Ecological Risks Arising in the Regional Water Resources in Inner Mongolia Due to a Large-Scale Afforestation Project

: In recent years, a large-scale afforestation campaign has been implemented in Inner Mongolia, China, to control desertiﬁcation and soil erosion. However, the water consumption associated with large-scale afforestation signiﬁcantly impacts the water resources in Inner Mongolia, resulting in a major ecological risk. This study aimed to evaluate the ecological risk of water resources caused by afforestation in the region. In this study, using land cover data, normalized difference vegetation index ( NDVI ) data, and meteorological data, we performed trend analysis and used the water balance equation and water security index (WSI) to analyze the ecological risks of water resources caused by afforestation in Inner Mongolia from 2000 to 2020. The results show that (1) the afforestation area in Inner Mongolia was 5.37 × 10 4 km 2 in 2000–2020; (2) afforestation in arid and semi-arid areas led to a reduction in water resources; (3) afforestation reduced water resources in the study area by 62 million cubic meters (MCM) per year; and (4) ~76% of afforestation regions faced ecological risks related to water resources. This study provides scientiﬁc suggestions for the sustainable development of regional water resources and afforestation.


Introduction
To protect the fragile ecosystem and economy, combat desertification, and control dust storms, the Chinese central and local governments have implemented a series of ecological engineering projects in China [1], such as the Slope Land Conversion Project, China's Natural Forest Protection Project, River Shelterbelt Project, and the Returning Farmland to Forest and Grassland [2].Ecological engineering slowed desertification and its expansion in China, significantly contributing to the world's "greening" trend [3].Chen et al. [3] indicated that the global green leaf area has increased by 5% since the early 2000s.China and India contributed 25% and 6.8%, respectively, to the increase in greening on land.
However, this significant land cover change resulting from afforestation very strongly affected the ecological environment, especially the water resources [4][5][6][7].It was confirmed that forests could increase regional evapotranspiration and reduce total runoff discharge compared to the absence of forests under the same precipitation conditions.Some studies pointed out that 10-40% of annual precipitation was lost by the canopy interception [8].It was shown that the total canopy interception was 76.6 mm in Shaanxi Province, northwestern China, accounting for 18.6% of the gross precipitation.Moreover, in forests, shallow roots extract soil water provided by precipitation, while groundwater is consumed through deep roots during the dry period [9].Karimov et al. [10] found that shallow groundwater contribution to plant transpiration exceeds 60% in the upstream area of the Syrdarya River, in Central Asia.In the Ejina Basin, China, Populus euphratica obtains 53% of its water from groundwater [11].In the dry season, the groundwater uptake accounts for 73.2% [12].Due to the effects of canopy interception and transpiration from soil and aquifer layers, the effective precipitation and groundwater recharge in forests are much lower than in other areas [13].Keese et al. [14] carried out an unsaturated flow modeling study, and through simulations, they found that forests significantly decreased groundwater recharge by factors of 2 to 30 relative to the recharge in non-vegetated areas [14].Therefore, it was concluded that large-scale tree planting in China could lead to regional water resource shortages.Xiao et al. [15] indicated that since 2000, the water consumption of plantations has increased to 40.42 billion m 3 in southwestern China.This amount accounts for 10.69% of the water resources for the entire year.Cao et al. [16] used seven models and estimated that afforestation will increase net water consumption by 559-2354 m 3 /ha annually compared with the potential amount of water natural vegetation would consume.Water resources are an important basis for maintaining the sustainability of ecosystem development in arid and semi-arid areas [17,18].Due to climate change and water resource development and utilization, the spatial distribution of water resources changes [19], leading to the risk of water supply-demand in regional ecosystems [20].
Since 2000, several large-scale ecological restoration projects have been implemented in Inner Mongolia, leading to a rapid increase in forest area.Inner Mongolia is situated in an arid and semi-arid region with a fragile environment, uneven spatial distribution of water resources, and high sensitivity to changes in water resources.It is a perfect natural laboratory for studying the impact of large-scale afforestation on regional water resources.Although studies have already acknowledged the influence of ecological restoration projects on the water balance of regional ecosystems [16,21,22], further assessment regarding their ecological risks remains lacking.Therefore, this paper specifically focuses on evaluating the ecological risks in regional water resources that are caused by large-scale afforestation.The primary objectives of this study are (1) identifying afforestation areas in Inner Mongolia from 2000 to 2020 and (2) assessing changes in water resources and associated ecological risks resulting from afforestation activities.This research aims to better regulate ecological engineering policies for the benefit of humankind while providing reliable insights into sustainable development concerning both regional water resources and afforestation.

