A Summary of the Survey Region

The Lhasa River Basin is a first-class tributary of the Yarlung Tsangpo River and is located in the central hinterland of the Tibet Autonomous Region (90°05′– 90°20′ E, 29°20′–31°15′ N). The Lhasa River flows through Nagqu City’s Gyali County, Lhasa City’s Lhunzhub County, Maizhokunggar County, Dazi District, Chengguan District, and Quxu County, as well as 10 additional districts and counties (Fig. 1). The terrain in the Lhasa River basin has a high degree of undulation (7074– 3568 m), with an average elevation of 4300 m; the basin belongs to a typical plateau temperate semi-arid climate zone, with an average annual temperature of −7–9.2 °C and an average precipitation of 340–700 mm, mostly concentrated in the rainfall area. The average evaporation for many years has been 1500–2200 mm; the vegetation is mainly alpine meadows and alpine grassland, the soil types are meadow soil and grass felt soil, and the soil’s organic matter content is higher.

As the principal portion of the Qinghai-Tibet Plateau, the majority of the Tibet Autonomous Region is underdeveloped, with protective development serving as the primary focus and a lack of capacity and ability for independent development. Other characteristics of the region include high altitude, a harsh climate, a sparse population, arid and low rainfall, infertile soil, and a lack of natural geographic conditions for large-scale development. The Lhasa River basin is situated in the Tibet Autonomous Region’s central hinterland. It encompasses the middle and lower reaches of the basin valley on both sides of the terrace development and is considered a typical wide valley river section. The land is relatively broad and flat, with a high soil organic matter content and a mild climate. The basin is highly populated and economically developed, with a high density of towns and citIes, a high road network density, and the Tibet Autonomous Region serving as an important area for agricultural and animal husbandry production. The greatest central city of the Tibet Autonomous Region is Lhasa, which is located in the basin. Although the basin area only makes up 3.02% of the region’s total territory, its population makes up 21.3% of the region’s total population. The basin has seen significant development over the last 20 years, and the Tibet Autonomous Region now relies heavily on it as its economic center. However, the basin is experiencing a number of ecological and environmental issues brought on by the reverse evolution of grassland, the worsening of soil erosion, and the decline in resource and environmental carrying capacity. These issues have resulted from the rapid economic development of the basin and the intensification of anthropogenic activities. In hindsight, the majority of relevant research in the Lhasa River Basin is concerned with studying the natural ecological environment. In order to provide helpful guidance and support for the sustainable development of the high intensity development region of the plateau, this study used the PLUS model and the InVEST model to simulate the spatial pattern of ecosystem carbon stock and its land use change response mechanism in the Lhasa River Basin under the future multi-scenario development mode [30].

Data Capturing

The study identified 14 factors, including natural geographic factors (desiccation, elevation, distance from the river, precipitation, etc.) and human environment factors (population, night lighting, GDP, etc.), as the drivers of land use change in the Tibetan Plateau. Temperature, ecological functional zoning, and elevation were selected as limiting factors in relation to the natural geography of watershed development and regional development policies. LULCC is a complex, dynamic change process that is driven by both the natural environment and anthropogenic effects. This study has identified a complex, dynamic process of change that is driven by both the natural environment and anthropogenic effects (Table 1).

