Open AccessArticle

Assessing Future Ecological Sustainability Shaped by Shared Socioeconomic Pathways: Insights from an Arid Farming–Pastoral Zone of China

Jiachen Ji

¹,

Sunxun Zhang

²,

Tingting Zhou

Fan Zhang

^1,*

Tianqi Zhao

⁴,

Xinying Wu

⁵,

Yanan Zhuo

¹,

Yue Zhang

¹ and

Naijing Lu

⁴

Jixian National Forest Ecosystem Observation and Research Station, National Ecosystem Research Network of China (CNERN), School of Soil and Water Conservation, Beijing Forestry University, Beijing 100083, China

State Key Laboratory of Simulation and Regulation of Water Cycle in River Basin, China Institute of Water Resources and Hydropower Research, Beijing 100048, China

State Key Laboratory of Environmental Criteria and Risk Assessment, Chinese Research Academy of Environmental Sciences, Beijing 100012, China

⁴

Yinshanbeilu Grassland Eco-Hydrology National Observation and Research Station, China Institute of Water Resources and Hydropower Research, Beijing 100038, China

⁵

National Institute of Natural Hazards, Ministry of Emergency Management of China, Beijing 100085, China

Author to whom correspondence should be addressed.

Remote Sens. 2024, 16(16), 2894; https://doi.org/10.3390/rs16162894

Submission received: 4 July 2024 / Revised: 28 July 2024 / Accepted: 6 August 2024 / Published: 8 August 2024

(This article belongs to the Special Issue Remote Sensing for Agricultural Water Management (RSAWM) (Second Edition))

Download

Browse Figures

Versions Notes

Abstract

Ecological sustainability quantifies the capacity of an ecological system to sustain its health while fulfilling human survival needs and supporting future development. An accurate projection of ecological dynamics for sustainability is crucial for decision-makers to comprehend potential risks. However, the intricate interplay between climate change and human activity has hindered comprehensive assessments of future ecological sustainability, leaving it inadequately investigated thus far. This study aimed to assess future ecological sustainability shaped by the Shared Socioeconomic Pathways (SSPs) using remote sensing data from a typical arid farming–pastoral zone located at the northern foot of Yinshan Mountain (NFYM), Inner Mongolia, China. Five machine learning models were employed to evaluate the relationship between ecological sustainability and its driving factors. The results indicate that (1) overall ecological sustainability initially decreased and then increased during 2003–2022; (2) the Geophysical Fluid Dynamics Laboratory Earth System Model version 4 (GFDL-ESM4) mode and random forest model demonstrated the best performance in climate and ecological sustainability simulations; and (3) the annual change rates of ecological sustainability from 2023 to 2099 are projected to be +0.45%, −0.05%, and −0.46% per year under the SSP1-2.6, SSP2-4.5, and SSP5-8.5 scenarios, respectively, suggesting that stringent environmental policies can effectively enhance ecological sustainability. The proposed framework can assist decision-makers in understanding ecological changes under different SSPs and calls for strategies to enhance ecosystem resilience in the NFYM and similar regions.

Keywords:

ecological sustainability; remote sensing data; machine learning; SSPs

1. Introduction

Climate change represents one of the foremost environmental challenges facing the world today. According to the Intergovernmental Panel on Climate Change (IPCC) Sixth Assessment Report (AR6), the global average surface temperature has risen by approximately 1.1 °C in 2011–2020 above pre-industrial levels. This warming trend is expected to intensify in the coming decades. The consequences of climate change—such as rising temperatures, altered precipitation patterns, and increased extreme weather events—exert substantial pressure on ecosystems, resulting in shifts in species distribution [1], loss of biodiversity [2], and degradation of ecosystem services [3]. Climate change poses significant threats to the health and stability of ecosystems, potentially undermining their ability to provide essential resources and support for future human generations, thus impacting ecological sustainability [4]. To better understand and project these impacts, Shared Socioeconomic Pathways (SSPs) are commonly used to depict future socioeconomic development scenarios, taking into account the effects of diverse social, economic, and environmental policies on climate change [5]. SSPs are widely utilized across various fields, including population expansion analysis [6], urban planning [7], and pollution projection [8]. Understanding future changes in ecological sustainability can guide the development of mitigation and adaptation strategies to achieve sustainability. Therefore, it is crucial to quantify future changes in ecological sustainability influenced by climate and human activities under different SSPs.

Climate change has profoundly impacted ecosystems worldwide through diverse pathways. It threatens tropical rainforest vegetation by pushing temperatures close to the maximum threshold for leaf photosynthesis [9], transforms snowfall into extreme rainfall in high-altitude regions, thereby heightening flood risks and water shortages [10], and diminishes water use efficiency in coastal wetlands, potentially altering these ecosystems [11]. As one of the most fragile ecosystems, the arid farming–pastoral zone is especially susceptible to climate change. In China, this region is markedly affected by environmental changes [12]. Variations in temperature and precipitation have already influenced landscape diversity and ecosystem stability in this region [13]. With the potential intensification of climate change, the pressure on the arid farming–pastoral zone ecosystem is expected to escalate [14]. However, the response of ecosystems to future climate change under various SSPs remains unclear [15], necessitating a quantitative analysis of the ecological impacts driven by changing environmental conditions.

Over-cultivation and overgrazing have subjected arid farming–pastoral zones to severe stress, encompassing water scarcity [16], vegetation degradation [17], and air pollution [18], thereby presenting substantial challenges to ecological sustainability. It is essential and urgent to conduct comprehensive and timely monitoring of dynamic changes in ecological sustainability for these regions. Remote sensing technology can be used to obtain spatial and temporal information on ecological conditions [19]. This provides a means to continuously monitor ecological changes [20]. The impact of vegetation degradation in arid farming–pastoral transition zones on ecological sustainability has garnered significant attention. Previous studies have utilized remote sensing technology to assess changes in vegetation productivity [21], greenness [22], and recovery [23] in these regions. However, few studies have integrated the dimensions of vegetation, hydrology, and atmosphere to construct a comprehensive ecological sustainability assessment index. Therefore, utilizing remote sensing data to identify ecological sustainability is promising, as it is expected to enhance the management of ecologically fragile areas.

Accurately quantifying the impacts and inter-relationships of a changing environment is crucial for projecting future ecological sustainability [24]. The intricate relationships between driving factors and ecological sustainability often cause traditional methods, such as linear regression, to fail in comprehensively capturing data characteristics [25], thus resulting in suboptimal projecting outcomes. In recent years, machine learning models have been proposed and applied in numerous instances, yielding new insights for projecting ecological sustainability, and have demonstrated excellent performance in element simulation [26] and mechanism discovery [27]. These models can process large datasets and discern complex nonlinear relationships [28]. They can be classified into three categories: neural network models (e.g., Backpropagation Neural Network, BPNN; Convolutional Neural Network, CNN; Long Short-Term Memory, LSTM), decision tree models (e.g., random forest, RF), and function approximation models (e.g., Radial Basis Function, RBF) [29]. Numerous practical applications of machine learning pertain to ecosystem studies. For instance, Wu et al. concluded that object-based machine learning models outperform others in classifying natural grassland types in northern China [30], while Li et al. discovered that RF is most effective in exploring the driving factors of soil organic carbon in China [31]. Although these successful applications demonstrate the efficacy of machine learning methods in elucidating complex relationships in ecosystems, the comparative performance of various machine learning models in projecting future ecological sustainability remains unclear. It is imperative to select the appropriate machine learning model based on projective performance to enhance the robustness of ecological sustainability projections.

In summary, this study aimed to assess future ecological sustainability in arid farming–pastoral zones using remote sensing data and SSPs. The methods proposed in this study offer several key advantages: (1) they facilitate the evaluation of ecological sustainability related to environmental issues in arid farming–pastoral zones, (2) they enable the selection of the most accurate CMIP6 climate mode for projecting future climates and the most precise machine learning model for projecting ecological sustainability in these regions, and (3) they allow for the identification of the potential impacts of different SSPs on ecological sustainability. The proposed methods are slated for application in real-world cases within China’s arid farming–pastoral zones and are anticipated to offer valuable insights for bolstering ecosystem sustainability in these zones and similar regions.

2. Materials and Methods

2.1. Study Framework

This study introduces a novel framework, depicted in Figure 1, designed to elucidate the spatiotemporal dynamics of ecological sustainability within farming–pastoral zones under various SSPs. This comprehensive framework is composed of three integral components: an assessment of historical ecological sustainability (2003–2022), the development of future scenarios under the SSPs, and the projection of future ecological sustainability through machine learning models (2023–2099).