Study Area
Inner Mongolia is located in northern China, covering an area of 1.18 million km 2 , or 12.3% of the national terrestrial area [23].It extends from 97-126 • E to 37-53 • N (Figure 1).The mean temperature is approximately 0-8 • C. The average annual precipitation is only 50-450 mm and gradually increases from west to east; approximately 75% of the rainfall is concentrated between July and September [24].The main vegetation types are forested prairies, temperate prairies, desert steppes, and deserts intergrading from east to west (Figure 1).Inner Mongolia contains the largest grassland and natural pasture in China.According to the results of the eighth forest resources inventory, the grassland area is 88 million hectares, accounting for ~74% of the total land area of Inner Mongolia and ~22% of the grassland area of the entire country.The forestland area is 45 million hectares, accounting for ~22% of the total land area of Inner Mongolia and ~14% of the forest area of the entire country.The ecology and environment of Inner Mongolia are fragile and highly vulnerable to water resource shortages.Ecological restoration projects have been implemented to address land desertification and mitigate the degradation of the environment, such as Returning Farmland to Forest, and Beijing-Tianjin Sand Control.
accounting for ~22% of the total land area of Inner Mongolia and ~14% of the forest ar of the entire country.The ecology and environment of Inner Mongolia are fragile an highly vulnerable to water resource shortages.Ecological restoration projects have bee implemented to address land desertification and mitigate the degradation of the enviro ment, such as Returning Farmland to Forest, and Beijing-Tianjin Sand Control.According to the Inner Mongolia Water Resource Bulletin, the area's total water r sources in 2020 were 50.39 billion cubic meters (BCM), surface water resources were 35.BCM, and groundwater resources were 24.39 BCM.However, water resources are main concentrated in the northeastern part of the study region, Hulunbuir City.In contrast, th per capita water resources in the northwestern part are only one fifth of the world averag and it is one of the most water-scarce regions in China.Water supply in Inner Mongol includes surface water and groundwater, with a total water consumption rate of abo 19.4 BCM.Surface water and groundwater consumption accounted for 54.4% and 41.9 of the total water consumption, respectively.Agricultural irrigation, industrial use, ec logical use, and human consumption accounted for 63.9%, 6.9%, 15.1%, and 6.0% of th total water consumption, respectively.Water consumption for agricultural irrigatio showed an obvious decreasing trend from 2000 to 2020, decreasing at a rate of 0.20 BC per year (Figure 2).According to the Inner Mongolia Water Resource Bulletin, the area's total water resources in 2020 were 50.39 billion cubic meters (BCM), surface water resources were 35.42 BCM, and groundwater resources were 24.39 BCM.However, water resources are mainly concentrated in the northeastern part of the study region, Hulunbuir City.In contrast, the per capita water resources in the northwestern part are only one fifth of the world average, and it is one of the most water-scarce regions in China.Water supply in Inner Mongolia includes surface water and groundwater, with a total water consumption rate of about 19.4 BCM.Surface water and groundwater consumption accounted for 54.4% and 41.9% of the total water consumption, respectively.Agricultural irrigation, industrial use, ecological use, and human consumption accounted for 63.9%, 6.9%, 15.1%, and 6.0% of the total water consumption, respectively.Water consumption for agricultural irrigation showed an obvious decreasing trend from 2000 to 2020, decreasing at a rate of 0.20 BCM per year (Figure 2).

