EN中文

Journal of Geographical Sciences >

2023 , Vol. 33 >Issue 7: 1397 - 1418

DOI: https://doi.org/10.1007/s11442-023-2135-5

Research Articles

Evaluation method and empirical application of human activity suitability of land resources in Qinghai-Tibet Plateau

  • XU Yong , 1, 2 ,
  • WANG Lijia , 1, 2, * ,
  • YANG Hua 1, 2
Expand
  • 1. Key Laboratory of Regional Sustainable Development Modelling, Institute of Geographic Sciences and Natural Resources Research, CAS, Beijing 100101, China
  • 2. University of Chinese Academy of Sciences, Beijing 100049, China
*Wang Lijia (1994-), specialized in regional sustainable development and human-environment interactions. E-mail:

Xu Yong (1964-), PhD and Professor, specialized in regional sustainable development and human-environment interactions. E-mail:

Received date: 2023-04-10

  Accepted date: 2023-05-15

  Online published: 2023-07-24

Supported by

The Second Tibetan Plateau Scientific Expedition and Research(2019QZKK0406)

The National Key Research and Development Program of China(2018YFD1100101)

Fold

Abstract

The current suitability evaluation methods for land resources human activity in China suffer from theoretical deficiencies related to fundamental data accuracy, elevation and slope classification, and suitability class judgment. Empirical application of these methods is also hindered by excessive evaluation indicators, data acquisition difficulties, and limited applicability to high altitude regions. To address these issues, this paper proposes a technical evaluation framework for the Qinghai-Tibet Plateau (QTP) that employs selected key parameters varying with elevation and slope to establish grid-scale evaluation models for construction land suitability (CLS) and arable land suitability (ALS). A generalized algorithm is then proposed for key parameters such as air density, air temperature, slope suitability for construction, and soil erosion resistance of sloping arable land. Empirical research is conducted using Milin County in southeast Tibet as a case study, with interval measurements of 100 m in elevation and 1° in slope. The evaluation model is tested using grid accuracies of 30 m, 50 m, 100 m, 250 m, 500 m, and 1000 m. The results reveal that: Firstly, the CLS and ALS can be categorized into five classes: highly suitable, suitable, moderately suitable, marginally suitable, and unsuitable, with varying area ratios under different grid accuracies. Secondly, existing construction lands in Milin County are mainly distributed in suitable, highly suitable, and moderately suitable CLS classes, accounting for over 94% of the total area studied under different grid accuracies. While arable land is mainly distributed in suitable, highly suitable, and moderately suitable ALS classes, accounting for over 96%. Thirdly, the empirical research in Milin County indicates that the evaluation method, quantitative model, and parameters algorithm for evaluating human activity suitability of land resources on the QTP are feasible and applicable, with a recommended grid accuracy within 100 m and a maximum of 250 m. Fourthly, the paper establishes a correspondence between land suitability (including construction land and arable land) and topographic factors (elevation and slope) that can be applied to the QTP. Finally, some professional defects in the evaluation methods of available land resources in Major Function Zoning and “Double Evaluations” of Territorial Spatial Planning in China when applied to the QTP are identified.

Key words： human activity; evaluation model; construction land suitability; arable land suitability; land resource; Qinghai-Tibet Plateau

Cite this article

XU Yong , WANG Lijia , YANG Hua . Evaluation method and empirical application of human activity suitability of land resources in Qinghai-Tibet Plateau[J]. Journal of Geographical Sciences, 2023 , 33(7) : 1397 -1418 . DOI: 10.1007/s11442-023-2135-5

1 Introduction

The suitability evaluation of land resources human activity aims to provide a scientific basis for the rational spatial allocation and development scale of various human activities according to the characteristics of land resources in the study area and the adaptive demand of different human activities towards land resources, selecting evaluation indicators and key parameters, establishing evaluation methods and quantitative formulas, and classifying the land resource suitability according to human activity types. Although research results on the land resources suitability for human activity are rare, studies on agricultural land suitability have been conducted for a long time (Zhou, 1964; Wu, 1989). In the 1930s, a preliminary analysis of agricultural land suitability was conducted in Britain based on land use survey (Wu, 1948). After the Second World War, Poland and the Soviet Union carried out similar suitability evaluations for agricultural land based on the agricultural regionalization (Wu, 1981). In the 1980s, China, drawing on international experience, conducted influential large-scale land resources suitability evaluation for agriculture, forestry, and animal husbandry in the Huang-Huai-Hai Plain by selecting a series of indicators, obtaining significant results (Li and Huang, 1989). However, there have been few influential studies on land resources suitability evaluations for agricultural, forestry, and pastoral in China over almost 30 years since then.
The suitability evaluation of land resource for construction, particularly for urban and rural residential or industrial land, has traditionally lagged behind its agricultural counterpart in China and abroad. The United States provided a notable case of such evaluation: the construction land resource suitability analysis in the upper reaches of the Gila River in Arizona, initiated in the late 1980s and early 1990s. The project developed a suitability evaluation framework for four types of target land, namely recreational land, industrial land, low-density housing land, and commercial land. Ten natural and nine socio-economic element indicators were selected to evaluate land use suitability in the study area according to matrix grid (Steiner et al., 2000). In China, the evaluation work began with resources and environment carrying capacity evaluation of post-Wenchuan earthquake reconstruction in 2008. Its central objective was to screen suitable land for construction, emphasizing terrain elevation and slope as key indicators. Suitable construction land for the reconstruction of destroyed towns and villages was determined based on classification and combination rules for terrain elevation and slope (Fan et al., 2009). Subsequently, this method has been extensively utilized in estimating of reserve construction land for Major Function Zoning (MFZ) in 31 provincial-level regions of China (GOSCPRC, 2011; Xu et al., 2011a; Fan, 2015; Fan et al., 2019), selecting disaster relocation land for post-disaster recovery and reconstruction after events like the Yushu and Lushan earthquakes and the Zhouqu debris flow, as well as calculating reserve construction land for urban and regional development planning and town planning at different scales (Dang et al., 2015; Tang et al., 2015). In 2017, the National Development and Reform Commission of China initiated provincial spatial planning pilot projects, treating the evaluation of resources and environmental carrying capacity and territorial development suitability (Double Evaluations) as the basic task of spatial planning. The method for evaluating the suitability of existing construction land was improved through factor extension, model refinement, and the addition of farmland suitability evaluation content (XNA, 2017; Fan, 2019). After undergoing empirical verification in the Territory Spatial Planning (TSP) pilot in Fujian Province and Liupanshui City, Guizhou Province, this methodology has been adopted in the Guidelines for Evaluation of Resource and Environment Carrying Capacity and Suitability of Land and Space Development (Trial) (Guidelines of Double Evaluations) issued by the Ministry of Natural Resources of China (Zhou et al., 2020).
Overall, considerable progress has been achieved both domestically and internationally regarding the key aspects of land resource suitability evaluation for agricultural and construction land use, which have found broad practical application. However, on the Qinghai-Tibet Plateau, there are particular issues such as excessive evaluation factors, data acquisition difficulties and limited applicability to high-altitude regions, particularly in the methods used to classify and combine terrain elevation and slope to determine suitability class. Although theoretically feasible, this approach lacks mechanistic professional scrutiny with respect to fundamental data accuracy, classification of elevation and slope, determination of suitability class among other elements. Referring to academic thought and framework developed from prior research achievements, our study aims to explore a straightforward and applicable quantitative model for the evaluation of construction and arable land suitability (CLS and ALS) on the Qinghai-Tibet Plateau. We selected key parameters that vary with changes in terrain elevation and slope, and conducted empirical research using Milin County in Southeast Tibet as an example. Our objectives include exploring a simple and applicable quantitative model for the CLS and ALS evaluation on the QTP, quantifying and analyzing the professional mechanism for suitability classification by narrowing down elevation and slop classification, and recognizing base data accuracy differences by comparing the evaluation results obtained from grids with varying degrees of accuracy and their compatibility with present land use.