(1) Nine periods of land use data, at a 30 m resolution, were taken from the Resource and Environmental Science Data Centre of the Chinese Academy of Sciences (https://www.resdc.cn, access date: 2023) and used as LULC datasets for the region in question. To assess the accuracy of the data, they were supplemented with visual interpretation, field surveys, and confusion matrix judgments. Ultimately, the watershed’s total categorization accuracy exceeded 90%. According to the primary and secondary classification standards of land use in the People’s Republic of China, combined with the land cover status of the watershed, the watershed region was classified into seven categories: cropland, shrub, waters, grassland, forest, constructed, and bare land. (2) Digital elevation data were obtained from the Geospatial Data Cloud Platform (https://www.gscloud.cn, access date: 16 May 2023; https://www.tpdc.ac.cn, access date: 16 May 2023), and slope extraction was completed on the basis of the DEM data, with a spatial resolution of 30 m. (3) Precipitation, dryness, and average annual temperature data were obtained from the Meteorological Bureau of the Tibet Autonomous Region, with a spatial resolution of 30 m, and the soil and soil erosion data were obtained from the National Tibetan Plateau Science Data Centre (https:// www.tpdc.ac.cn, access date: 16 May 2023); soil and soil erosion data were obtained from the National Tibetan Plateau Science Data Centre (https://www.tpdc.ac.cn), with a spatial resolution of 30 m. The Chinese Academy of Sciences Resource and Environmental Science Data Centre (https://www.resdc.cn, accessed 16 May 2023) provided four sets of socioeconomic data, primarily focusing on the spatial distribution of GDP and population; Open Street Map (https://www.openstreetmap.org) provided vector data useful for calculating the distances to roads; the National Geographic Information Resource Service (https://www.webmap.cn) provided vector data for watershed streams; the World Food Organization’s FAO’s Livestock Density Dataset (https:///data.apps.fao. org/) used grazing data, and the National Tibetan Plateau Science Data Centre (https://www.tpdc.ac.cn) provided night lighting data. (5) A database of the carbon density of the soil and vegetation in the watershed was created by combining pertinent experimental data with data on the carbon density of the terrestrial ecosystems found in scholarly publications published between 2000 and 2022. Furthermore, a uniform spatial resolution of 30 m was applied for raster data resampling.

Fig. 1.

A digital elevation model of the research area. GS (2020) 4619 should be noted.

InVEST Model

The InVEST model, which is made up of multiple sub-modules and algorithms, is a tool used extensively in the estimation of ecosystem carbon stocks because it can quantitatively evaluate the value of ecosystem services. It can also simulate changes in ecosystem carbon storage under various LULCCs (Table 2). The four primary components of the carbon stock of terrestrial ecosystems are dead organic matter, soil organic carbon, aboveground biomass, and below-ground biomass [29]. To estimate the entire regional carbon stock, the InVEST model computes the average carbon densities of various land-use categories and LULCs. The carbon stock can be calculated using the following formula:

where Ci is the carbon stock of all land use types. Ci_above is the above-ground carbon density, which is the carbon stock present in all vegetation surviving above the ground, such as tree trunks, branches, leaves, etc.; belowground carbon density, Ci_below, is the carbon stock in plants’ active root systems; Ci_soil is soil carbon density; Ci_dead is the carbon stock in mineral and organic soils, and the carbon density in dead organic matter. Ctotal is total carbon stock across all land use categories and land cover types.

Data on land use in the study region and the associated carbon density value of each land use type are needed for the InVEST model to calculate carbon stock. The carbon density value of each land use type in the study area can be found in the literature [29]. In order to calculate the carbon density of land types, temperature and precipitation factors are used to correct the carbon density data of the local land use types. The carbon stock in the Lhasa River Basin is then obtained through inversion. Scholars believe that there is little variation in the carbon densities of land types in similar climatic zones. The following is the calculation formula:

where C′ denotes the carbon density of the Lhasa River Basin; C′′ denotes the carbon density of the land types in similar climatic zones; MAP denotes the air temperature; MAT denotes the precipitation; CBT denotes the vegetation carbon density based on the precipitation; CSP denotes the vegetation carbon density based on the air temperature; denotes the soil carbon density based on the precipitation and the carbon density of the land use type.

PLUS Models

Scholars have developed a large number of land use models to simulate land use change in the context of future development scenarios, with the goal of better assessing and simulating regional land use changes. However, most of these models are linear simulations that do not adequately reflect the evolution of patch landscapes and do not adequately simulate changes in the underlying nonlinear relationships of LUCC [20]. In order to better represent the complex land use–land change system and achieve higher simulation accuracy with more similar landscapes, the PLUS model employs land use expansion rule deep mining and the multi-CA model, using the stochastic seeding mechanism of the types of scenarios. This model can excavate the deep driving factors of land expansion and landscape change, and LEAS can be used to analyze the potential land use conversion rules.