2.2. Study Area

The northern foot of Yinshan Mountain (NFYM) (40°72′–45°07′N, 105°18′–116°91′E) is situated in the transitional zone between the Yinshan Mountains and the Mongolian Plateau, representing a typical arid farming–pastoral zone in northern China. The region is bordered by the Otindag Sandy Land in the east, the Yinshan Mountains in the south, the Urad Desert in the west, and the Mongolian Plateau in the north, encompassing an area of approximately 22.72 × 10⁴ km². Administratively, the area spans 18 counties (banners) in the central part of the Inner Mongolia Autonomous Region, China (Table 1).

The NFYM ranges in elevation from 801 to 2356 m, with the terrain gradually sloping downward from south to north. The landform transitions from middle mountains to low and undulating hills, and finally to a wavy plateau (Figure 2). The NFYM falls within the continental monsoon climate zone, with an average annual temperature of 1.5–3.7 °C and a multi-year average precipitation of 163.42–349.01 mm, decreasing from east to west. Precipitation is primarily concentrated from July to September, and the annual evaporation rate is high. Due to climatic constraints, the native vegetation is primarily forest steppe and desert steppe, with the main soil types being chestnut and brown soils.

The southern hilly area of the NFYM is predominantly cultivated, with dryland farming as the main agricultural practice. The central wavy plateau area is a mixed farming–pastoral region, while the northern part is dominated by grasslands, focusing on livestock husbandry [32]. The NFYM is situated in a transitional zone of climate and topography, making the local ecological environment highly sensitive. Over the past century, the region has been increasingly impacted by human activities, leading to issues such as excessive cultivation, overgrazing, deforestation, unsustainable water resource utilization, and air pollution, all of which have affected the sustainability of the local ecosystem [33]. Evaluating ecological sustainability, quantifying the relationships between ecological sustainability and its driving factors, and projecting future changes in ecological sustainability have become critical issues for decision-makers in the NFYM. The objective is to enhance the ecosystem’s resilience in a targeted manner.

2.3. Ecological Sustainability Assessment Based on Remote Sensing Data

Arid farming–pastoral zones face several ecological challenges. Key among these are vegetation degradation, air pollution, and the depletion of water resources. This study evaluates ecological sustainability by selecting indicators from three dimensions: vegetation, atmosphere, and hydrology.

First, the fragile transitional zone vegetation is susceptible to overgrazing, deforestation, and excessive cultivation, leading to vegetation degradation and affecting ecological sustainability. Therefore, the normalized difference vegetation index (NDVI) is selected as the vegetation dimension indicator, as it is widely used to reflect vegetation vigor and nutrient status. The NDVI is calculated using the following formula:

N D V I = \frac{N I R - R}{N I R + R},

(1)

where NDVI represents the normalized difference vegetation index; NIR represents the reflectance in the near-infrared band; and R represents the reflectance in the red band.

Secondly, in the arid farming–pastoral zone, surface dust and the predominant use of primary energy sources contribute to air pollutant emissions. Air pollution is a significant environmental issue in the region, affecting both human and ecosystem health. Consequently, PM2.5 concentration is chosen as the atmospheric dimension indicator.

Meanwhile, the region’s rivers are mostly seasonal and intermittent, with small water volumes, and most usable water resources are groundwater. In recent years, the increasing population and rapidly growing agricultural and industrial water demands have led to the unsustainable exploitation of water resources. The depletion of water resources can impact the ecosystem’s hydrological processes. Thus, terrestrial water storage (TWS) is selected as the hydrological dimension indicator.

In summary, the formula for calculating the ecological sustainability of the arid farming–pastoral zone is as follows:

E S = ω_{N D V I} N D V I + ω_{P M 2.5} P M 2.5 + ω_{T W S} T W S,

(2)

where ES represents ecological sustainability, with values ranging from 0 to 1; PM2.5 represents the normalized PM2.5 concentration, taken as the inverse; and TWS represents the normalized terrestrial water storage. ω_NDVI, ω_PM2.5, and ω_TWS are the respective weights of the indicators, calculated using the analytic hierarchy process [34].

The calculation of annual ecological sustainability from 2003 to 2022 was based on the following remote sensing datasets (Table 2).

The data processing workflow is as follows. The monthly NDVI data were synthesized using the maximum value composite method to obtain the annual NDVI series. The PM2.5 concentration dataset was produced by integrating ground-based observations, atmospheric reanalysis, and emission inventory data. The TWS dataset is based on monthly land water storage data from the Gravity Recovery and Climate Experiment (GRACE) satellites and their follow-up satellites (GRACE-FO), with missing values during the observation period interpolated. The annual average TWS anomaly was then derived from the monthly data (Figure 3).

2.4. Generation of Future Climate Scenarios

2.4.1. Selection of CMIP6 Climate Mode

The future climate mode data utilized in this study are sourced from the Coupled Model Intercomparison Project Phase 6 (CMIP6) organized by the World Climate Research Programme. The climate mode selection is based on the Scenario Mode Intercomparison Project (ScenarioMIP) experiments, which involve different combinations of SSPs. This study selected the SSP1-2.6, SSP2-4.5, and SSP5-8.5 scenarios to simulate climate conditions from 2015 to 2099 (Table 3).

The CMIP6 mode development teams comprise numerous institutions, and the diverse Earth system models developed by these organizations yield varying future climate model data projections. In this study, future temperature and precipitation projection modes from seven institutions under the SSP1-2.6, SSP2-4.5, and SSP5-8.5 scenarios were utilized (Table 4). The selection of climate modes comprehensively considered resolution, scenario relevance, variable availability, and institutional credibility. Mode performance was evaluated, and the mode with the most optimal simulation results was employed in the subsequent analysis.

2.4.2. Accuracy Evaluation Metrics

To assess the accuracy of future climate modes and identify the optimal mode, three statistical metrics, including the mean absolute error (MAE), root mean square error (RMSE), and coefficient of determination (R²), were employed. The MAE is a commonly used metric for assessing simulation performance. It intuitively reflects the average deviation of projected values from observed values. Lower MAE values indicate more accurate mode simulations. The RMSE is a metric used to quantify the magnitude of projection errors. Compared to MAE, RMSE assigns greater weight to larger error values. Smaller RMSE values indicate higher projection accuracy. The R² reflects the model’s ability to explain the variability in the data. The value of R² ranges from 0 to 1. Higher R² values indicate a more accurate model projection. The formulas are represented as follows [38]:

M A E = \frac{1}{n} \sum_{i = 1}^{n} |y_{i} - {\hat{y}}_{i}|,

(3)

R M S E = \sqrt{\frac{1}{n} \sum_{i = 1}^{n} {(y_{i} - {\hat{y}}_{i})}^{2}},

(4)

R^{2} = 1 - \frac{\sum_{i - 1}^{n} {(y_{i} - {\hat{y}}_{i})}^{2}}{\sum_{i - 1}^{n} {(y_{i} - \bar{y})}^{2}},

(5)

where n represents the number of samples;

y_{i}

represents the i-th observed value;

{\hat{y}}_{i}

represents the i-th projected value; and

\bar{y}

represents the average of all measured values.

2.4.3. Climate Scenario Bias Correction

If the mode’s bias remains consistent over time for each grid, the following formula is employed to rectify the discrepancy in the projected temperature of future climate modes:

T_{s i m}^{D M} (x, t) ∶ = T_{e m p} (x, 0) + (T_{s i m}^{r a w} (x, t) - T_{s i m}^{r a w} (x, 0)) = T_{s i m}^{r a w} (x, t) + (T_{e m p} (x, 0) - T_{s i m}^{r a w} (x, 0)),

(6)

where

T_{s i m}^{D M}

represents the processed temperature;

T_{e m p}

represents the observed temperature; and

T_{s i m}^{r a w}

represents the raw temperature before processing.

The following formula can be used to rectify discrepancies in the projected precipitation of future climate modes [39]:

P_{s i m}^{D M} (x, t) ∶ = P_{o b s} (x, 0) \cdot \frac{P_{s i m}^{r a w} (x, t)}{P_{s i m}^{r a w} (x, 0)} = P_{s i m}^{r a w} (x, t) \cdot \frac{P_{o b s} (x, 0)}{P_{s i m}^{r a w} (x, 0)},

(7)

where

P_{s i m}^{D M}

represents the processed precipitation;

P_{o b s}

represents the observed precipitation; and

P_{s i m}^{r a w}

represents the raw precipitation before processing.