Data Collection and Processing
The meteorological data mainly included monthly precipitation and temperature from 2000 to 2020.The meteorological data from 37 stations were collected from the Chi-

Data Collection and Processing
The meteorological data mainly included monthly precipitation and temperature from 2000 to 2020.The meteorological data from 37 stations were collected from the Chinese National Meteorological Information Center (http://data.cma.cn,accessed on 10 May 2023).The precipitation and temperature data were used for spatial interpolation using the ArcGIS 10.2 (https://www.esri.com,accessed on 18 July 2023) "Spatial Analyst Tools", and resampling was performed into 1000 m raster data.Land use/land cover remote sensing monitoring data in China from 2000 to 2020 were collected from the Data Center for Resources and Environmental Sciences, Chinese Academy of Sciences (RESDC, http://www.resdc.cn,accessed on 11 July 2023).The NDVI dataset originated from the Global Inventory Modeling and Mapping Studies (GIMMS) covering the period from 2000 to 2020, and data were collected from the Ecological Forecasting Lab of NASA's Ames Research Center (https://ecocast.arc.nasa.gov,accessed on 12 April 2023).The annual evapotranspiration (ET) data (MOD16 data product) from 2000 to 2020 were obtained from the Land Processes Distributed Active Archive Center (https://lpdaac.usgs.gov/,accessed on 10 February 2023).The MOD16 evapotranspiration dataset is based on the logic of the Penman-Monteith equation, which includes inputs of daily meteorological data and MODIS remote sensing data products such as vegetation property dynamics and land cover [25].NDVI and ET data were resampled to 1000 × 1000 m resolution.We obtained the 2000-2020 afforestation data from China Forestry Statistical Yearbooks and Inner Mongolia Forest Resources Inventory data (Forestry and Grassland Bureau of Inner Mongolia Autonomous Region, https://lcj.nmg.gov.cn/,accessed on 21 August 2023).We obtained the 2000-2020 water resource data from the Inner Mongolia Water Resource Bulletin (Water Resources Department of Inner Mongolia Autonomous Region, http://slt.nmg.gov.cn/,accessed on 21 August 2023).

Identifying the Afforestation Regions
The identification of afforestation regions is an important prerequisite to evaluating the ecological risk of water resources caused by afforestation [26,27].Compared with natural forestland, the NDVI of artificial forestland is more easily affected by human activities, and the change trend is more obvious [28].In this study, the NDVI and land cover data were used to analyze the spatial characteristics of vegetation and forestland, afforestation areas were identified through the combination of the two, and finally, sample points were used to verify the results (Figure 3).The main steps are as follows: (1) Trend analysis was used to identify areas with significant increases in the NDVI in Inner Mongolia from 2000 to 2020.Trend analysis is a linear regression analysis of time-dependent variables [29,30].The constantly changing properties of vegetation may be reflected in the trend of changes in the NDVI for each grid by applying linear trend analysis.In this study, the trend of NDVI change in Inner Mongolia from 2000 to 2020 was determined using a unitary linear regression model, and the slope of the trend was calculated using the least-square method as follows: where Slope refers to the trend of vegetation change, n is the number of years studied (n = 21 in this study), i is the ordinal number of a given year, and NDV I i denotes the NDVI value for year i; in the case of a Slope > 0, this indicates that the NDVI tends to increase.
A significance test is often used to assess the accuracy of the trend change.In this study, we assessed the significance of trends using the F-test (p < 0.05).The calculation's formula is as follows: where U refers to the sum of squares of errors, Q is the regression square sum, y i is the NDVI value for year i, ŷi is the NDVI regression value for fear i, y is the average NDVI value in n year, and i is the ordinal number of a given year.
(1) Trend analysis was used to identify areas with significant increases in the NDVI in Inner Mongolia from 2000 to 2020.Trend analysis is a linear regression analysis of time-dependent variables [29,30].The constantly changing properties of vegetation may be reflected in the trend of changes in the NDVI for each grid by applying linear trend analysis.In this study, the trend of NDVI change in Inner Mongolia from 2000 to 2020 was determined using a unitary linear regression model, and the slope of the trend was calculated using the least-square method as follows: where  refers to the trend of vegetation change,  is the number of years studied ( = 21 in this study),  is the ordinal number of a given year, and   denotes the NDVI value for year ; in the case of a  > 0, this indicates that the NDVI tends to increase.A significance test is often used to assess the accuracy of the trend change.In this study, we assessed the significance of trends using the F-test (p < 0.05).The calculation's formula is as follows: (3) where  refers to the sum of squares of errors,  is the regression square sum,   is the NDVI value for year ,   ̂ is the NDVI regression value for fear ,  ̅ is the average NDVI value in n year, and  is the ordinal number of a given year.4).The second series of data were sampling points, which were used to verify the spatial distribution of afforestation and mainly comprised 23 field survey samples and 147 samples collected from the Tsinghua University global land cover dataset (Figure 5).The overall identification accuracy was 78.24%.afforestation identification was mainly verified with two kinds of data.First, the afforestation area determined during 2000-2020 in this study was verified with statistical data.From 2000 to 2020, the government in Inner Mongolia carried out four forest resource inventories, the forest area increased from 20.51 × 10 4 km 2 to 26.15 × 10 4 km 2 , and the afforestation area was 5.64 × 10 4 km 2 (Figure 4).The second series of data were sampling points, which were used to verify the spatial distribution of afforestation and mainly comprised 23 field survey samples and 147 samples collected from the Tsinghua University global land cover dataset (Figure 5).The overall identification accuracy was 78.24%.To better demonstrate the ecological risks of water resources caused by afforestation, the extracted afforestation raster data (1000 × 1000 m) were converted into point data through areal averaging.