2 Materials and methods

The basic requirements of land resources for human habitation and production are mainly reflected in the flatness of terrain, the abundance of oxygen in the air, the suitability of temperature, and the convenience of water use, all of which are closely related to elevation and slope (Fan et al., 2009; Xu et al., 2011a). As elevation increases, atmospheric pressure gradually decreases causing air to become thin, temperatures to decrease, and the degree of restriction on human habitation and production activities to increase. As slope increases, the cost of human habitation and production increases, and soil erosion is intensified. Therefore, elevation and slope can be considered as the two key factors for evaluating the suitability of human activities on land resources (Fan, 2019). Given that construction land and arable land are the main sites of human activity and population concentration on the QTP, and also the regions with the most severe impact of human activity on nature, the methodological model construction and key parameter estimation for evaluating land suitability will mainly focus on construction land and arable land.

2.1 Evaluation technical process

Firstly, data preparation involves the digital elevation model (DEM), land use and administrative boundary maps (Figure 1). Since graphic data from various sources may present inconsistencies in coordinate systems and graphic boundaries, it is necessary to conduct coordinate transformation and graphic boundary correction processing. The land use map typically serves as a reference for registering other map data to facilitate graphic overlay analysis. Key data include corresponding values of air density and temperature variations with elevation, soil erosion with slope changes.
Figure 1 Analytical framework
Secondly, determination of threshold key parameters, which mainly refers to the boundary values for classifying elevation and slope. Considering the large span of elevation and slope classifications for the MFZ and TSP on the QTP, it cannot effectively reflect its characteristics of high altitude, rugged terrain, and vertical differentiation. Thus, elevation and slope are finely classified according to different thresholds. Key parameters include air density parameters and slope suitability coefficient for evaluating construction land suitability (CLS), and air temperature parameters and soil erosion resistance coefficient for evaluating arable land suitability (ALS).
Furthermore, Suitability evaluation involves using a quantitative formula composed of air density parameters and slope suitability coefficient, combined with elevation and slope classifications, to conduct full coverage evaluation of CLS for the study area and classify suitability classes. Similarly, a quantitative formula composed of temperature parameters and soil erosion resistance coefficient, combined with elevation and slope classifications, is utilized to conduct full coverage evaluation of ALS.
Lastly, adaptability analysis entails extracting and generating layers of urban and rural residential land, independent industrial and mining land, and transportation land from the land use status map. By overlaying them with the CLS classes, the differentiation status of various types of different CLS classes is analyzed. Similarly, the arable land and garden land layers are extracted and generated from the land use status map, and by overlaying them with the ALS classes, the differentiation status of current ALS classes is analyzed.

2.2 Evaluation model

(1) Evaluation model of construction land suitability
Referring to the idea of using the intersection of elevation and slope to select suitable construction land in China, this study replaces elevation and slope with their closely related air density parameter and coefficient of slope suitability for construction, respectively. The multiplication of these two parameters can quantitatively evaluate CLS. The quantitative formula is defined as follows:
$\begin{matrix} H{{A}_{B}}={{\delta }_{TE}}{{\theta }_{TS}} \\ \end{matrix}$
where HAB is the characteristic value of construction land suitability, referred to as the construction land suitability index (The abbreviation is CLSI). δTE refers to the air density parameter of elevation TE. θTS represents the coefficient of slope suitability for construction of slope TS.
(2) Arable land suitability evaluation model
Similar to the quantitative formula of CLSI, ALSI can be determined by replacing elevation and slope with air temperature parameter and coefficient of soil erosion resistance of sloping arable land, respectively. Multiplying these two parameters yields a quantitative evaluation of ALS. The formula is represented as:
$\begin{matrix} H{{A}_{P}}={{\lambda }_{TE}}~{{\beta }_{TS}} \\ \end{matrix}$
where HAP represents the characteristic value of arable land suitability, referred to as the arable land suitability index (The abbreviation is ALSI). λTE is the temperature parameter of elevation TE. βTS is the coefficient of soil erosion resistance of sloping arable land of slope TS.

2.3 Study area and data sources

Milin County is located in the southeast of Tibet Autonomous Region and in the middle and lower reaches of Yarlung Zangbo River, between 28°39′-29°50′N and 93°07′-95°12′E, with a total land area of about 9500 km2. The terrain is high in the west and low in the east, with Nyenchen Tanglha Mountains in the north, Himalayas in the south, and the Yarlung Zangbo River Valley in the middle. The lowest altitude in the territory is about 2200 m, the highest is about 7300 m, and the average altitude is about 3700 m. It belongs to a plateau temperate semi-humid monsoon climate with significant vertical zonation features. The annual average temperature is 8.2℃, the annual precipitation is 641 mm, with 85% of precipitation occurring from June to September, and the frost-free period is 170 days. It is rich in flora and fauna, with a forest coverage rate of 48.11% and a total wood stock volume of 91.5 million m3. There are many rivers in the region, and the Yarlung Zangbo River runs from west to east across the entire region. Human activities are mainly concentrated in the Valley of Yarlung Zangbo River and along its tributaries.
Milin County has 66 administrative villages in 3 towns and 5 townships with a resident population of 26,200 as of 2020. In 2019, the gross domestic product (GDP) was 1.814 billion yuan, with the proportions of the primary, secondary and tertiary industries being 9.43%, 44.54%, and 46.03%, respectively. Major crops include corn, barley, wheat, peas and alfalfa. In 2020, the per capita disposable income of urban residents and farmers and herdsmen was 33,000 yuan and 19,700 yuan, respectively. The elevation difference of nearly 5000 m in Milin County reflects the overall vertical zonation feature of the Qinghai-Tibet Plateau (QTP). The land use characteristics in cities, towns, agricultural areas, and pastoral areas reflect the spatial types and distribution of human activities on the QTP. The moderate land area scale and approximate east-west spatial distribution form are conducive to the empirical application of the evaluation method model under site conditions and the comparison of grid element accuracy. Taking Milin County as a case, the evaluation method and empirical application of human activity suitability in land resources on the QTP have strong representativeness (Zhen and Yang, 1985; Zheng, 1996; Fan et al., 2015).