2.5. Machine Learning and Statistical Models

Machine learning models are generally considered as effective tools to capture the intricate interactions among multiple variables. This study applies machine learning models to project future ecological sustainability. We employ five machine learning models to construct projection models: LSTM, BPNN, RF, CNN, RBF. The principles of each model are presented in Table 5. Each model possesses unique structures and optimization strategies, making them suitable for various types of data and problems.

In addition, Partial Least Squares (PLS) has been applied as statistical modeling. PLS projects data into a new subspace and generates composite variables, known as score variables, through linear combinations [45]. We evaluated the accuracy of these models in projecting ecological sustainability and ultimately selected the one with the best projective performance for this task.

3. Results Analysis and Discussion

3.1. Spatiotemporal Changes in Ecological Sustainability during 2003–2022

Based on the 20-year average ecological sustainability values calculated at the raster scale, the land in the NFYM was classified into five levels of ecological sustainability—great, good, moderate, low, and poor—using the natural breaks method (Table 6).

The spatial distribution of ecological sustainability in the NFYM is illustrated in Figure 4a, revealing a general decreasing trend from east to west. The southern and eastern counties exhibit great ecological sustainability (i.e., WC, ZL, DL, and the eastern part of AB). The transition area from the east to the central region (i.e., GY, the southern part of DMJ, the southern part of SZW, and the western part of AB) shows moderate to good sustainability. The central and northern regions (i.e., ER, SR, SL, the northern part of SZW, and the northern part of DMJ) and the western part of UM exhibit low ecological sustainability. The western part of UB shows poor ecological sustainability. Overall, areas with great, good, moderate, low, and poor sustainability accounted for 13.37%, 23.96%, 45.65%, 12.10%, and 4.92%, respectively, of the NFYM in 2022.

Ecological sustainability in the NFYM experienced an initial decline followed by a subsequent increase from 2003 to 2022. Changes in average ecological sustainability were calculated for each county (Figure 4b). Specifically, all counties exhibited a fluctuating downward trend in ecological sustainability from 2003 to 2014, with decreases ranging from 11.49% to 29.68%. Fourteen of the eighteen counties experienced decreases of more than 15%. However, ecological sustainability improved from 2014 to 2022. Every county experienced an increase in ecological sustainability, ranging from 11.68% to 39.65%. Fifteen of the eighteen counties experienced increases of more than 15%. Overall, by 2022, ecological sustainability in the NFYM had returned to levels close to those of 2003.

Rapid population growth and relatively low productivity levels have led to predatory exploitation and inefficient use of water resources since 2003. The collection of fuelwood, overgrazing, and excessive cultivation have severely damaged existing forest and grassland ecosystems, resulting in a significant decline in ecological sustainability between 2003 and 2014. Since around 2010, the government has placed greater emphasis on ecological environment construction and protection, advancing projects such as the Three-North Shelterbelt Program and the Beijing-Tianjin Sandstorm Source Control Project. The NFYM has implemented ecological projects such as artificial reforestation and grassland restoration, and the restoration and management of degraded grasslands. Additionally, the government formulated water-saving plans, including regulating water demand based on availability, effectively preventing the overuse of water resources, and improving the water environment. Furthermore, the Sandstorm Source Control Project and stricter factory emission controls have alleviated air pollution issues. Consequently, ecological sustainability recovered between 2014 and 2022. However, average ecological sustainability values in some counties dominated by desert grasslands remained low in 2022 (e.g., ER 0.41, SR 0.39, DMJ 0.40, UB 0.30, UM 0.39). Restoring fragile ecosystems requires continuous effort, and any relaxation can result in setbacks. Therefore, projecting future ecological sustainability changes in the NFYM based on SSPs is important to help address future risks and enhance ecosystem resilience in the NFYM.

3.2. Selection of Optimal Future Climate Mode and Projecting Human Activities

Ecological sustainability is influenced by a multitude of factors pertaining to climate conditions and human activities. Variations in temperature and precipitation directly influence the hydrothermal conditions of ecosystems. Population size is a significant factor impacting ecosystem stress. Agriculture and livestock husbandry are the primary industries upon which people in the arid farming–pastoral zone rely. Fluctuations in population and livestock numbers result in changes in the demand for natural resources, with livestock numbers being a crucial factor impacting the sustainability of grassland ecosystems. Changes in cropland area directly influence land use structure and productivity, thereby impacting ecological sustainability. Accordingly, this study identifies temperature, precipitation, population, cropland area, and livestock numbers as the primary driving factors of ecological sustainability in the arid farming–pastoral zone. The data processing methods for each driving factor are provided in Appendix A.1.

Initially, the simulation accuracy of the CMIP6 mode was compared to observed data. Observed data of temperature and precipitation from seven national meteorological stations in the NFYM, spanning January 2015 to July 2023, were used as the observed values. The data projected by the various CMIP6 modes were extracted based on the latitude and longitude of the meteorological stations and used as the projected values. On this basis, the MAE, RMSE, and R² for the temperature and precipitation data over 104 months were calculate.

All seven modes present relatively good performance in simulating temperature in the NFYM (Figure 5a–c). The CAMS-CSM1-0 (MAE = 1.94 °C, RMSE = 2.49, R² = 0.96), GFDL-ESM4 (MAE = 2.01 °C, RMSE = 2.59, R² = 0.96), and NorESM2-MM (MAE = 2.05 °C, RMSE = 2.71, R² = 0.96) modes demonstrate excellent temperature simulation performance. However, the simulation performance of the modes for precipitation in this region varies greatly, ranked from high to low as follows: GFDL-ESM4, NorESM2-MM, CESM2-WACCM, BCC-CSM2-MR, CAMS-CSM1-0, CIESM, and CAS-ESM2-0 (Figure 5d–f). Compared to other modes, GFDL-ESM4 has the lowest MAE (19.97 mm) and RMSE (34.48), and a significantly higher R² (0.36). Meanwhile, CIESM (MAE = 24.11 mm, RMSE = 39.50, R² = 0.15) and CAS-ESM2-0 (MAE = 23.89 mm, RMSE = 40.29, R² = 0.12) demonstrate poorer simulation performance.

The temperature simulation performance of each mode is superior to their precipitation simulation performance. The reason for this is the significant intra-annual and inter-annual variability observed in the NFYM continental monsoon climate zone, which makes precipitation simulation more challenging than temperature simulation. Additionally, the precipitation process is more complex than temperature changes, involving atmospheric water vapor condensation, cloud formation, and microphysical processes of precipitation. The GFDL-ESM4 mode couples comprehensive physical, chemical, and biogeochemical processes, allowing for a more thorough simulation of climate changes in the earth system. Therefore, it can provide more accurate simulations of future precipitation in the NFYM. The GFDL-ESM4 mode improved temperature simulation metrics (MAE, RMSE, R²) by an average of 15.60%, 13.58%, and 1.47% compared to the other six climate modes. For precipitation simulation, the improvements were 10.86%, 8.85%, and 59.55%. Based on the comparison of simulation performance, the GFDL-ESM4 mode is the only one that shows excellent performance in both temperature (MAE = 2.01 °C, RMSE = 2.59, R² = 0.96) and precipitation (MAE = 19.97 mm, RMSE = 34.48, R² = 0.36) simulations among all modes. Therefore, the GFDL-ESM4 mode is selected as the climate simulation mode.

Based on the GFDL-ESM4 mode, the temperature is projected to increase consistently during 2022–2099 (Figure 6a). The SSP5-8.5 scenario shows the strongest increase, with a rate of 76.00%, followed by the SSP2-4.5 (29.24%) and SSP1-2.6 (5.49%) scenarios. The historical multi-year average temperature is 4.86 °C. From 2023 to 2060, the multi-year average temperatures under the SSP1-2.6, SSP2-4.5, and SSP5-8.5 scenarios increase to 5.51 °C, 6.07 °C, and 6.53 °C, respectively. From 2061 to 2099, temperatures continue to rise, with the multi-year average values under the three scenarios further increasing to 5.82 °C, 6.98 °C, and 8.41 °C, respectively.