Water Balance
To understand the influence of afforestation on water resources in Inner Mongolia, the water balance equation was used to calculate the change in water resources in afforestation regions [31], which can be described as follows: where Q refers to the water resource (mm); P is the precipitation (mm); ET a is the actual evapotranspiration (mm); and ∆S indicates basin water resource change (mm), which is generally assumed to be zero in the long term [15,16,21,22].In the study area, deep groundwater is usually extracted for agricultural irrigation, and afforestation activities mainly affect shallow groundwater.Therefore, the influence of agricultural irrigation water on the water balance of afforestation is ignored in the calculation.
In order to detect the changing trend of water resources caused by afforestation during 2000-2020, the least-square linear regression model was used.

Ecological Risk
The water security index (WSI) was used to evaluate the ecological risk of water resources caused by afforestation, which quantified regional water security from the perspective of regional supply and demand balance [32].The equation is as follows: where P refers to the water resource supply (mm), assuming that the afforestation area receives its water supply only through precipitation; Pe is the effective precipitation (mm) and is calculated using the soil conservation service method developed by the U.S. Department of Agriculture (USDA); D is the water resource requirement (mm); and ET a was considered the water resource requirement of the afforestation area.Thus, WSI < −0.5 indicates high risk; −0.5 ≤ WSI < 0 indicates low risk; 0 ≤ WSI < 0.5 indicates low security; and 0.5 ≤ WSI < 1 indicates high security.

