEI caused by PAR was 0.0266, 0.0715, and 0.0662 in 2017, 2018, and 2020, respectively.

The spatial distribution of the coupling coordination between the relocation population density and RSEI variables showed that there was a significant positive correlation between the increase in RSEI and the relocation population density. The larger the relocated population, the greater and more significant the increase in RSEI in the region. This also led to a higher corresponding level of coupling coordination. As measured by the ARIMA prediction model, the EEQ of the study area, with many people undergoing PAR and in a cluster shape, will continue to evolve for the better. Furthermore, the EEQ of the developed urban areas represented by Xingyi will continue to decrease.

**Author Contributions:** Conceptualization, Z.Z.; data curation of PAR, W.L.; data curation of Landsat, L.W. and X.Z.; writing—original draft preparation, Q.F.; writing—review and editing, Z.Z. and Q.F.; visualization, Q.F.; supervision, C.Z.; project administration, L.Z. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the National Natural Science Foundation of China (41661088), jointly funded by the Program in Guizhou Planning of Philosophy and Social Science (21GZZD39) and the High-level Innovative Talents Training Program in Guizhou Province (2016-5674).

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** The data presented in this study are available on request from the corresponding author. Some of the data are not publicly available due to privacy constraints.

**Conflicts of Interest:** The authors declare no conflict of interest.

**Figure A1.** Distribution of RSEI from 2000 to 2020.

**Figure A2.** Population density of 8 counties (2016).


**Table A1.** Statistical table of changing EEQ image factors.

#### **References**