2.4 Data sources

Our analysis is primarily based on digital elevation models (DEM), land use data, and administrative boundaries. We used NASA newly released one-arcsecond (~30 m) NASADEM dataset, which has a higher elevation accuracy and less noise problems that are particularly severe in cloudy areas. The land use data was sourced from the Third National Land Survey Data of Milin County, while the county-level administrative boundary was extracted from the land use data. The data corresponding to air density, temperature, and elevation (altitude) were gathered from the comparison table of altitude and air pressure, air density and gravitational acceleration (http://www.doczj.com). Due to the absence of research concerning soil erosion on the QTP, we used the average values of the simulated results for five types of crops across eight years using Win-Yield software on the Loess Plateau as a substitute. The simulations were run on the Yangou watershed in Yan’an, with daily meteorological data obtained from at Yan’an Station (located roughly 4 km away from the mouth of Yangou). The crops simulated include alfalfa, corn, potatoes, sorghum, and soybeans with a simulation period spanning from 1997 to 2006 (Xu et al., 2005a; 2005b; 2008; 2011b; Yang and Xu, 2010). According to data from the Chinese Soil Database, simulations indicate that sloping arable land in the Yangou watershed features mainly yellow clay soil (loessial soil with a sandy texture). This soil type is similar to the alluvial wet ash soil found in planting land in Milin County, with both types predominantly composed of sandy clay loam. The average annual precipitation in the Yangou watershed is about 507 mm, mainly occurring between June and September, while in Milin County is about 652 mm, also largely confined to the period from June to September. Although Milin County has slightly higher precipitation levels compared to the Yangou watershed, both areas experience similar precipitation distribution patterns. The primary cause of soil erosion in both areas is hydraulic erosion. The coefficient of soil erosion resistance of sloping arable land in this paper focused on reflecting the relative relationship between soil erosion on sloping arable land and slope. Specifically, we examined the differences in soil erosion between various degrees of slope on sloping arable land. While we considered the impacts of crop type, soil texture and precipitation differences on soil erosion, these variables were not the primary focus of our study.
We divided elevation between 2000 m and 5000 m into classes at 100 m intervals, while altitude above 5000 m was set as a separate class. Slope classification was divided into 1° intervals, with a maximum gradient of 30°, and the grids precision are established at six levels: 30 m×30 m, 50 m×50 m, 100 m×100 m, 250 m×250 m, 500 m×500 m, and 1000 m× 1000 m. The elevation and slope classification of Milin County are shown in Figure 2.
Figure 2 Topographic elevation (a) and slope (b) classification in Milin County, Tibet

3 Key parameters

(1) Air density parameter
Oxygen is the most common and critical factor for evaluating the suitability of a regional space for long-term human living and non-agricultural production activities. The content of oxygen in air generally depends on the air density, which increases with higher air density resulting in higher oxygen content and vice versa. On the earth’s surface, the quantitative relationship between air density and altitude has been confirmed by atmospheric state equation.
Air density parameter is a characteristic value used to describe the relationship between human activity comfort and topographical altitude. Air density data corresponding to elevation (range from 1000 m to 6000 m) is standardized using the Maximum Difference Normalization Method and this normalized value is defined as the air density parameter. To simplify raster calculations, air density parameter was fitted to elevation for correlation, and they both follow a linear variation function (Figure 3a). The fitting equation is as follows:
$\begin{matrix} {{\delta }_{TE}}=1.1822-0.0002TE \\ \end{matrix}$
$\begin{matrix} {{R}^{2}}=0.9979 \\ \end{matrix}$
Figure 3 Fitting curve of key parameters with elevation and slope
This formula can be used to calculate the air density parameter at different elevations ($TE\in \left[ 2000,5000 \right]$) on the QTP, including the value of air density parameter for each raster grid on the DEM. To reduce computational complexity and data volume, the air density parameter values are calculated using the average value of elevation classifications.
(2) Coefficient of slope suitability for construction
Step transformation of sloping land is an important way to expand construction land in mountainous and hilly areas. After terrace construction, the area available for construction use is smaller than that on flat land due to the deduction of slope protection and reasonable avoidance. The steeper the slope, the smaller the area available for construction. The algorithm of step transformation of sloping land can be obtained in References (Mao, 2009; Xu et al., 2020). The ratio of actual construction land area after step transforming of sloping lands to the slope vertical projection area is standardized using the Maximum Difference Normalization Method, and this standardized value is defined as the coefficient of slope suitability for construction. To simplify raster calculations, a quadratic function relationship was found existing between the coefficient of slope suitability for construction and slope by fitting (Figure 3b). The expression is:
$\begin{matrix} {{\theta }_{TS}}=0.9904-0.0098TS-0.0004T{{S}^{2}} \\ \end{matrix}$
$\begin{matrix} {{R}^{2}}=0.9995 \\ \end{matrix}$
This formula can be used to calculate the coefficient of slope suitability for construction at different slope values ($TS\in \left[ 0,43 \right]$) on the QTP. Similar to air density parameter, the coefficient of slope suitability for construction values are calculated using the average value of slope classifications. When TS > 43°, θTS = 0.
(3) Air temperature parameter
Generally, solar radiation, light, temperature, and precipitation are all crucial climatic elements for crop cultivation. However, in China (especially on the QTP), solar radiation and light are non-limiting factors for crop cultivation. Temperature can indicate the maturation of crops and better reflect the zonal differentiation of crop cultivation. In this paper, we will not consider the impact of precipitation on crop cultivation as it is a key factor in water resources evaluation. Similar to air density, the quantitative relationship between air temperature and elevation on earth’s surface has been confirmed by atmospheric state equations. The air temperature parameter is a characteristic value reflecting the relationship between crop planting or maturity and elevation from the perspective of temperature or accumulated temperature. The temperature data corresponding to the elevation (ranging from 1000 m to 6000 m) were standardized using the Maximum Difference Normalization Method, and this normalized value is defined as the air temperature parameter. The relationship between air temperature parameter and elevation conforms to linear variation (Figure 3c), and the fitting equation is:
$\begin{matrix} {{\lambda }_{TE}}=1.1999-0.0002TE \\ \end{matrix}$
$\begin{matrix} {{R}^{2}}=1 \\ \end{matrix}$
This formula can be used to calculate the air temperature parameter on the QTP at different elevations ($TE\in \left[ 2000,5000 \right]$), including the calculated value of air temperature parameter for each raster grid on the DEM map taken from the mean value of elevation classifications, similar to the air density parameter.
(4) Coefficient of soil erosion resistance of sloping arable land
The coefficient of soil erosion resistance of sloping arable land is a characteristic value that reflects the relationship between the amount of arable land soil erosion and slope from the opposite perspective. It is expressed by the difference between the standardized soil erosion modulus value and 1. The soil erosion modulus of sloping arable land uses the 8-year simulated average value (Xi) of 5 crops in the Yan’an Yangou watershed of the Loess Plateau, and the following formula is used to standardize the data:
$\begin{matrix} {{\beta }_{TS}}=1-{{X}_{i}}/5000 \\ \end{matrix}$
It should be noted that the denominator value of soil erosion modulus standardization should be greater than its maximum value, since water and soil loss and soil erosion can also occur on flat farmland. Generally, it should be controlled between 110% to 120% of the maximum value. Thus, 5000 t/km2·a meet this requirement. The relationship between coefficient of soil erosion resistance of sloping arable land and slope conforms to quadratic function (Figure 3d), and the fitting equation is:
$\begin{matrix} {{\beta }_{TS}}=0.535-0.0094TS-0.0002T{{S}^{2}} \\ \end{matrix}$
$\begin{matrix} {{R}^{2}}=0.9616 \\ \end{matrix}$
This equation can be used to calculate the coefficient of soil erosion resistance of sloping arable land with different slopes ($TS\in \left[ 0,\text{ }33 \right]$) on the QTP. The coefficient of soil erosion resistance of sloping arable land takes the calculated value of mean slope classification. When TS > 33°, θTS = 0.