The Spatial Distribution of Afforestation in Inner Mongolia
According to the changing trend of the NDVI in Inner Mongolia from 2000 to 2020, ~65% of the region's NDVI exhibited an increasing trend, and vegetation coverage significantly improved, mainly distributed in the central and eastern parts of Inner Mongolia (Figure 6).Within these significantly increased land areas based on the NDVI, their forest cover was designated as afforestation.The spatial distribution of afforestation showed pronounced spatial heterogeneity, and afforestation was mainly distributed in the eastern part of Inner Mongolia (Figure 7).The forest coverage area in the study area increased from 20.74 × 10 4 km 2 in 2000 to 26.11 × 10 4 km 2 in 2020.The afforestation area in Inner Mongolia was 5.37 × 10 4 km 2 from 2000 to 2020.This is consistent with the results of the forest resource inventories in Inner Mongolia from 2000 to 2020.The growth rate was 0.27 × 10 4 km 2 /year.The forest coverage rate increased from 17.53% to 22.07%.Hulunbuir had the largest afforestation area, with an area of 2.46 × 10 4 km 2 , accounting for 45.81% of the total afforestation area.The afforestation area of Chifeng was second only to Hulunbuir, accounting for 10.75% of the total afforested area.The afforestation area of Wuhai and Alxa in western Inner Mongolia was relatively small, accounting for 0.24% and 1.57%, respectively.The spatial distribution of afforestation showed pronounced spatial heterogeneity, and afforestation was mainly distributed in the eastern part of Inner Mongolia (Figure 7).The forest coverage area in the study area increased from 20.74 × 10 4 km 2 in 2000 to 26.11 × 10 4 km 2 in 2020.The afforestation area in Inner Mongolia was 5.37 × 10 4 km 2 from 2000 to 2020.This is consistent with the results of the forest resource inventories in Inner Mongolia from 2000 to 2020.The growth rate was 0.27 × 10 4 km 2 /year.The forest coverage rate increased from 17.53% to 22.07%.Hulunbuir had the largest afforestation area, with an area of 2.46 × 10 4 km 2 , accounting for 45.81% of the total afforestation area.The afforestation area of Chifeng was second only to Hulunbuir, accounting for 10.75% of the total afforested area.The afforestation area of Wuhai and Alxa in western Inner Mongolia was relatively small, accounting for 0.24% and 1.57%, respectively.
to 2020.This is consistent with the results of the forest resource inventories in Inner Mon-golia from 2000 to 2020.The growth rate was 0.27 × 10 4 km 2 /year.The forest coverage rate increased from 17.53% to 22.07%.Hulunbuir had the largest afforestation area, with an area of 2.46 × 10 4 km 2 , accounting for 45.81% of the total afforestation area.The afforestation area of Chifeng was second only to Hulunbuir, accounting for 10.75% of the total afforested area.The afforestation area of Wuhai and Alxa in western Inner Mongolia was relatively small, accounting for 0.24% and 1.57%, respectively.

ET Comparison between Artificial Forestland and Natural Forestland
The effects of afforestation on regional water resources were studied by comparing ET changes in artificial and natural forests.The annual ETa of artificial forests in Inner Mongolia was 488.77 mm, slightly higher than that of natural forests (418.82mm), and the potential ET (PET) of artificial forests was 1186.19 mm, much higher than that of natural forests (950.14 mm) (Figure 8a).Against the backdrop of global warming due to climate change and wetting in the Inner Mongolia Plateau [33], ET showed an obvious increasing trend from 2000 to 2020, with an increase at a rate of 7.15 mm/year in artificial forests and 5.25 mm/year in natural forests.The consumption of water resources in artificial forests was greater than that in natural forests.

ET Comparison between Artificial Forestland and Natural Forestland
The effects of afforestation on regional water resources were studied by comp ET changes in artificial and natural forests.The annual ETa of artificial forests in Mongolia was 488.77 mm, slightly higher than that of natural forests (418.82mm), an potential ET (PET) of artificial forests was 1186.19 mm, much higher than that of na forests (950.14 mm) (Figure 8a).Against the backdrop of global warming due to cl change and wetting in the Inner Mongolia Plateau [33], ET showed an obvious incre trend from 2000 to 2020, with an increase at a rate of 7.15 mm/year in artificial forest 5.25 mm/year in natural forests.The consumption of water resources in artificial fo was greater than that in natural forests.

Changes in Water Resources Caused by Afforestation
The change in water resources caused by afforestation during 2000-2020 was c lated using the water balance equation (Figure 9).Based on the results, the change in resources in Inner Mongolia caused by afforestation showed a decreasing trend and ous spatial heterogeneity.An increase in water resources was observed in ~43% o afforestation regions, mainly in the eastern areas with sufficient precipitation, espe Hulunbuir and Hinggan, whereas ~57% of the afforestation regions experienced crease in water resources, mainly in the central and western regions with insufficien