The GFDL-ESM4 mode projects that precipitation will also increase under all three future scenarios, with small discrepancies of less than 10 mm in annual averages (Figure 6b). The historical multi-year average precipitation is 244.83 mm. Under the SSP1-2.6, SSP2-4.5, and SSP5-8.5 scenarios, the future multi-year average precipitation is projected to be 330.35 mm, 320.77 mm, and 320.31 mm, respectively, corresponding to increases of 34.93%, 31.02%, and 30.83% relative to historical data.

Future cropland areas show a decreasing trend, with the rate of decline increasing from the SSP1-2.6 to the SSP5-8.5 scenarios (Figure 6c). The cropland area is 1.41 × 10⁴ km² in 2022, decreasing to 1.21 × 10⁴, 1.11 × 10⁴, and 0.97 × 10⁴ km², respectively, under the three scenarios by 2099. The number of livestock also shows a decreasing trend, with the rate of decline increasing from the SSP1-2.6 to the SSP5-8.5 scenarios (Figure 6d). In 2022, the number of livestock is 973.39 × 10⁴ heads. By 2099, the number of livestock under the three scenarios is projected to be 651.43 × 10⁴, 670.73 × 10⁴, and 748.58 × 10⁴ heads, respectively.

The population shows an increasing trend from 2003 to 2007, reaching 2.41 × 10⁶ people in 2007, and a decreasing trend from 2008 to 2022, with the population decreasing to 2.22 × 10⁶ people by 2022 (Figure 6e). Projections indicate that the population will continue to decline in the future. From 2023 to 2050, the rate of decline is similar across all three scenarios. From 2050 to 2099, the SSP2-4.5 scenario shows a slower rate of decline, with the population decreasing to 1.20 × 10⁶ people by 2099. Both the SSP1-2.6 and SSP5-8.5 scenarios show a faster rate of decline, with the population decreasing to 0.90 × 10⁶ and 0.97 × 10⁶ people, respectively, by 2099.

3.3. Selection of the Optimal Machine Learning Model

The dataset of annual ecological sustainability and its driving factors for each county from 2003 to 2022 was divided into training and validation sets with a ratio of 70% and 30%, respectively. Simulations of ecological sustainability were conducted separately for each machine learning model and statistical model PLS (Figure 7). The accuracy of the training set projections, ranked from highest to lowest, is as follows: CNN, RF, BPNN, LSTM, RBF, and PLS. The projection performance of the various models on the validation set is relatively close, with R² values ranging from 0.46 to 0.66. Among these, the RF and RBF models exhibit superior projection performance. The accuracy of the validation set projections, ranked from highest to lowest, is as follows: RBF, RF, CNN, LSTM, PLS, and BPNN.

The CNN (MAE = 0.025, RMSE = 0.031, R² = 0.86) and the RF (MAE = 0.026, RMSE = 0.032, R² = 0.86) demonstrated superior performance in projecting the training set data. The RBF (MAE = 0.040, RMSE = 0.050, R² = 0.66) and the RF (MAE = 0.037, RMSE = 0.045, R² = 0.65) demonstrated the most effective performance in projecting the validation set data (Figure 8). The characteristics of a given dataset can influence the extent to which different models can leverage their respective strengths. Among the six models, only the RF consistently demonstrated superior performance in projecting both the training and validation set data. The RF model improved the MAE, RMSE, and R² on the training set by an average of 30.74%, 30.40%, and 25.43%, respectively, compared to the other five models. On the validation set, these metrics improved by 13.11%, 13.41%, and 13.81%, respectively. RF reduces model variance and improves generalization ability by constructing multiple decision trees and performing voting or averaging. RF uses randomly selected subsets of features when building each decision tree. This approach increases model diversity and improves performance. Compared to CNN, RF can automatically perform feature selection, has a clear structure, and trains quickly. Additionally, RF is more stable and has better generalization ability compared to PLS. Therefore, RF demonstrates superior performance in simulating ecological sustainability. In general, RF stands out with relatively strong performance and is used to project the future ecological sustainability of the NFYM.

3.4. Projected Changes in Ecological Sustainability during 2023–2099

Changes in ecological sustainability for each county in the NFYM from 2023 to 2099 were simulated using the driving factors and the RF model. The annual average value of ecological sustainability in the NFYM was calculated (Figure 9a). Future ecological sustainability shows an upward trend under the SSP1-2.6 scenario, a slight decline under the SSP2-4.5 scenario, and a significant decline under the SSP5-8.5 scenario. Specifically, under the SSP1-2.6 scenario, ecological sustainability maintains a growth trend similar to that of 2015–2022. The average value of ecological sustainability in the NFYM in 2022 was 0.42, which is projected to increase to 0.57 by 2099, representing a growth of 34.92%. The growth trend is projected to slow down relatively after the mid-21st century, with an average change rate of 2.23 × 10⁻³/year from 2023 to 2050 and a slightly slower growth rate of 2.02 × 10⁻³/year from 2051 to 2099. Under the SSP2-4.5 scenario, future ecological sustainability shows a declining trend but remains generally stable. There is a slight decline from 2023 to 2070 at an average change rate of −0.93 × 10⁻³/year. After 2070, despite large interannual fluctuations, the overall trend tends to stabilize. Under the SSP5-8.5 scenario, ecological sustainability shows a declining trend, with a relatively large decline from 2023 to 2060 at an average change rate of −5.68 × 10⁻³/year. From 2061 to 2099, the trend tends to stabilize but continues to decline at an average change rate of −0.50 × 10⁻³/year, reaching 0.27 by 2099, a decrease of 35.80% compared to the current situation.

The spatial distribution maps of ecological sustainability in the NFYM for the years 2030, 2050, 2070, and 2099 are shown in Figure 9b. Overall, ecological sustainability in each county of the NFYM continues to recover under the SSP1-2.6 scenario. Under the SSP2-4.5 scenario, ecological sustainability slightly deteriorates from 2030 to 2070 and then slightly improves by 2099. Ecological sustainability in each county continues to deteriorate under the SSP5-8.5 scenario.

In 2030, as projected by the SSP1-2.6 scenario, the central region shows moderate ecological sustainability, with average values of 0.41 for SL and 0.43 for SR. In contrast, the western region has low ecological sustainability, with values of 0.35 for UM and 0.33 for UB, while other counties exhibit good ecological sustainability. Ecological sustainability in the central and western regions improves year by year. By 2099, the western region (0.47 for UM and 0.44 for UB) reaches a good level of ecological sustainability, while the average ecological sustainability value of the other 16 counties ranges from 0.41 to 0.56, all at a high level.

Under the SSP2-4.5 scenario, areas with moderate or lower ecological sustainability show a gradual expansion trend from 2030 to 2070. Nine counties in the eastern and southern regions have good ecological sustainability in 2030. However, only two counties maintain this level by 2070, while the ecological sustainability of the western region also decreases. Ecological sustainability improves from 2070 to 2099, with the number of counties with good ecological sustainability increasing to five, while most other counties have moderate ecological sustainability.

Ecological sustainability continues to deteriorate under the SSP5-8.5 scenario. In 2030, ecological sustainability shows a trend of gradually decreasing from northeast to southwest. Two counties in the western region have poor ecological sustainability, the central region has low ecological sustainability, and the eastern and southern regions have moderate ecological sustainability. From 2030 to 2099, ecological sustainability in each county continues to decline. By 2099, all 18 counties have poor ecological sustainability, with the lowest being UB (0.22) and the highest being DL (0.29).

The differences among the three future scenarios illustrate the impact of human activities on the future ecological sustainability of the NFYM. Reduced reliance on fossil fuels and sustainable climate policies decrease greenhouse gas emissions and mitigate the negative impact of human activities on the environment under the SSP1-2.6 scenario. Humans adopt a sustainable development policy path, including effective emission reduction actions and ecological protection measures, such as artificial reforestation and grassland restoration. Additionally, climate change leads to a slight increase in temperature and increased precipitation [46]. The warmer and wetter climate benefits vegetation growth. The reduction in human impact and favorable natural climate conditions lead to ecological recovery and increased sustainability. In the SSP2-4.5 scenario, there is moderate reliance on fossil fuels. This results in a slightly higher warming trend. Moderate sustainability policies cause a smaller negative impact on the environment from human use of natural resources [47]. Therefore, ecological sustainability slightly decreases between 2023 and 2070. Furthermore, under the SSP2-4.5 scenario, increased precipitation and reduced human activity pressure between 2070 and 2100 lead to slight ecological recovery [48]. Weaker climate actions and higher greenhouse gas emissions exacerbate environmental degradation under the SSP5-8.5 scenario. The intensity of human agricultural, pastoral, and industrial activities is greater, negatively impacting vegetation, atmosphere, soil, and water cycles, and leading to resource depletion [49]. The focus on short-term economic interests and the lack of effective ecological protection measures results in a vicious cycle. This cycle leads to overexploitation and worsening resource depletion.