As was previously stated, 14 LULCC driving factors were chosen for use in this work to simulate future changes in the watershed’s land use and serve as a guide for future planning [50–54].

The PLUS model simulates patch evolution under various land use types using a multi-objective stochastic patch seeding technique based on threshold reduction.

where μK is the minimum number of new land use type patches required to generate land use type K, and is a random number between 0 and 1. New land use patches are created from the land use type K value. Twelve characteristics are utilized as a random forest for training, while 50 decision trees are used, and the sampling rate is 0.01.

Setting the Scenario and Site Needs

Scenario Setting

This study involved three steps. First, the conversion rule of LULCC in the study area was established, and nine periods of LULC data with a 30 m resolution (1980, 1990, 1995, 2000, 2005, 2010, 2015, 2018, and 2020) as well as data on the causes affecting land use change in the study region were prepared. Second, we simulated three distinct development scenarios for the study area and maximized the study area’s LULC regional space using the PLUS model and Markov chain. Third, we investigated how the ecosystem of the Lhasa River Basin responds to changes in land use and the geographic pattern of carbon stock under various scenarios (Fig. 2).

Fig. 2.

A framework for policy that connects ecological and institutional data.

The Lhasa River Basin, located in the Tibet Autonomous Region (TAR), is a significant region for agricultural and animal husbandry production. Its economic development is vital for driving the overall economic growth of the TAR. Recognizing its importance, both the central government of China and the TAR government have designated it as a key area for investment and development. Consequently, substantial investments in capital and technology have been made in the basin, resulting in a strong impetus for its development. We analyze the land use pattern that aligns with the current regional economic development and predict the land demand that aligns with future economic development. We propose a business as usual (BAU) land use change scenario that prioritizes regional economic development and pays less attention to the impact of reduced ecological carrying capacity on the regional ecological environment. Over the past two decades, the Lhasa River has emerged as a significant area in the Tibet Autonomous Region (TAR) due to extensive development. Following two decades of extensive development, the Lhasa River has emerged as a significant economic hub in the Tibet Autonomous Region (TAR). Despite the fact that the watershed area only constitutes 3.02% of the TAR’s total area, it is home to 21.3% of the region’s population. Additionally, the Lhasa River Basin is experiencing a substantial influx of migrants from other areas within the TAR. Given the population size, food security concerns, and the government’s policy stance, we suggest implementing a land use change scenario called cropland protection (CLP). This scenario is based on a land-use model that aligns with the existing cropland in the region. It prioritizes expanding cropland to meet the needs of the growing population and economic development of the basin, with less emphasis on ecological protection. The Tibet Autonomous Region (TAR) is a vital green ecological barrier in China and Central Asia, serving as the primary section of the Qinghai-Tibetan Plateau and the source of numerous major rivers in the region. It plays a critical role in preserving the ecological security of both China and Central Asia. The Lhasa River Basin, located in the Tibet Autonomous Region (TAR), serves as a crucial ecological barrier area. The ecological state of this region significantly influences the overall ecological environment of the TAR. In light of the Chinese government’s present objective to achieve carbon peaking and carbon neutrality, we suggest implementing an ecological land protection (ELP) land use change scenario. This scenario prioritizes the improvement of the ecological quality of the basin. We present an ELP land use alteration scenario that aligns with the Chinese government’s existing objectives for achieving carbon peaking and carbon neutrality. The following are the scenario designs’ goals and guiding principles:

1. The BAU scenario presupposes that land use will largely follow historical patterns and that there will not be any notable changes in the future. The probability of changes in the watershed’s Markov chain of land use from 2010 to 2020 is used to calculate land demand for the BAU scenario in 2030, and the probability of change in the watershed’s Markov chain of land use from 2030 to 2040 is used to calculate the BAU scenario in 2040 [55, 56];