4 Results

4.1 Construction land suitability

Based on 30 m × 30 m DEM data, six grid accuracy are resampled for Milin County, including 30 m×30 m, 50 m×50 m, 100 m×100 m, 250 m×250 m, 500 m×500 m, and 1000 m×1000 m. After spatial overlay of elevation and slope, 3 fields are added to the attribute table for the air density parameter (δTE), coefficient of slope suitability for construction (θTS), and construction land suitability index (HAB), and the corresponding parameter values are input and calculated. According to the differentiation characteristics of the construction land suitability index, the CLS in Milin County was classified into five classes: highly suitable, suitable, moderately suitable, marginally suitable, and unsuitable with the classification threshold of 0.58, 0.48, 0.38, and 0.28 (Figure 4 and Table 1).
Table 1 Classes area of construction land suitability under different grid accuracy
CLS
classes		 CLS
value		 30 m 50 m 100 m 250 m 500 m 1000 m
Area
(km2)		 Ratio
(%)		 Area
(km2)		 Ratio
(%)		 Area
(km2)		 Ratio
(%)		 Area
(km2)		 Ratio
(%)		 Area
(km2)		 Ratio
(%)		 Area
(km2)		 Ratio
(%)
Highly
suitable		 ≥0.58 120.41 1.26 135.95 1.43 152.77 1.60 147.39 1.55 118.23 1.24 70.30 0.74
Suitable (0.58,0.48] 277.97 2.92 267.33 2.80 246.20 2.58 228.79 2.40 228.64 2.40 295.09 3.10
Moderately
suitable		 (0.48,0.38] 201.93 2.12 199.91 2.10 199.34 2.09 219.98 2.31 310.34 3.26 553.25 5.80
Marginally
suitable		 (0.38,0.28] 263.95 2.77 264.69 2.78 273.75 2.87 364.36 3.82 599.31 6.29 1218.69 12.79
Unsuitable <0.28 8666.56 90.93 8662.95 90.89 8658.77 90.85 8570.30 89.92 8274.30 86.82 7393.50 77.57
The general characteristic of the CLS classes under a 30-m grid accuracy is that the suitability decreases with the increase of elevation or slope, while within the same suitability class, slope decreases with elevation.
(1) Highly suitable class covers an area of 120.41 km2, accounting for 1.26% of the total area, with the main distribution near the Yarlung Tsangpo River’s mainstream and tributaries river estuaries on both sides. The terrain elevation lies between 2200-3000 m, with a maximum slope of 11° occurring between elevations of 2200-2300 m while the slope at elevation of 2900-3000 m is 2°.
(2) Suitable class covers an area of 277.97 km2, accounting for 2.92% of the total area, mainly distributed along the mainstream of the Yarlung Tsangpo and the major tributary valley. The terrain elevation is below 3500 m, with a maximum slope value of 15°occurring at an elevation of 2200-2300 m, while the slope at the elevation of 3400-3500 m is 2°.
(3) Moderately suitable class covers an area of 201.93 km2, accounting for 2.12% of the total area, mainly distributed in the middle and lower reaches of the Yarlung Tsangpo’s tributaries or secondary and above tributary valleys. The elevation is below 4000 m, with a maximum slope value of 19° appears at elevation of 2200-2300 m and a slope of 2° at elevation of 3900-4000 m.
(4) Marginally suitable class covers an area of 263.95 km2, accounting for 2.77% of the total area, relatively concentrated in the valley land of the Yarlung Tsangpo River’s upstream or secondary and above tributary valleys. The terrain elevation is below 4500 m, with a maximum slope of 22° occurring between 2200-2300 m, while the slope between 4400-4500 m is 3°.
(5) Unsuitable class covers an area of 8666.56 km2, accounting for a high proportion of the total area (90.93%). It is extensively distributed throughout the territory, with high altitude or steep slope being the key limiting factors.
The evaluation results obtained under the other five grid accuracies are consistent with the 30-m grid in terms of elevation and slope control, but there are differences in the areas of various classes and their spatial distribution (Figures 5 and 6). Compared with the 30-m grid, the highly suitable class area of 50 m, 100 m, and 250 m increased by 12.91%, 26.87%, and 22.41%, respectively. Meanwhile, the area of 500-m and 1000-m grids decreased by 1.81% and 41.62%, respectively. The suitable class area of 50-m, 100-m, 250-m, and 500-m grids decreased 3.83%, 11.43%, 17.69%, and 17.75%, respectively, while area of 1000-m grid increased by 6.16%. The moderately suitable class’s area of 50-m and 100-m grids is reduced by 1% and 1.28%, respectively, and the area of 250-m, 500-m, and 1000-m grids are increased by 8.94%, 53.69%, and 173.98%, respectively. As for the marginally suitable class, its area showed an increasing trend as the grid accuracy decreased compared to the 30-m grid. The area of 50-m, 100-m, 250-m, 500-m, and 1000-m grids increases by 0.28%, 3.71%, 38.04%, 127.05%, and 361.71%, respectively. The area for each unsuitable class decreased by 0.04%, 0.09%, 1.11%, 4.53%, and 14.69% for the other five grid accuracies, respectively. The area increases or decrease in each suitability class under different grid accuracies was due to the difference in grid accuracy and actual topographic relief, indicating the complexity of topographic relief in the study area (Zhang et al., 2004). The proportion of suitability classes area in the total area differed slightly between the 50-m and 100-m grid compared to the 30-m grid with an error of ±0.34%. The error was ±1.11% for the 250-m grid, while the 500-m and 1000-m grids differed significantly with the error ±4.11% and ±13.36%, respectively.
Figure 5 Partial map of construction land suitability under different grid accuracy
Figure 6 Classes area change of construction land suitability under different grid accuracy
As for the spatial distribution of suitability classes, the results for 50-m, 100-m, and 250-m grids were consistent with those for the 30-m grid, while those for the 500-m and 1000-m grids were quite different. The above analysis shows that the grid accuracy of 30 m, 50 m, and 100 m are applicable to the CLS evaluation on the QTP, whereas the grid accuracy of 250 m is slightly rough.