Further analysis was conducted on the magnitude of changes in ecological sustainability for each county from 2022 to 2099 (Figure 10). The ecological sustainability of all counties increases under the SSP1-2.6 scenario. The three counties with the smallest increases are DL (rising from 0.58 to 0.64, an increase of 10.77%), UM (rising from 0.39 to 0.47, an increase of 19.55%), and WC (rising from 0.47 to 0.56, an increase of 19.68%). Under SSP2-4.5 scenario, except for UB and XH, the ecological sustainability of the remaining counties declines slightly. The three counties with the largest decreases are DL (falling from 0.58 to 0.50, a decrease of 14.35%), HD (falling from 0.48 to 0.43, a decrease of 10.45%), and WC (falling from 0.47 to 0.43, a decrease of 8.23%). The ecological sustainability of all counties decreases under the SSP5-8.5 scenario. Five counties experience a decrease of more than 40%, namely DL (49.56%), HD (46.62%), ZL (42.97%), WC (41.93%), and ZXB (40.61%).

The results indicate that counties with higher risks of future ecological sustainability decline are all located in the southern part of the NFYM. The southern counties of the NFYM face higher risks of ecological degradation under future SSPs. The southern region currently has higher ecological sustainability, providing more room for decline, which could lead to more severe ecological degradation consequences. Additionally, the southern region has higher economic development levels and population density than the northern region, resulting in greater ecological pressure. The southern ecosystem is mainly composed of forests and farmland, making it more sensitive to human activities. Excessive agricultural development is more likely to lead to land degradation. The DL, HD, and WC counties show the greatest risks of ecological sustainability decline. It is recommended to prioritize strengthening ecological environment protection policies in areas with high ecological degradation risks.

In conclusion, the SSP2-4.5 and SSP5-8.5 scenarios project a heightened risk of future ecological sustainability decline within the NFYM region. However, the anticipated increase in ecological sustainability under the SSP1-2.6 scenario suggests that proactive human attitudes towards the environment can have a significant impact. The NFYM faces substantial risks of future ecological sustainability decline. Decision-makers can address these challenges by implementing enhanced climate adaptation measures, promoting sustainable land management practices, and optimizing the utilization of sustainable water resources.

This study projects the spatiotemporal changes in ecological sustainability within the NFYM from 2023 to 2099. The advantages of the research framework are as follows: (1) the selection of the most accurate CMIP6 mode for simulating future climate, (2) five distinct machine learning models were employed to project ecological sustainability, with the RF model identified as the optimal model based on validation set results, enhancing projection accuracy and reliability, and (3) the spatiotemporal changes in ecological sustainability under the SSP1-2.6, SSP2-4.5, and SSP5-8.5 scenarios were analyzed, with corresponding policy recommendations proposed. Furthermore, the calculation methods for NDVI, TWS, and PM2.5 concentration data utilized in this study are well established. Additionally, data on population, cropland area, and livestock numbers can be readily obtained from yearbooks. This makes the proposed research framework more feasible for practical application in regional ecological sustainability assessments and projections compared to other complex models. The GFDL-ESM4 climate mode selected from the CMIP6, along with the RF model chosen from the machine learning models, have demonstrated high accuracy in their application to the NFYM region. This can serve as a reference when selecting models for other arid farming–pastoral zones.

Nevertheless, this study is subject to several limitations and uncertainties. This study assesses ecological sustainability based on remote sensing technology. In the future, we will explore the development of field measurement methods. Field measurements will be used to evaluate and validate the accuracy of remote sensing assessments. Additionally, future ecological sustainability projections depend on the accuracy of SSPs, which inherently possess uncertainties. These limitations necessitate further validation and adjustment in subsequent research.

In future studies, the ecological changes identified in this study can be used to propose water resource allocation plans or sustainable land use strategies. This approach will furnish decision-makers with additional references and options to enhance ecological sustainability.

4. Conclusions

This paper introduces a methodology for projecting future ecological sustainability under various SSPs. Applied to the NFYM region, it assesses historical spatiotemporal changes in ecological sustainability utilizing remote sensing data. This study verifies and selects the most accurate CMIP6 mode to simulate future climate driving factors in the NFYM. Five machine learning models are employed to project the relationship between driving factors and ecological sustainability, with the most accurate model selected based on validation set results. This study projects future ecological sustainability under SSPs and identifies counties with potential ecological risks. Finally, it provides policy recommendations to enhance ecological sustainability. The main conclusions are summarized as follows:

Ecological sustainability initially decreased and then increased from 2003 to 2022. By 2022, the proportions of areas with great, good, moderate, low, and poor ecological sustainability were 13.37%, 23.96%, 45.65%, 12.10%, and 4.92%, respectively, returning to levels close to those observed in 2003.
The GFDL-ESM4 mode and RF model demonstrate the best performance in climate and ecological sustainability projections, respectively. The GFDL-ESM4 mode improved temperature projection metrics (MAE, RMSE, R²) by an average of 15.60%, 13.58%, and 1.47% compared to the other six climate modes. For precipitation projections, the improvements were 10.86%, 8.85%, and 59.55%. Meanwhile, the RF model enhanced ecological sustainability projection metrics (MAE, RMSE, R²) by 13.11%, 13.41%, and 13.81% on average compared to the other five machine learning and statistical models.
The annual change rates of ecological sustainability from 2023 to 2099 are projected to be +0.45%, −0.05%, and −0.46% per year under the SSP1-2.6, SSP2-4.5, and SSP5-8.5 scenarios, respectively, suggesting that stringent environmental policies can effectively enhance ecological sustainability. The DL, HD, and WC counties exhibit the highest risks of ecological sustainability decline.

The research framework proposed in this study can be utilized to evaluate and project the ecological sustainability of arid farming–pastoral zones. The results of selecting the optimal CMIP6 climate modes and machine learning models can provide valuable references for ecological sustainability projections in similar regions. However, certain research limitations mentioned in the result analysis and discussion section still need to be addressed in future work.

Author Contributions

Conceptualization, F.Z. and J.J.; methodology, F.Z., J.J., S.Z. and X.W.; software, J.J.; validation, F.Z., S.Z., T.Z. (Tingting Zhou), T.Z. (Tianqi Zhao) and X.W.; data curation, J.J., Y.Z. (Yanan Zhuo) and Y.Z. (Yue Zhang); writing—original draft preparation, J.J.; writing—review and editing, F.Z., J.J., S.Z., X.W., T.Z. (Tingting Zhou), T.Z. (Tianqi Zhao) and N.L.; visualization, F.Z. and J.J.; supervision, F.Z.; project administration, F.Z.; funding acquisition, F.Z. All authors have read and agreed to the published version of the manuscript.

Funding

This paper is supported by the National Key Research and Development Program of China (No. 2023YFF1305101) and the Opening Foundation of Yinshanbeilu Grassland Eco-Hydrology National Observation and Research Station (No. YSS202315).

Data Availability Statement

Data will be made available on request.

Acknowledgments

Acknowledgements are extended to the institutions that provided data for this study. Future climate data were obtained from the World Climate Research Programme’s CMIP6 (https://esgf-node.llnl.gov/search/cmip6/, accessed on 25 April 2024). Meteorological station data were sourced from the China Meteorological Data Network (http://data.cma.cn/, accessed on 25 April 2024). Historical cultivated land area data were sourced from the Land Cover Classification Gridded Maps (ESA-CCI) by the European Space Agency (https://doi.org/10.24381/cds.006f2c9a, accessed on 27 April 2024). NDVI data were sourced from the MOD13A3 dataset provided by the Land Processes Distributed Active Archive Center (https://search.earthdata.nasa.gov/search/, accessed on 28 April 2024).

Conflicts of Interest

The authors declare no conflicts of interest.

Appendix A

Appendix A.1. Data Processing for Each Driving Factor

Appendix A.1.1. Historical Data

The details of temperature, precipitation, population, cropland area, and livestock numbers for each county in the NFYM region from 2003 to 2022 are provided in Table A1.