2. The CLP scenario is based on the basic farmland protection regulations in the TAR and findings from a questionnaire distributed in the region on the minimum and maximum stocks of grain in various municipalities. These documents demonstrate the TAR governments and various municipalities’ strict commitments to farmland protection. This scenario presupposes the strict protection of farmland; promotes the allocation of some constructed land, some forest land, some grassland, and some waters for conservation; and returns all farmland that has previously been converted to other uses to a state of conservation, thus reversing the 60% conversion probability rate [57];

3. The ELP scenario is based on the greening of the TAR’s north and south mountains, a project to turn farmland back into forests, the 14th Five-Year Plan for Ecological Environmental Protection set out by the TAR government, and the ecological environmental protection plans of Tibetan cities and counties. These initiatives demonstrate the increased commitment of the TAR government and local governments to ecological environmental protection. This scenario is based on the supposition that the government strictly limits the growth of constructed land, promotes the return of farms to forests, grasslands, and lakes, and safeguards habitats such as forests and grasslands [58]. In this case, all farmland that slopes between 6° and 15° is turned into forest and scrubland, and all farmland that slopes more than 15° is turned into grassland. Furthermore, within the vicinity of 100 m from the water, a buffer zone has been established.

LULC Simulation in Several Situations

Using the PLUS model, the dynamics of land use change under the three development scenarios in the two periods of 2030 and 2040 were calculated, and the spatial distribution of land use in the Lhasa River Basin of the Tibet Autonomous Region under various scenarios was simulated (Table 4 and Table 5, and Fig. 3). Grassland and bare land, together making up 82.57% of the basin area, were determined to be the predominant land use types in the Tibetan Lhasa River basin. When subjected to the BAU scenario, the Lhasa River Basin will exhibit an overall declining trend of urbanization, while the watershed LULC under future development scenarios will display distinct development trends. The dynamic indices of the constructed areas for 2030 and 2040 are, respectively, −0.026593 and 0.013046. In other words, the growth in built land under the BAU scenario for the Lhasa River Basin will show a slow rate, while the natural environment shows an expanding trend. Constructed land in the Lhasa River Basin showed a decrease in scale under the CLP scenario; its dynamic indices for 2030 and 2040 were found to be −0.045171 and −0.014782, respectively. Under the CLP scenario, cropland clearly expanded, and its dynamic indices for 2030 and 2040 were found to be 0.020449 and 0.011836, respectively. Under the ELP scenario, the scale of constructed land continued to decrease, with the dynamic indices of the constructed land for 2030 and 2040 being −0.057994 and −0.038242, respectively, due to the afforestation of the north and south mountains of the Lhasa River Basin and the regional ecological and environmental policies. A relatively large increase in the area of forests, scrubs, watersheds, and grasslands was also seen.

Fig. 3.

Matrix of land use types transferred under various land use scenarios in the Lhasa River Basin at various times between 2020 and 2040. The land use transfer matrix for the 2020–2030 BAU scenario is shown in (a); the land use transfer matrix for the 2020–2030 CLP scenario is (b); the land use transfer matrix for the 2020–2030 ELP scenario is (c); the land use transfer matrix for the 2020–2030 BAU scenario is (d); the land use transfer matrix for the 2030 BAU to 2040 BAU scenario is (e); the land use transfer matrix for the 2030 BAU to 2040 CLP scenario is (f).

Fig. 4.

Variations in the ecosystem categories’ spatiotemporal patterns under each scenario between 2020 and 2040.