Changes in Water Resources Caused by Afforestation
The change in water resources caused by afforestation during 2000-2020 was calculated using the water balance equation (Figure 9).Based on the results, the change in water resources in Inner Mongolia caused by afforestation showed a decreasing trend and obvious spatial heterogeneity.An increase in water resources was observed in ~43% of the afforestation regions, mainly in the eastern areas with sufficient precipitation, especially Hulunbuir and Hinggan, whereas ~57% of the afforestation regions experienced a decrease in water resources, mainly in the central and western regions with insufficient precipitation, especially in Alxa, Ordos, Hohhot, Baotou, and Ulanqab.In terms of the changing trend of water resource consumption in afforestation areas over the period under study (Figure 10), it can be seen that the water resource consumption in afforestation areas was between 6.0 and 9.0 BCM from 2000 to 2020, showing an increasing trend, with a change rate of 62 MCM per year; the water resource consumption in 2011 was the smallest, and the minimum value was 6.24 BCM, while the water resource consumption in 2013 was the largest, and the maximum value was 8.51 BCM.The mean value across multiple years was 7.31 BCM.

Ecological Risk of Water Resources in Afforestation Area
The ecological risk caused by afforestation in terms of water supply was calculated In terms of the changing trend of water resource consumption in afforestation areas over the period under study (Figure 10), it can be seen that the water resource consumption in afforestation areas was between 6.0 and 9.0 BCM from 2000 to 2020, showing an increasing trend, with a change rate of 62 MCM per year; the water resource consumption in 2011 was the smallest, and the minimum value was 6.24 BCM, while the water resource consumption in 2013 was the largest, and the maximum value was 8.51 BCM.The mean value across multiple years was 7.31 BCM.In terms of the changing trend of water resource consumption in afforestation areas over the period under study (Figure 10), it can be seen that the water resource consumption in afforestation areas was between 6.0 and 9.0 BCM from 2000 to 2020, showing an increasing trend, with a change rate of 62 MCM per year; the water resource consumption in 2011 was the smallest, and the minimum value was 6.24 BCM, while the water resource consumption in 2013 was the largest, and the maximum value was 8.51 BCM.The mean value across multiple years was 7.31 BCM.

Ecological Risk of Water Resources in Afforestation Area
The ecological risk caused by afforestation in terms of water supply was calculated with the WSI.Assuming precipitation as the only source of water supply, most areas in the study area face ecological risks caused by afforestation (Figure 11).The ecological risk

Ecological Risk of Water Resources in Afforestation Area
The ecological risk caused by afforestation in terms of water supply was calculated with the WSI.Assuming precipitation as the only source of water supply, most areas in the study area face ecological risks caused by afforestation (Figure 11).The ecological risk increased from east to west, indicative of spatial heterogeneity.Notably, ~24% of the afforestation regions were at high risk, mainly in the central and western parts of Inner Mongolia, while ~52% of the afforestation regions were at low risk, mainly in the eastern part of Inner Mongolia, especially Chifeng and Tongliao.In addition, ~24% of the afforestation regions had low water security, mainly in the eastern part of Hulunbuir.There was no afforestation area with high water security in Inner Mongolia.increased from east to west, indicative of spatial heterogeneity.Notably, ~24% of the afforestation regions were at high risk, mainly in the central and western parts of Inner Mongolia, while ~52% of the afforestation regions were at low risk, mainly in the eastern part of Inner Mongolia, especially Chifeng and Tongliao.In addition, ~24% of the afforestation regions had low water security, mainly in the eastern part of Hulunbuir.There was no afforestation area with high water security in Inner Mongolia.
From the temporal changing trend of the WSI in afforestation regions (Figure 12), it can be seen that the WSI in afforestation areas gradually changed from low risk to high risk from 2000 to 2020, showing a decreasing trend.In 2008, the WSI was the largest, at −0.

Identification of Afforestation
Inner Mongolia is one of the key regions targeted by Chinese ecological restoration programs [23], and ecological restoration projects such as afforestation have a significant impact on regional water resources.An important basis for calculating and evaluating the ecological risk of water resources is the accurate identification of afforestation regions.In this study, the change characteristics of land cover data and NDVI data were combined to identify the afforestation areas in Inner Mongolia.The afforestation area, growth rate, and spatial distribution characteristics were consistent with the results of previous studies [23,26,27,34], indicating that large-scale ecological restoration projects have made some progress since 2000.Due to the large study area, the identification of forestland and shrub vegetation with low vegetation coverage was poor, and the actual afforestation area may be underestimated.In future studies, we plan to combine fieldwork and deep learning methods with high-resolution remote sensing data to identify afforestation areas at different time scales and analyze afforestation at different times and geographic locations [35][36][37][38].