Table A1. Historical data sources for various driving factors.

Category	Data	Source
Natural environment	Temperature	China meteorological data network
Natural environment	Precipitation	China meteorological data network
Human activities	Population	Inner Mongolia statistical yearbook and the statistical yearbooks of each county
	Livestock numbers
	Cropland area	Land cover classification gridded maps (ESA-CCI) from the European Space Agency

Appendix A.1.2. Future Data

Future temperature and precipitation data were obtained from the CMIP6. The CMIP6 mode data processing steps are as follows:

Obtain multidimensional grid data for temperature and precipitation variables from the CMIP6 modes.
Perform multidimensional grid data preprocessing, including coordinate system definition, interpolation, and cropping to the NFYM region.
Convert the units of the precipitation data from its original units to millimeters and the temperature data from its original units to degrees Celsius.
Calculate the annual average temperature and the total annual precipitation for each county.

Figure A1 illustrates the corresponding data processing workflow.

Figure A1. CMIP6 mode data processing workflow.

Future population data are derived from the gridded population dataset under the SSPs. This dataset employs the Population–Development–Environment model and the Cobb–Douglas model to calibrate and validate population model parameters [50]. Future cropland area data are derived from the global 1 km land use dataset. This dataset is generated using the Patch-level Land Use Simulation model and the Global Change Assessment Model [51]. Future cropland area data for each county are subsequently extracted from the land use dataset.