We counted and computed the areas of land use types under various scenarios in 2020–2040 in order to analyze the geographical and temporal variations in land use types in the Lhasa River Basin under the three scenarios. The distributions of grassland, constructed forests, and cropland spatial patterns in the Lhasa River Basin under the BAU, CLP, and ELP scenarios from 2020 to 2040 are depicted in Fig. 4. Under the BAU and CLP scenarios, the area of cropland in the Lhasa River Basin can be expected to grow significantly, by 132.2 km² and 220.2 km², respectively, in 2030. This increase will be mostly centered in the urban and rural areas located in the middle and lower reaches of the Lhasa River. The Lhasa River Basin’s forest area decreased to varying degrees under the BAU and CLP scenarios, by 19.8 km² and 89.8 km², respectively, mainly in the middle and lower reaches of the basin, which are more and less populated, respectively. Under the ELP scenario, the cropland area decreased by 92.4 km², mainly in the middle and lower reaches of the basin, which may be due to large-scale afforestation encroaching on this land type. The Lhasa River Basin’s forest area decreased by 19.8 km² and 89.8 km², respectively, under the BAU and CLP scenarios. These decreases were mostly concentrated in the urban and rural agricultural areas in the middle and lower portions of the basin, where we see higher concentrations of people, and high-elevation sections can be seen in the basin’s northwest. Constructed land showed a notable contraction under each of the three scenarios, falling by 64.7 km², 109.9 km², and 141.1 km², respectively, with the majority of this contraction occurring in the basin’s middle and lower reaches’ gently sloping areas.

Changes in the Carbon Stock in the Lhasa River Basin throughout Time and Space under Various Scenarios

Data on the carbon pools of various land types and the geographical environmental parameters of the basin were used to calculate and assess the spatial carbon stock in the Lhasa River Basin via the InVEST model. In Fig. 5, the findings show that, under various development scenarios, there are clear geographical disparities in the carbon stocks of terrestrial ecosystems in the Lhasa River Basin. The Lhasa River Basin’s carbon stock increased by 72,957.8 Mg under the 2030 BAU scenario and by 110,696.4 Mg and 173,918.1 Mg under the 2030 CLP and ELP scenarios, respectively, compared to the 2030 BAU scenario. This increase could be attributed to the carbon stock in the terrestrial ecosystems rising under the conditions of strict ecological protection, afforestation, and large-scale farmland expansion. These increases in carbon stock within terrestrial ecosystems as a result of stringent ecological protection, tree plantations, and large-scale cropland expansion are closely linked to this. Under the 2040 BAU scenario, the spatial carbon stock in the Lhasa River Basin decreased by a significant amount, −84,730.3 Mg, when compared to the 2030 BAU scenario. In contrast, the CLP and ELP scenarios showed increases in the carbon stock in the Lhasa River Basin amounting to 47,025.5 Mg and 157,904.1 Mg, respectively, when compared to the 2030 BAU scenario, because of the substantial contraction in construction and the huge growth in grasslands and forests under the CLP and ELP scenarios, which increased the watershed’s total carbon supply. It is important to note that the Lhasa River Basin’s overall spatial carbon stock decreased significantly under the same development scenarios in 2040 and 2030. This decline was caused by an increase in constructed land and a decrease in terrestrial ecosystems, such as forests, grasslands, scrubs, and waters.

Analysis of Potential Changes in Land Use under Various Scenarios

This study has proposed a land use prediction framework for future watershed sustainable development [59], which can effectively integrate regional sustainable development with global sustainable development and serve as a basis for decision-making in the context of regional sustainable development. The framework is based on the concept of watershed sustainable development from the perspective of ecosystem evolution. Furthermore, we have concentrated our research on terrestrial ecosystems because the United Nations Environment Programme (UNEP) views these as a crucial component of the global Sustainable Development Goals (SDGs). The importance of ecosystem services cannot be overstated in the context of local and national development planning, since they are crucial to the achievement of regional and global sustainable development [60, 61].