4.2 Arable land suitability

In the same process as the CLS evaluation, 3 fields were added to the attribute table for the air temperature parameter (λTB), coefficient of soil erosion resistance of sloping arable land (βTS), and arable land suitability index (HAP), and their corresponding parameter values were input and calculated on the elevation classification. According to the differentiation characteristics of the arable land suitability index, the ALS of Milin County was classified into five classes: highly suitable, suitable, moderately suitable, marginally suitable, and unsuitable with the classification threshold of 0.32, 0.274, 0.21, and 0.18 (Figure 7 and Table 2).
Figure 7 Distribution of arable land suitability under different grid accuracy
Table 2 Classes area of arable land suitability under different grid accuracy
ALS
classes		 ALS
value		 30 m 50 m 100 m 250 m 500 m 1000 m
Area
(km2)		 Ratio
(%)		 Area
(km2)		 Ratio
(%)		 Area
(km2)		 Ratio
(%)		 Area
(km2)		 Ratio
(%)		 Area
(km2)		 Ratio
(%)		 Area
(km2)		 Ratio
(%)
Highly
suitable		 ≥0.32 68.41 0.72 72.74 0.76 83.64 0.88 85.01 0.89 69.01 0.72 36.68 0.38
Suitable (0.32, 0.274] 292.14 3.07 293.62 3.08 278.86 2.93 256.68 2.69 237.79 2.49 271.82 2.85
Moderately
suitable		 (0.274, 0.21] 249.31 2.62 246.27 2.50 246.53 2.59 267.41 2.81 370.66 3.89 633.13 6.64
Marginally
suitable		 (0.21, 0.18] 128.89 1.35 130.16 1.37 136.66 1.43 178.39 1.87 288.00 3.02 551.70 5.79
Unsuitable <0.18 8792.08 92.25 8788.03 92.21 8785.14 92.18 8743.33 91.74 8565.37 89.87 8037.49 84.33
The general characteristic of the ALS classes with a 30-m grid accuracy are similar to the results of CLS, except for the differences in class area and limited elevation or slope.
(1) Highly suitability class covers an area of 68.41 km2, accounting for 0.72% of the total area, mainly distributed in the river valley along the Yarlung Tsangpo River’s mainstream and its tributaries into the river estuary on both sides. The elevation ranges between 2200-2300 m, with a maximum slope of 10° occurring between elevations of 2200-3000 m and a slope of 1° at 2900-3000 m.
(2) Suitable class covers an area of 292.14 km2, accounting for 3.07% of the total area, mainly distributed along the mainstream of the Yarlung Tsangpo and the main tributary valley. The elevation is below 3400 m, with a maximum slope of 15° appears at an elevation of 2200-2300 m and a slope of 2° at elevation of 3300-3400 m.
(3) Moderately suitable class covers an area of 249.31 km2, accounting for 2.62% of the total area, mainly distributed in the middle and lower reaches of the tributaries of Yarlung Tsangpo River or the valley land of the tributaries above the secondary. The elevation is below 4000 m, with a maximum slope of 20° at an elevation of 2200-2300 m and with a slope of 2° at 3900-4000 m.
(4) Marginally suitable class covers an area of 128.89 km2, accounting for 1.35% of the total area, and is relatively concentrated in the valley of the upper reaches of the tributaries or the secondary and above tributaries of Yarlung Tsangpo River. The elevation is below 4300 m, with a maximum slope of 22° when the elevation is 2200-2300 m and a slope is 2° at elevation of 4200-4300 m.
(5) Unsuitable class covers an area of 8792.08 km2, accounting for 92.25% of the total area, spatially distributed throughout the territory, with high altitude or steep slope being the key limiting factors.
The evaluation results obtained under the other five grid accuracies are consistent with the 30-m grid in elevation and slope control, but differences exist in the areas of various classes and their spatial distribution (Figures 8 and 9). Compared with the grid accuracy of 30 m, the highly suitable class area of 50 m, 100 m, 250 m, and 500 m increased by 6.33%, 22.26%, 24.27%, and 0.88%, respectively. Meanwhile, the area of 1000 m grid decreased by 46.38%. The suitable class area of 100 m, 250-m, 500-m, and 1000-m grids decreased 4.55%, 12.14%, 18.60%, and 6.95%, respectively, while area of 50-m grid increased by 0.51%. The moderately suitable class’s area of 50-m and 100-m grids is reduced by 1.22% and 1.12%, respectively, while the area of 250-m, 500-m, and 1000-m grids are increased by 7.26%, 48.67%, and 153.95%, respectively. With the decrease in grid accuracy, the area of the marginally suitable class shows an increasing trend compared with a 30-m grid. The area of 50 m, 100 m, 250 m, 500 m, and 1000 m increases by 0.98%, 6.03%, 38.41%, 123.44%, and 328.04%, respectively. The area for each unsuitable class under other five grid accuracies decreased by 0.05%, 0.09%, 0.55%, 2.58%, and 8.58%, respectively.
Figure 8 Partial map of arable land suitability under different grid accuracy
Figure 9 Classes area change of arable land suitability under different grid accuracy
The proportion of suitability classes area in the total area differed slightly between the grid accuracy of 50 m and 100 m compared the 30-m grid with the error of ±0.35% and ±0.16%, respectively, the error was ±0.52% for the 250-m grid, while the 500-m and 1000-m grids differed significantly with the error of ±2.38% and ±7.82%, respectively.
According to the spatial distribution of suitability classes, the results for 50 m, 100 m, and 250 m grids were consistent with those for the 30-m grid, while those for the 500-m and 1000-m grids were quite different. The above analysis shows that the grid accuracy of 30 m, 50 m, 100 m, and 250 m are applicable to the ALS evaluation on the QTP.