References

Montràs-Janer, T.; Suggitt, A.J.; Fox, R.; Jönsson, M.; Martay, B.; Roy, D.B.; Walker, K.J.; Auffret, A.G.; Sveriges, L. Anthropogenic climate and land-use change drive short- and long-term biodiversity shifts across taxa. Nat. Ecol. Evol. 2024, 8, 739–751. [Google Scholar] [CrossRef] [PubMed]
Portner, H.O.; Scholes, R.J.; Arneth, A.; Barnes, D.; Burrows, M.T.; Diamond, S.E.; Duarte, C.M.; Kiessling, W.; Leadley, P.; Managi, S.; et al. Overcoming the coupled climate and biodiversity crises and their societal impacts. Science 2023, 380, eabl4881. [Google Scholar] [CrossRef] [PubMed]
Willcock, S.; Hooftman, D.; Neugarten, R.A.; Chaplin-Kramer, R.; Barredo, J.I.; Hickler, T.; Kindermann, G.; Lewis, A.R.; Lindeskog, M.; Martinez-Lopez, J.; et al. Model ensembles of ecosystem services fill global certainty and capacity gaps. Sci. Adv. 2023, 9, eadf5492. [Google Scholar] [CrossRef] [PubMed]
Conradi, T.; Eggli, U.; Kreft, H.; Schweiger, A.H.; Weigelt, P.; Higgins, S.I. Reassessment of the risks of climate change for terrestrial ecosystems. Nat. Ecol. Evol. 2024, 8, 888–900. [Google Scholar] [CrossRef] [PubMed]
Michalek, A.T.; Villarini, G.; Husic, A. Climate change projected to impact structural hillslope connectivity at the global scale. Nat. Commun. 2023, 14, 6788. [Google Scholar] [CrossRef] [PubMed]
Yang, D.; Luan, W.; Zhang, X. Projecting spatial interactions between global population and land use changes in the 21st century. NPJ Urban Sustain. 2023, 3, 53. [Google Scholar] [CrossRef]
Ren, Q.; He, C.; Huang, Q.; Shi, P.; Zhang, D.; Güneralp, B. Impacts of urban expansion on natural habitats in global drylands. Nat. Sustain. 2022, 5, 869–878. [Google Scholar] [CrossRef]
Jones, E.R.; Bierkens, M.F.; van Puijenbroek, P.J.; van Beek, L.R.P.; Wanders, N.; Sutanudjaja, E.H.; van Vliet, M.T. Sub-Saharan Africa will increasingly become the dominant hotspot of surface water pollution. Nat. Water 2023, 1, 602–613. [Google Scholar] [CrossRef]
Doughty, C.E.; Keany, J.M.; Wiebe, B.C.; Rey-Sanchez, C.; Carter, K.R.; Middleby, K.B.; Cheesman, A.W.; Goulden, M.L.; Da Rocha, H.R.; Miller, S.D.; et al. Tropical forests are approaching critical temperature thresholds. Nature 2023, 621, 105–111. [Google Scholar] [CrossRef]
Ombadi, M.; Risser, M.D.; Rhoades, A.M.; Varadharajan, C. A warming-induced reduction in snow fraction amplifies rainfall extremes. Nature 2023, 619, 305–310. [Google Scholar] [CrossRef]
Wei, S.; Chu, X.; Sun, B.; Yuan, W.; Song, W.; Zhao, M.; Wang, X.; Li, P.; Han, G. Climate warming negatively affects plant water-use efficiency in a seasonal hydroperiod wetland. Water Res. 2023, 242, 120246. [Google Scholar] [CrossRef]
Li, M.; Wu, P.; Ma, Z.; Liu, J. Climate change reshapes bioclimatic environments in China’s dry–wet transition zones. J. Hydrol. 2024, 634, 131122. [Google Scholar] [CrossRef]
Li, Z.; Ma, T.; Cai, Y.; Fei, T.; Zhai, C.; Qi, W.; Dong, S.; Gao, J.; Wang, X.; Wang, S. Stable or unstable? Landscape diversity and ecosystem stability across scales in the forest–grassland ecotone in northern China. Landsc. Ecol. 2023, 38, 3889–3902. [Google Scholar] [CrossRef]
Li, C.; Fu, B.; Wang, S.; Stringer, L.C.; Zhou, W.; Ren, Z.; Hu, M.; Zhang, Y.; Rodriguez-Caballero, E.; Weber, B.; et al. Climate-driven ecological thresholds in China’s drylands modulated by grazing. Nat. Sustain. 2023, 6, 1363–1372. [Google Scholar] [CrossRef]
Li, H.; Li, Z.; Chen, Y.; Liu, Y.; Hu, Y.; Sun, F.; Kayumba, P. Projected Meteorological Drought over Asian Drylands under Different CMIP6 Scenarios. Remote Sens. 2021, 13, 4409. [Google Scholar] [CrossRef]
Jones, E.R.; Bierkens, M.F.P.; van Vliet, M.T.H. Current and future global water scarcity intensifies when accounting for surface water quality. Nat. Clim. Chang. 2024, 14, 629–635. [Google Scholar] [CrossRef]
Cai, Q.; Chen, W.; Chen, S.; Xie, S.; Piao, J.; Ma, T.; Lan, X. Recent pronounced warming on the Mongolian Plateau boosted by internal climate variability. Nat. Geosci. 2024, 17, 181–188. [Google Scholar] [CrossRef]
Wen, Z.; Ma, X.; Xu, W.; Si, R.; Liu, L.; Ma, M.; Zhao, Y.; Tang, A.; Zhang, Y.; Wang, K.; et al. Combined short-term and long-term emission controls improve air quality sustainably in China. Nat. Commun. 2024, 15, 5169. [Google Scholar] [CrossRef] [PubMed]
Pascual, U.; Balvanera, P.; Anderson, C.B.; Chaplin-Kramer, R.; Christie, M.; Gonzalez-Jimenez, D.; Martin, A.; Raymond, C.M.; Termansen, M.; Vatn, A.; et al. Diverse values of nature for sustainability. Nature 2023, 620, 813–823. [Google Scholar] [CrossRef]
Wang, H.; Li, Z.; Chen, L.; Li, D.; Jiang, Y. The Disparity in Normalized Difference Vegetarian Index Response to Climate Warming and Humidification in the Tibetan Plateau before and after 1998. Remote Sens. 2024, 16, 2121. [Google Scholar] [CrossRef]
Zhang, X.; Evans, J.P.; Burrell, A.L. Less than 4% of dryland areas are projected to desertify despite increased aridity under climate change. Commun. Earth Environ. 2024, 5, 300–309. [Google Scholar] [CrossRef]
Li, Z.; Sun, F.; Wang, H.; Wang, T.; Feng, Y. Detecting the interactions between vegetation greenness and drought globally. Atmos. Res. 2024, 304, 107409. [Google Scholar] [CrossRef]
Zhang, Y.; Feng, X.; Zhou, C.; Sun, C.; Leng, X.; Fu, B. Aridity threshold of ecological restoration mitigated atmospheric drought via land-atmosphere coupling in drylands. Commun. Earth Environ. 2024, 5, 381. [Google Scholar] [CrossRef]
Song, Y.; Penuelas, J.; Ciais, P.; Wang, S.; Zhang, Y.; Gentine, P.; McCabe, M.F.; Wang, L.; Li, X.; Li, F.; et al. Recent Water Constraints Mediate the Dominance of Climate and Atmospheric CO2 on Vegetation Growth Across China. Earth’s Future 2024, 12, e2023EF004395. [Google Scholar] [CrossRef]
Liu, L.; Zhou, W.; Guan, K.; Peng, B.; Xu, S.; Tang, J.; Zhu, Q.; Till, J.; Jia, X.; Jiang, C.; et al. Knowledge-guided machine learning can improve carbon cycle quantification in agroecosystems. Nat. Commun. 2024, 15, 357. [Google Scholar] [CrossRef]
Yu, Q.; Jiang, L.; Wang, Y.; Liu, J. Enhancing streamflow simulation using hybridized machine learning models in a semi-arid basin of the Chinese loess Plateau. J. Hydrol. 2023, 617, 129115. [Google Scholar] [CrossRef]
Feng, J.; Li, J.; Xu, C.Y.; Wang, Z.; Zhang, Z.; Wu, X.; Lai, C.; Zeng, Z.; Tong, H.; Jiang, S. Viewing Soil Moisture Flash Drought Onset Mechanism and Their Changes Through XAI Lens: A Case Study in Eastern China. Water Resour. Res. 2024, 60, e2023WR036297. [Google Scholar] [CrossRef]
Zhang, Q.; Miao, C.; Gou, J.; Zheng, H. Spatiotemporal characteristics and forecasting of short-term meteorological drought in China. J. Hydrol. 2023, 624, 129924. [Google Scholar] [CrossRef]
Greener, J.G.; Kandathil, S.M.; Moffat, L.; Jones, D.T. A guide to machine learning for biologists. Nat. Rev. Mol. Cell Biol. 2022, 23, 40–55. [Google Scholar] [CrossRef] [PubMed]
Wu, N.; Crusiol, L.G.T.; Liu, G.; Wuyun, D.; Han, G. Comparing Machine Learning Algorithms for Pixel/Object-Based Classifications of Semi-Arid Grassland in Northern China Using Multisource Medium Resolution Imageries. Remote Sens. 2023, 15, 750. [Google Scholar] [CrossRef]
Li, H.; Wu, Y.; Liu, S.; Xiao, J.; Zhao, W.; Chen, J.; Alexandrov, G.; Cao, Y. Decipher soil organic carbon dynamics and driving forces across China using machine learning. Glob. Chang. Biol. 2022, 28, 3394–3410. [Google Scholar] [CrossRef] [PubMed]
Zhang, G.; Chen, X.; Zhou, Y.; Zhao, H.; Jin, Y.; Luo, Y.; Chen, S.; Wu, X.; Pan, Z.; An, P. Land use/cover changes and subsequent water budget imbalance exacerbate soil aridification in the farming-pastoral ecotone of northern China. J. Hydrol. 2023, 624, 129939. [Google Scholar] [CrossRef]
Pei, H.; Liu, M.; Shen, Y.; Xu, K.; Zhang, H.; Li, Y.; Luo, J. Quantifying impacts of climate dynamics and land-use changes on water yield service in the agro-pastoral ecotone of northern China. Sci. Total Environ. 2022, 809, 151153. [Google Scholar] [CrossRef] [PubMed]
Saaty, T.L. Axiomatic foundation of the analytic hierarchy process. Manag. Sci. 1986, 32, 841–855. [Google Scholar] [CrossRef]
Wei, J.; Li, Z.; Lyapustin, A.; Sun, L.; Peng, Y.; Xue, W.; Su, T.; Cribb, M. Reconstructing 1-km-resolution high-quality PM2.5 data records from 2000 to 2018 in China: Spatiotemporal variations and policy implications. Remote Sens. Environ. 2021, 252, 112136. [Google Scholar] [CrossRef]
Chen, W.; Xiong, Y.; Zhong, M.; Yang, Z.; Shum, C.K.; Li, W.; Liang, L.; Li, Q. Twenty-Year Spatiotemporal Variations of TWS over Mainland China Observed by GRACE and GRACE Follow-On Satellites. Atmosphere 2023, 14, 1717. [Google Scholar] [CrossRef]
Zhou, J.; Lu, H.; Yang, K.; Jiang, R.; Yang, Y.; Wang, W.; Zhang, X. Projection of China’s future runoff based on the CMIP6 mid-high warming scenarios. Sci. China Earth Sci. 2023, 66, 528–546. [Google Scholar] [CrossRef]
Liu, Z.; Wang, J.; Ding, J.; Xie, X. Analysis of spatial–temporal evolution trends and influential factors of desert-oasis thermal environment in typical arid zone: The case of Turpan–Hami region. Ecol. Indic. 2023, 154, 110747. [Google Scholar] [CrossRef]
Wu, Y.; Miao, C.; Fan, X.; Gou, J.; Zhang, Q.; Zheng, H. Quantifying the Uncertainty Sources of Future Climate Projections and Narrowing Uncertainties with Bias Correction Techniques. Earth’s Future 2022, 10, e2022EF002963. [Google Scholar] [CrossRef]
Ishida, K.; Ercan, A.; Nagasato, T.; Kiyama, M.; Amagasaki, M. Use of one-dimensional CNN for input data size reduction in LSTM for improved computational efficiency and accuracy in hourly rainfall-runoff modeling. J. Environ. Manag. 2024, 359, 120931. [Google Scholar] [CrossRef]
Jiao, L.; Wang, L.; Lu, H.; Fan, Y.; Zhang, Y.; Wu, Y. An assessment model for urban resilience based on the pressure-state-response framework and BP-GA neural network. Urban Clim. 2023, 49, 101543. [Google Scholar] [CrossRef]
Zhang, X.; Shen, H.; Huang, T.; Wu, Y.; Guo, B.; Liu, Z.; Luo, H.; Tang, J.; Zhou, H.; Wang, L.; et al. Improved random forest algorithms for increasing the accuracy of forest aboveground biomass estimation using Sentinel-2 imagery. Ecol. Indic. 2024, 159, 111752. [Google Scholar] [CrossRef]
Elbeltagi, A.; Srivastava, A.; Ehsan, M.; Sharma, G.; Yu, J.; Khadke, L.; Gautam, V.K.; Awad, A.; Jinsong, D. Advanced stacked integration method for forecasting long-term drought severity: CNN with machine learning models. J. Hydrol. Reg. Stud. 2024, 53, 101759. [Google Scholar] [CrossRef]
Granata, F.; Di Nunno, F. Neuroforecasting of daily streamflows in the UK for short- and medium-term horizons: A novel insight. J. Hydrol. 2023, 624, 129888. [Google Scholar] [CrossRef]
Sanches Fernandes, L.F.; Fernandes, A.C.P.; Ferreira, A.R.L.; Cortes, R.M.V.; Pacheco, F.A.L. A partial least squares–Path modeling analysis for the understanding of biodiversity loss in rural and urban watersheds in Portugal. Sci. Total Environ. 2018, 626, 1069–1085. [Google Scholar] [CrossRef]
Kotz, M.; Levermann, A.; Wenz, L. The effect of rainfall changes on economic production. Nature 2022, 601, 223–227. [Google Scholar] [CrossRef]
Chen, G.; Li, X.; Liu, X.; Chen, Y.; Liang, X.; Leng, J.; Xu, X.; Liao, W.; Qiu, Y.A.; Wu, Q.; et al. Global projections of future urban land expansion under shared socioeconomic pathways. Nat. Commun. 2020, 11, 537. [Google Scholar] [CrossRef] [PubMed]
Sutradhar, U.; Spearing, L.; Derrible, S. Depopulation and associated challenges for US cities by 2100. Nat. Cities 2024, 1, 51–61. [Google Scholar] [CrossRef]
Chakraborty, T.C.; Qian, Y. Urbanization exacerbates continental- to regional-scale warming. One Earth 2024, 7, 1–15. [Google Scholar] [CrossRef]
Huang, J.; Qin, D.; Jiang, T.; Wang, Y.; Feng, Z.; Zhai, J.; Cao, L.; Chao, Q.; Xu, X.; Wang, G.; et al. Effect of Fertility Policy Changes on the Population Structure and Economy of China: From the Perspective of the Shared Socioeconomic Pathways. Earth’s Future 2019, 7, 250–265. [Google Scholar] [CrossRef]
Zhang, T.; Cheng, C.; Wu, X. Mapping the spatial heterogeneity of global land use and land cover from 2020 to 2100 at a 1 km resolution. Sci. Data 2023, 10, 748. [Google Scholar] [CrossRef] [PubMed]

Figure 1. Study framework.