We used the PLUS model to simulate the LULCs in the Lhasa River Basin in 2030 and 2040 in order to verify the model’s accuracy in predicting the LULC in the region. The model’s high accuracy in simulating the LULC in the Lhasa River Basin is indicated by its kappa value of 0.928 and overall simulation accuracy of 0.974. Furthermore, in order to simulate the spatial pattern of land use in the Lhasa River Basin in 2030–2040, we set up three development scenarios based on the basin’s development policy, current conditions, and projected development trends. The BAU scenario revealed a slight increase in the basin’s cropland, a decrease in the area of forests, scrubs, grasslands, and waters, and a decrease in the area used for construction. The larger magnitude is primarily caused by the fact that a large number of people are stranded in the countryside, reclaiming a large amount of cropland, which results in the conversion of forests, scrublands, grasslands, and waters into cropland. The main reasons for this are these things, along with more people living in the watershed and more activities caused by humans, and the policy that doesn’t take into account how the lower ecological carrying capacity will affect the regional ecological environment. This aligns with the findings of Hao et al. [62] in the “one river and two rivers” watershed in Tibet; in the CLP scenario, cropland appeared to increase significantly, while at the same time, the watershed and grassland appeared to decrease to some extent. The reduction area is primarily concentrated in the river valley, which is a relatively flat and low terrain area, primarily because of cropland reclamation that has been done by humans. As a result of the influence of government policy, local residents have a greater relationship with the decrease in water and grassland caused by the large-scale development of cropland in the lower terrain area. Under the ELP scenario, there is a relatively large increase in forests, scrubs, grasslands, and water areas due to the greater emphasis on the protection of the ecological environment of the river basin. At the same time, strict restrictions on the expansion of the city’s scale to improve the efficiency of the utilization of the construction land appear to have increased. In addition to the watershed’s ecological environment increasing as a result of the policy of greater attention to ecological environmental protection, utilization efficiency, and artificial patches also appeared to decrease relatively significantly. This also caused the watershed’s economic development mode to shift from one of unrestricted expansion to one of high-quality, refined development, thereby achieving the coordination of the environment, population growth, and watershed ecology.

Fig. 5.

Spatial distribution characteristics of the carbon reserves under each scenario during 2020–2040.

Impact Analysis of LULC on Carbon Stock

The project mapped the spatial patterns and carbon stocks of terrestrial ecosystems in the Lhasa River Basin at high resolution under several future development scenarios. We were able to successfully integrate the contributions of human activity and the natural environment, as demonstrated by the medium- and high-resolution mapping [20, 21]. To ensure the measurability of the spatial raster of carbon density, the study also included field measurement techniques. In order to guarantee the correctness of the spatial patterns of carbon stocks in the watershed, this study also carried out the quantitative calibration and spatial accuracy matching of the LULC raster.

Prospects and Limitations

This study evaluated and analyzed the spatial pattern of carbon stock and carbon stock in the basin under the synchronized spatial riparian pattern. However, there are still limitations and uncertainties in the analysis [11, 12]. The PLUS model was used to generate the spatial pattern of land use under various future development scenarios in the Lhasa River Basin. The current local development policies and national development goals used in the study only represent three of the future development scenarios of the Lhasa River Basin; these do not fully represent the future development scenarios of the basin. Numerous factors, including national and regional policies, the political environment both at home and abroad, the conditions surrounding economic development, the availability of regional resources, and the environment, all constrain and influence regional development. Our three suggested land use change scenarios, BAU, CLP, and ELP, only take into account the watershed’s actual existing state and anticipated future development trend. According to the watershed’s current carbon stock, land use, and economic development, it appears that the watershed’s ecological quality has somewhat declined under the BAU scenario, which is unfavorable to the watershed’s development and the preservation of its ecological environment. There is also a risk of harm to the watershed’s resources and environment. When creating regional development policies, the TAR government should consider the relationship between the ecological environment and the effectiveness of resource utilization. Under the CLP scenario, the watershed addresses issues of population growth and food security, but the effects of these policies may be limited and have little bearing on the preservation of the ecological environment and the efficient use of land. Under the ELP scenario, the development of the watershed and the preservation of the ecological environment are not supported, even though the construction of the watershed ecosystems and the state of land use have somewhat declined. The ELP scenario has a restricted amount of construction land, but it can still have a positive impact on the watershed’s ecological environment protection and increase the productivity of both cropland and construction land. This has a direct impact on the watershed’s sustainable development. We plan to investigate more thorough development options for the watershed in later research. To satisfy the demands of different stakeholders in the watershed for the best possible land allocation within the watershed, consideration is given, for instance, to the effects on LULCs of future changes in the plateau climate and the implementation of the Tibetan Plateau Ecological Protection Law.