Impacts of Climate Change on Large-Scale Afforestation
Forest plays an important role in regulating regional climate and has a significant influence on the regional hydrological cycle [39].Afforestation can not only increase the carbon capacity of ecosystems and reduce the impact of climate change [40] but it can also lead to a decrease in the surface temperature, thus reducing the occurrence of drought events [41].The spatial heterogeneity of precipitation led to the change in afforestation activities from east to west in Inner Mongolia.From east to west, the planting of tree species changed from trees to shrubs, and the ecological function changed from water conservation to wind protection and sand fixation [42,43].The suitable density of afforestation is related to climate, and it is determined based on the water balance of the soil-vegetation system: Some precipitation on the soil surface evaporates into the atmosphere, while some rainfall penetrates the soil to replenish the groundwater and maintain the normal growth of plants [44].When the density of afforestation exceeds the water supply capacity of the region, groundwater is consumed faster, which affects the survival of vegetation and leads to greater ecological risks [45,46].

The Impact of Large-Scale Afforestation on the Water Resources and Ecological Environment
Some studies show that the water consumption associated with afforestation is greater than that of natural vegetation [47].Water consumption in forests increases significantly, resulting in an imbalance in regional ecological water resources [2,16,21].In addition, afforestation in arid and semi-arid areas, where precipitation is insufficient, will have an impact on groundwater recharge [48].Precipitation is less in western Inner Mongolia than in eastern Inner Mongolia, and vegetation is more dependent on groundwater.Afforestation increases vegetation coverage.Water consumption through vegetation canopy interception and vegetation transpiration increases as vegetation coverage increases.The soil water mainly moves upward, decreasing the groundwater recharge and groundwater table [7,49].It is very difficult to restore the groundwater level once the water table depth has decreased [50], which suggests that these changes may lead to permanent decreases in the ground's capacity to store water [16,51].At the same time, other studies have shown that afforestation can alleviate groundwater depletion in areas with sufficient precipitation [45,52,53].In future research, we will study the influence of afforestation on groundwater in different climate areas in Inner Mongolia.
Large-scale afforestation increases water consumption and may exacerbate land degradation, especially through the plantation of fast-growing and short-lived vegetation [54].In arid and semi-arid regions with sparse precipitation and considerable evaporation, the stability of the ecosystem is poor due to the simple species composition and structure.Large-scale vegetation restoration destroys the original stable state of the ecosystem [5].In this process, the planted vegetation competes with the original vegetation for water, which changes the regional eco-hydrological processes and causes more severe ecological problems [55].The decrease in groundwater depth causes the degradation of surface vegetation and land [56].In future studies, we will more comprehensively consider the impact of afforestation on regional groundwater depth.

Future Ecological Risks to Afforestation
Afforestation in China, considered an expensive ecological restoration policy, has yielded far fewer returns than expected, with only 5-34% survival rates in the northwestern provinces [16,51].According to China's ecological plan for the next 15 years, in order to achieve carbon neutrality and carbon peak, the national forest coverage rate will reach 26% by 2035, which indicates that China will continue its afforestation program in the future.Therefore, we suggest (1) avoiding afforestation in arid and semi-arid areas, so as not to cause a permanent decline in regional water storage capacity; (2) selecting suitable vegetation according to local conditions, as this is an important prerequisite for promoting ecological restoration and sustainable development [57]; and (3) developing an optimized groundwater extraction plan since this is one of the most effective management strategies to protect and maintain the current ecological environment and ecosystem [58][59][60].It should also be noted that as the climate of the Inner Mongolia Plateau is warmer and wetter, the increase in precipitation in the future may reduce the ecological risks caused by afforestation [33,61].