Figure 2. Geolocation of the study area.

Figure 3. Remote sensing data of NDVI (a), PM2.5 concentration (b), and TWS anomaly (c) in 2022.

Figure 4. Ecological sustainability assessment of the NFYM. (a) The spatial distribution of ecological sustainability at the grid scale for the years 2003, 2007, 2011, 2015, 2019, and 2022; (b) changes in the average values of ecological sustainability for each county from 2003 to 2022. The definitions corresponding to the abbreviations of the county names can be found in Table 1.

Figure 5. Comparison of the simulation performance for temperature (a–c) and precipitation (d–f) of different CMIP6 modes based on the MAE, RMSE, and R².

Figure 6. Changes in temperature (a), precipitation (b), cropland area (c), livestock number (d), and population (e) during 2003–2099 under different scenarios.

Figure 7. Effectiveness of simulation of ecological sustainability by six machine learning models. (a) Simulation effect of the training set; (b) simulation effect of the validation set.

Figure 8. Comparison of projection performance of different machine learning models for training (a–c) and validation (d–f) sets based on MAE, RMSE, and R².

Figure 9. Changes in future ecological sustainability of the NFYM under the three SSPs. (a) Changes in the average ecological sustainability values of the NFYM in the historical period (2003–2022) and future (2023–2099); (b) spatial distribution of ecological sustainability at the county scale for each typical year (i.e., 2022, 2030, 2050, 2070, 2099).

Figure 10. Changes in ecological sustainability across different counties between 2023 and 2099.

Table 1. Names and abbreviations of the counties.

County Name	Abbreviation
Duolun County	DL
Wuchuan County	WC
Qahar Middle Banner	QM
Zhenglan County	ZL
Huade County	HD
Qahar Back Banner	QB
Shangdu County	SD
Guyang County	GY
Zhengxiangbai County	ZX
Abag Banner	AB
Xianghuang County	XH
Darhan Mumingan Joint Banner	DMJ
Siziwang Banner	SZW
Sonit Left Banner	SL
Sonit Right Banner	SR
Urat Middle Banner	UM
Erenhot	ER
Urat Back Banner	UB

Table 2. Remote sensing data sources.

Data	Observation Sensor	Data Sources
NDVI	Satellite-based Moderate Resolution Imaging Spectroradiometer	NASA MOD13A3
PM2.5 concentration	Satellite-based Moderate Resolution Imaging Spectroradiometer	China High Air Pollutants Dataset [35]
Terrestrial water storage	K-band Ranging System and Accelerometers	China Terrestrial Water Storage Dataset [36]

Table 3. CMIP6 climate scenarios.

Scenario	Scenario Description
SSP1-2.6	It is an upgrade of the RCP2.6 scenario (relatively low greenhouse gas emissions) based on the SSP1 (sustainable development) pathway. The radiative forcing in this scenario reaches 2.6 W/m² by 2099 [37].
SSP2-4.5	It is an upgrade of the RCP4.5 scenario (medium greenhouse gas emissions) based on the SSP2 (middle-of-the-road) pathway. The radiative forcing in this scenario reaches 4.5 W/m² by 2099 [37].
SSP5-8.5	It is an upgrade of the RCP8.5 scenario (high greenhouse gas emissions) based on the SSP5 (conventional development) pathway. The radiative forcing in this scenario reaches 8.5 W/m² by 2099 [37].

Table 4. CMIP6 modes.

Mode Name Abbreviation	Mode Name	Institution	Institution Abbreviation	Country
BCC-CSM2-MR	Beijing Climate Center Climate System Mode version 2 Medium Resolution	National Climate Center	BCC	China
CAMS-CSM1-0	Chinese Academy of Meteorological Sciences Climate System Mode version 1.0	Chinese Academy of Meteorological Sciences	CAMS	China
CAS-ESM2-0	Chinese Academy of Sciences Earth System Mode version 2.0	Institute of Atmospheric Physics, Chinese Academy of Sciences	CAS	China
CIESM	Community Integrated Earth System Mode	Tsinghua University	THU	China
NorESM2-MM	Norwegian Earth System Mode version 2 Medium-Resolution	Norwegian Climate Center	NCC	Norway
CESM2-WACCM	Community Earth System Mode Version 2 with the Whole Atmosphere Community Climate Mode	National Center for Atmospheric Research	NCAR	United States
GFDL-ESM4	Geophysical Fluid Dynamics Laboratory Earth System Mode version 4	National Oceanic and Atmospheric Administration	NOAA	United States

Table 5. Descriptions of machine learning models.

Model Name	Description	Reference
Long Short-Term Memory (LSTM)	The fundamental architecture of an LSTM network comprises an input gate, a forget gate, an output gate, and a cell state. These gates regulate the flow of information at each time step, updating the cell state and generating the output.	[40]
Backpropagation Neural Network (BPNN)	The basic structure of a BPNN includes an input layer, one or more hidden layers, and an output layer. Each layer consists of several neurons, and the layers are connected by weights. The training process mainly includes two phases: forward propagation and backpropagation. Through continuous adjustment of weights and biases, the network minimizes the output error.	[41]
Random Forests (RF)	RF consists of multiple decision trees, each generated by randomly sampling the training data and selecting random features during training, resulting in relatively independent models. The final projection is based on the aggregated results of all trees.	[42]
Convolutional Neural Networks (CNN)	The fundamental architecture of a CNN includes convolutional layers, pooling layers, and fully connected layers. Each layer performs distinct operations on the input data, progressively extracting higher-level features. The convolutional layer is the core component of a CNN, responsible for extracting local features from the data through convolution operations.	[43]
Radial Basis Function (RBF)	The fundamental architecture of an RBF network includes an input layer, a hidden layer, and an output layer. The hidden layer performs nonlinear transformations, while the output layer executes linear combinations of the transformed data.	[44]

Table 6. Criteria for classifying ecological sustainability.

Ecological Sustainability Scale	Range of Values
Great sustainability	0.5–1
Good sustainability	0.43–0.5
Moderate sustainability	0.37–0.43
Low sustainability	0.3–0.37
Poor sustainability	0–0.3

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

© 2024 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).

Share and Cite

MDPI and ACS Style

Ji, J.; Zhang, S.; Zhou, T.; Zhang, F.; Zhao, T.; Wu, X.; Zhuo, Y.; Zhang, Y.; Lu, N. Assessing Future Ecological Sustainability Shaped by Shared Socioeconomic Pathways: Insights from an Arid Farming–Pastoral Zone of China. Remote Sens. 2024, 16, 2894. https://doi.org/10.3390/rs16162894

AMA Style

Ji J, Zhang S, Zhou T, Zhang F, Zhao T, Wu X, Zhuo Y, Zhang Y, Lu N. Assessing Future Ecological Sustainability Shaped by Shared Socioeconomic Pathways: Insights from an Arid Farming–Pastoral Zone of China. Remote Sensing. 2024; 16(16):2894. https://doi.org/10.3390/rs16162894

Chicago/Turabian Style

Ji, Jiachen, Sunxun Zhang, Tingting Zhou, Fan Zhang, Tianqi Zhao, Xinying Wu, Yanan Zhuo, Yue Zhang, and Naijing Lu. 2024. "Assessing Future Ecological Sustainability Shaped by Shared Socioeconomic Pathways: Insights from an Arid Farming–Pastoral Zone of China" Remote Sensing 16, no. 16: 2894. https://doi.org/10.3390/rs16162894

APA Style

Ji, J., Zhang, S., Zhou, T., Zhang, F., Zhao, T., Wu, X., Zhuo, Y., Zhang, Y., & Lu, N. (2024). Assessing Future Ecological Sustainability Shaped by Shared Socioeconomic Pathways: Insights from an Arid Farming–Pastoral Zone of China. Remote Sensing, 16(16), 2894. https://doi.org/10.3390/rs16162894

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Menu