4.3 Adaptability analysis of suitability evaluate result to existing land use

The current construction land in Milin County mainly includes urban and rural residential land, independent industrial and mining land, and transportation land, with a total area of 32.05 km2 in 2019. Among them, the urban and rural areas and industrial and mining land covered 10.62 km2, while transportation land covered 21.43 km2. After spatial overlaid with the CLS at grid accuracies of 30 m, 50 m, and 100 m, it was found that the urban and rural areas as well as industrial and mining land, are mainly distributed in suitable class, accounting for 63.85%, 61.34%, and 56.07%, respectively. The highly suitable class follows closely behind, accounting for 23.01%, 26.17%, and 28.99%, respectively. The moderately suitable class account for 9.52%, 8.85%, and 10.04%, respectively. The area in marginally suitable class is very small, accounting for 2.28%, 2.21%, and 3.27%, respectively. The proportion of unsuitable class is low. However, the distribution of transportation land within every suitability class is tiny due to the linear character of traffic roads (Table 3).
Table 3 Existing construction land area in CLS classes under different grid accuracy
Class Urban, rural, industrial and mining land Transportation land
30 m 50 m 100 m 30 m 50 m 100 m
Area
(km2)		 Ratio
(%)		 Area
(km2)		 Ratio
(%)		 Area
(km2)		 Ratio
(%)		 Area
(km2)		 Ratio
(%)		 Area
(km2)		 Ratio
(%)		 Area
(km2)		 Ratio
(%)
Highly
suitable		 2.44 23.01 2.78 26.17 3.08 28.99 2.61 12.17 3.24 15.13 3.87 18.06
Suitable 6.78 63.85 6.52 61.34 5.96 56.07 8.36 39.03 7.82 36.47 6.97 32.52
Moderately suitable 1.01 9.52 0.94 8.85 1.07 10.04 3.71 17.31 3.68 17.18 3.81 17.76
Marginally suitable 0.24 2.28 0.23 2.21 0.35 3.27 1.88 8.78 1.96 9.16 2.31 10.78
Unsuitable 0.14 1.33 0.15 1.43 0.17 1.63 4.87 22.72 4.73 22.06 4.48 20.88
Total 10.62 100 10.62 100 10.62 100 21.43 100 21.43 100 21.43 100
The current agricultural land in Milin County primarily consists of arable land and garden land, covering areas of 71.23 km2 and 9.89 km2, respectively, in 2019. After spatial overlaid with ALS at accuracies of 30 m, 50 m, and 100 m, current arable land and plantation land were mainly distributed among highly suitable, suitable, and moderately suitable ALS classes, with minimal areas in marginally suitable and unsuitable classes (Table 4). Across various grid accuracies, over 96% of arable land and garden land were distributed within highly suitable, suitable, and moderately suitable ALS classes, while the marginally suitable class representing less than 1.8%, and the unsuitable class having a small proportion of under 1.40%.
Table 4 Area of arable land and garden land in ALS classes under different grid accuracy
Class Arable land Garden land
30 m 50 m 100 m 30 m 50 m 100 m
Area
(km2)		 Ratio
(%)		 Area
(km2)		 Ratio
(%)		 Area
(km2)		 Ratio
(%)		 Area
(km2)		 Ratio
(%)		 Area
(km2)		 Ratio
(%)		 Area
(km2)		 Ratio
(%)
Highly
suitable		 4.96 6.96 7.39 10.38 11.08 15.55 0.69 6.95 0.90 9.15 1.39 14.05
Suitable 53.45 75.05 51.36 72.11 45.99 64.57 7.69 77.74 7.55 76.38 6.81 68.89
Moderately suitable 11.41 16.02 10.96 15.39 11.77 16.52 1.28 12.91 1.23 12.43 1.39 14.04
Marginally suitable 0.73 1.02 0.75 1.05 1.20 1.69 0.12 1.24 0.11 1.06 0.17 1.70
Unsuitable 0.67 0.94 0.76 1.07 1.18 1.66 0.11 1.16 0.10 0.98 0.13 1.32
Total 71.23 100 71.23 100 71.23 100 9.89 100 9.89 100 9.89 100

5 Discussion

5.1 Pedigree of suitability class corresponding to topographic elevation and slope

The quantitative evaluation model of CLS and ALS constructed in this study is a scientific mechanism analysis and quantitative improvement of the current evaluation methods in China. Through empirical application and verification in Milin County, convincing results have been achieved. As the current domestic evaluation of land resource suitability for construction land and arable land still adopts a combination rule of elevation and slope to select the suitable land, it is necessary to convert the suitability class in this study into a combination of elevation and slope. According to the threshold of the CLS and ALS in Milin County, the corresponding pedigree of the CLS classes and elevation and slope is shown in Figure 10, and the corresponding pedigree of ALS classes and elevation and slope is shown in Figure 11. These two pedigree charts can serve as benchmark references to determine whether the current methods related to the CLS and ALS in China are reasonable and whether the results on the QTP are accurate.
Figure 10 Corresponding pedigree between CLS classes and elevation and slope
Figure 11 Corresponding pedigree between ALS classes and elevation and slope

5.2 Comparative analysis of available land resources in the Major Function Zoning

The purpose of available land resources evaluation in the MFZ is to measure the quantity and spatial distribution of reserve suitable construction land for large-scale urbanization and industrialization. This method first divides the elevation into five classes at the national level: <500 m, 500-1000 m, 1000-2000 m, 2000-3000 m, and ≥3000 m, and the slope is divided into five classes: <3°, 3°-8°, 8°-15°, 15°-25°, and ≥25°. Then, the available land resource is extracted according to the rules, such as elevation below 2000 m with slope less than 15°, elevation between 2000 m and 3000 m with slope less than 8°, and elevation above 3000 m with slope less than 3° (Figure 12). Compared with the corresponding pedigree of CLS classes and elevation and slope, the available land resource area under 2000 m calculated by the MFZ is roughly equivalent to the sum of suitable and highly suitable classes of CLS in this study. Meanwhile, areas between 2000 m and 3000 m have lost the highly suitable or suitable classes, while areas above 3000 m contain large areas of marginally suitable and unsuitable classes, with a marked lack of maximum elevation control over available land resources with slopes less than 3°.
Figure 12 Corresponding pedigree between available land resources and elevation and slope of the MFZ