Conclusions
In arid and semi-arid regions with limited water resources, it is crucial to assess the ecological risks of afforestation on water resources for achieving sustainable regional development.Based on the land cover data, NDVI data, and meteorological data, in this study, we analyzed the ecological risks of water resources caused by afforestation in Inner Mongolia from 2000 to 2020.The results show that afforestation in Inner Mongolia increased by 5.37 × 10 4 km 2 , mainly distributed in eastern areas.Water resources are scarce in Inner Mongolia, and therefore the region cannot support the sustainable growth of large-scale forestland.Large-scale afforestation increases total water consumption, leading to an increase in regional water resource consumption and becoming a potential source of ecological risks in the region.The central and western parts of Inner Mongolia will face greater ecological risk of water resources than the eastern part in the future.Developing optimized afforestation schemes and improving water resource use efficiency are crucial to avoid potential ecological risks.

Figure 1 .
Figure 1.The location and land use type of the study area.

Figure 1 .
Figure 1.The location and land use type of the study area.

Figure 3 .
Figure 3.The workflow of identifying the afforestation regions.

( 2 )Figure 3 .
Figure 3.The workflow of identifying the afforestation regions.(2)Five periods of land cover data (2000, 2005, 2010, 2015, and 2020) were used to analyze the increase in forestland regions during 2000-2020.The histogram analysis method was used to calculate the NDVI values of the increased forestland regions, and the NDVI range of 20-80% was considered the NDVI threshold of the afforestation regions.(3) Afforestation areas were determined by overlapping the results obtained in the prior two steps.The NDVI of the increased region that fell within the NDVI threshold of the afforestation region was used to identify an afforestation area.The accuracy of afforestation identification was mainly verified with two kinds of data.First, the afforestation area determined during 2000-2020 in this study was verified with statistical data.From 2000 to 2020, the government in Inner Mongolia carried out four forest resource inventories, the forest area increased from 20.51 × 10 4 km 2 to 26.15 × 10 4 km 2 , and the afforestation area was 5.64 × 10 4 km 2 (Figure4).The second series of data were sampling points, which were used to verify the spatial distribution of afforestation and mainly comprised 23 field survey samples and 147 samples collected from the Tsinghua University global land cover dataset (Figure5).The overall identification accuracy was 78.24%.

Figure 4 .
Figure 4. Results of the 6th to 9th forest resource inventory in Inner Mongolia.

Figure 4 .
Figure 4. Results of the 6th to 9th forest resource inventory in Inner Mongolia.

Figure 4 .
Figure 4. Results of the 6th to 9th forest resource inventory in Inner Mongolia.

Figure 5 .
Figure 5.The spatial distribution of afforestation verification points.

Figure 5 .
Figure 5.The spatial distribution of afforestation verification points.

Figure 8 .
Figure 8. ET comparison between artificial forestland and natural forestland: (a) comparison and PET; (b) variations in ETa for artificial forestland and natural forestland (2000-2020).

Figure 8 .
Figure 8. ET comparison between artificial forestland and natural forestland: (a) comparison of ETa and PET; (b) variations in ETa for artificial forestland and natural forestland (2000-2020).
25.The WSI in 2013 and 2020 were the smallest, at −0.57, indicating the highest ecological risk of water resources caused by afforestation.The ecological risk was high in 2003, 2009, 2011, 2013, 2015, and 2020.

Figure 11 .
Figure 11.Ecological risks in afforestation areas in Inner Mongolia.

Figure 11 .
Figure 11.Ecological risks in afforestation areas in Inner Mongolia.From the temporal changing trend of the WSI in afforestation regions (Figure12), it can be seen that the WSI in afforestation areas gradually changed from low risk to high risk from 2000 to 2020, showing a decreasing trend.In 2008, the WSI was the largest, at −0.25.The WSI in 2013 and 2020 were the smallest, at −0.57, indicating the highest ecological risk of water resources caused by afforestation.The ecological risk was high in2003, 2009, 2011, 2013, 2015, and 2020.

Figure 11 .
Figure 11.Ecological risks in afforestation areas in Inner Mongolia.