5.3 Comparative analysis of land resources availability oriented to urban construction and agriculture function in the Territory Spatial Planning

The “Double Evaluation” is an important foundation for the TSP in China, and in the Guidelines of Double Evaluation regarding the availability evaluation of land resources oriented to urban construction and agriculture function, with the elevation being divided into five classes at the national level: <1000 m, 1000-2000 m, 2000-3000 m, 3000-4000 m, and ≥4000 m, and five classes of slope: higher, high, medium, relatively low, and low according to the rules of <3°, 3°-8°, 8°-15°, 15°-25°, and ≥25°. For elevations below 3000 m, the availability level remains unchanged based on the slope classification. While for elevations between 3000 m and 4000 m, the availability decreases by one class, and decreases by two levels when the elevation is above 4000 m (Figure 13a). After that, the availability oriented to urban construction function is adjusted according to the relief degree of land surface; for regions with elevations exceeding 5000 m are directly delineated as low-level. The guidelines related to the evaluation of arable land availability is similar to that of urban construction land, with the same elevation classification and five classes of slope determining usability at the national level: higher, high, medium, low, and lower according to the rules of <2°, 2°-6°, 6°-15°, 15°-25°, and ≥25°. Based on slope classification, the availability level remains unchanged when elevation below 2000 m, while decreases by one class between 2000 m and 3000 m, and decreases by two classes when above 3000 m (Figure 13b). Then, the availability oriented to agriculture function is adjusted according to soil texture, crop maturity, etc. (MNRPRC, 2020).
Figure 13 Corresponding pedigree between land resources availability and elevation and slope oriented to urban construction (a) and agriculture function (b) of the Territory Spatial Planning
While the evaluation of land resources availability in “Double Evaluation” has made significant improvements over the evaluation of available land resources in the MFZ, it classifies the availability oriented to urban construction and agriculture function, several issues remain, particularly for the TSP. These include rough classification and a lack of maximum elevation control for agricultural land usage. For urban construction, the dual evaluation’s usability level classification (Figure 13a) below 3000 m depends solely on slope, without considering elevation variation. Compared with the corresponding pedigree of the CLS classes and the elevation and slope (Figure 10), it can be observed that the availability class oriented to urban construction function depends entirely on slope below the elevation of 3000 m, without reflecting any variances in slope that would occur with changes in elevation; above 3000 m, the class is dominated by human subjectivity, has fewer categories, and exhibits a sudden transition in types; the threshold for the limitation at 5000 m is significantly higher than the 4500 m of CLS unsuitable class. Comparing the corresponding pedigree of the ALS classes and the elevation and slope (Figure 11), it becomes evident that the availability class oriented to agriculture function (Figure 13b) suffers from similar problems as the availability oriented to urban construction mentioned before. More importantly, it fails to include limited elevations of agricultural land such as 3400 m and 4300 m. According to Zheng’s investigation on the natural belt in Southeast Tibet, the upper elevation limit of wheat planting is about 3400 m, and highland barley can reach 4100 m in the Changdu City area. The field investigation in Shannan City found that the upper elevation limit of highland barley can reach 4300 m in some small mountain basins, valleys, and other local areas with suitable temperature and irrigation conditions (Zheng, 1985).

5.4 Extension and application of the evaluation method

The evaluation model established in this study addresses the deficiencies associated with land resources availability of urban construction and agriculture function of the TSP on the QTP according to the “Double Evaluation” guidelines. However, when promoting the applicability of this method to other regions or even the entire region of the QTP, it is necessary to follow the framework of the “Double Evaluation” guidelines and focus on key issues such as the spatial scale of study area, the continuity and scale of suitability classes, the abundance of water resources, the convenience of agricultural water usage, and the crop maturity periods, among others. The evaluation model is established based on the site conditions or grid units, and the evaluation results are only applicable to small-scale areas. For larger-scale areas, the results of construction land evaluation need to be revised in terms of connectivity, scale, and water usage requirements of the specific construction objects. In addition, the results of ALS must consider crop maturity periods, with barley’s required accumulated temperature (≥0℃) generally between 1300 and 1400℃ being an indicator crop for annual maturation on the QTP. Finally, empirical evidence from typical cases is needed to determine whether this evaluation model is applicable to low-altitude central and eastern regions in China.

6 Conclusions

Based on the existing research, we constructed a research framework for evaluating the human activity suitability of land resources on the QTP, developed a quantitative model based on raster grid for evaluating the CLS and ALS by mapping the key parameters onto elevation and slope classification, proposed universal formulas of key parameters such as air density, air temperature, slope suitability for construction, and soil erosion resistance of sloping arable land. Based on six different grid accuracies with elevation intervals of 100 m and slope intervals of 1°, the CLS and ALS of Milin County was evaluated, and the results are divided into five classes: highly suitable, suitable, moderately suitable, marginally suitable, and unsuitable. The main conclusions are as follows:
(1) Under a 30-m grid accuracy, the CLS classes of highly suitable, suitable, and moderately suitable in Milin County covers an area of 600.31 km2, accounting for 6.30% of the total area with the area of marginally suitable and unsuitable classes accounting for 2.77% and 90.93%, respectively. The grid accuracy errors of the results obtained with 50-m and 100-m grids were small compared with 30 m, and that of 250-m grid is acceptable, while the error of 500-m and 1000-m grids were greater. Under the grid accuracy of 30 m, the ALS classes of highly suitable, suitable, and moderately suitable covers an area of 609.85 km2, accounting for 6.40%, with marginally suitable and unsuitable class at 1.35% and 92.25%, respectively. Minor errors were observed under grid of 50 m, 100 m, and 250 m compared to the 30-m grid, while significant errors were found under intervals of 500 m and 1000 m. This indicates that the suitable raster grid accuracy for evaluating the CLS and ALS should be 100 m, with a maximum not exceeding 250 m.
(2) The existing urban, village, and industrial and mining land in Milin County are mainly distributed among CLS classes of highly suitable, suitable, and moderately suitable, accounting for 95.96% under a 30-m grid accuracy, with this proportion reaching 95.96% and 94.74% at 50-m and 100-m grids, respectively. The existing arable land and garden land are predominantly distributed amongst the ALS classes of highly suitable, suitable, and moderately suitable, which accounted for 96% or more under grid of 30 m, 50 m, and 100 m, with the marginally suitable class comprising less than 1.8%, and the unsuitable class being below 1.4%. This indicates that the proposed evaluation method, model, and parameter algorithms are feasible and applicable on the QTP.
(3) Based on the empirical research results obtained from Milin County, a pedigree has been established that corresponds to the elevation and slope and applies to QTP of CLS and ALS classes. It is used to compare and analyze professional deficiencies found in the evaluation methods for available land resources specified by the Chinese MFZ and TSP when applied to the QTP.

References

Publishing order | Descend order by publishing year | Descend order by cited within
[1]
Dang L J, Xu Y, Tang Q, 2015. The pattern of available construction land along the Xijiang River in Guangxi, China. Land Use Policy, 42: 102-112.

DOI

[2]
Fan J, 2015. Draft of major function oriented zoning of China. Acta Geographica Sinica, 70(2): 186-201. (in Chinese)

DOI

[3]
Fan J, 2019. Assessment Guidelines for Resource and Environmental Carrying Capacity and Territorial Development Suitability. Beijing: Science Press. (in Chinese)

[4]
Fan J, Wang Y, Wang C S et al., 2019. Reshaping the sustainable geographical pattern: A major function zoning model and its applications in China. Earth’s Future, 7(1): 25-42.

DOI

[5]
Fan J, Xu Y, Wang C S et al., 2015. The effects of human activities on the ecological environment of Tibet over the past half century. Chinese Science Bulletin, 60(32): 3057-3066. (in Chinese)

[6]
Fan J, Zhou C H, Gu H F et al., 2009. State Planning for Post-Wenchuan Earthquake Restoration and Reconstruction:Resource Environment Carrying Capacity Evaluation. Beijing: Science Press. (in Chinese)

[7]
General Office of the State Council of the People’s Republic of China GOSCPRC, 2011. Major function oriented zoning of China. http://www.gov.cn/zwgk/2011-06/08/content_1879180.htm2011-06-30/ 2021-12-29. (in Chinese)

[8]
Li X B, Huang R J, 1989. Evaluation of agricultural land adaptability in the Huang-Huai-Hai Plain. Natural Resources, 11(4): 32-38. (in Chinese)

[9]
Mao Y L, 2009. Mountainous urban planning and architectural design according to climate. Mountain Research, 27(5): 605-611. (in Chinese)

[10]
Ministry of Natural Resources of the People’s Republic of China MNRPRC, 2020. Assessment guidelines for resource and environmental carrying capacity and territorial development suitability. http://www.gov.cn/zhengce/zhengceku/2020-01/22/content_5471523.htm2020-01-19/ 2021-12-29. (in Chinese)

[11]
Steiner F, Mcsherry L, Cohen J, 2000. Land suitability analysis for the upper Gila River watershed. Landscape and Urban Planning, 50(4): 199-214.

DOI

[12]
Tang Q, Xu Y, Dong X et al., 2015. Appraisal of land resources security in the post-earthquake reconstruction area of Lushan earthquake. Acta Geographica Sinica, 70(4): 650-663. (in Chinese)

DOI

[13]
Wu C J, 1948. Education and enterprise of geography at Britain British. Acta Geographica Sinica, 15(2-4): 47-56. (in Chinese)

[14]
Wu C J, 1981. Promoting areal specialization of agriculture through development areal predominance. Acta Geographica Sinica, 36(4): 349-357. (in Chinese)

[15]
Wu C J, 1989. A review of agricultural geography development. Geographic Environment Research, 1(1): 10-17. (in Chinese)

[16]
Xinhua News Agency XNA, 2017. Provincial spatial planning pilot program issued by General Office of the Central Committee of the Communist Party of China and General Office of the State Council of the People’s Republic of China. http://www.gov.cn/zhengce/2017-01/09/content_5158211.htm2017-01-30/ 2021-12-29. (in Chinese)

[17]
Xu Y, Gan G H, Wang Z Q, 2005a. Topographic differentiation simulation of crop yield based on WIN-YIELD software in the loess hilly-gully region. Transactions of the Chinese Society of Agricultural Engineering, 21(7): 61-64. (in Chinese)

[18]
Xu Y, Tang Q, Fan J et al., 2011a. Assessing construction land potential and its spatial pattern in China. Landscape and Urban Planning, 103(2): 207-216.

DOI

[19]
Xu Y, Tian J L, Liu P L et al., 2005b. Topographic differentiation simulation of soil and water loss of slope farmland in Loess Plateau. Journal of Soil and Water Conservation, 19(5): 20-23. (in Chinese)

[20]
Xu Y, Yang B, Liu G B et al., 2008. Topographic differentiation simulation of crop yield and soil and water loss on the Loess Plateau. Acta Geographica Sinica, 63(11): 1218-1226. (in Chinese)

DOI

[21]
Xu Y, Yang B, Tang Q et al., 2011b. Analysis of comprehensive benefits of transforming slope farmland to terraces on the Loess Plateau: A case study of the Yangou Watershed in Northern Shaanxi Province, China. Journal of Mountain Science, 8(3): 448-457.

DOI

[22]
Xu Y, Zhao S, Fan J, 2020. Urban planning construction land standard and its revision based on climate and topography in China. Journal of Geographical Sciences, 31(4): 603-620.

DOI

[23]
Yang B, Xu Y, 2010. Topographic differentiation simulation of Alfalfa yield and soil and water loss in the Loess Plateau. Progress in Geography, 29(5): 530-534. (in Chinese)

DOI

[24]
Zhang C C, Qin K, Lu Y et al., 2004. The Theories and Methods of Spatial Analysis in GIS. Wuhan: Wuhan University Press. (in Chinese)

[25]
Zheng D, 1996. Research on the natural territory system of Qinghai-Tibet Plateau. Scientia Sinica Terrae, 26(4): 336-341. (in Chinese)

[26]
Zheng D, Yang Q Y, 1985. Some problems on the altitudinal belts in southeastern Qinghai-Xizang (Tibetan) Plateau. Acta Geographica Sinica, 40(1): 60-69. (in Chinese)

DOI

[27]
Zhou D J, Xu Y, Wang Y F et al., 2020. Methodology and role of “Double Evaluation” in optimization of spatial development pattern. Bulletin of Chinese Academy of Sciences, 35(7): 814-824. (in Chinese)

[28]
Zhou L S, 1964. Researching formation, development, internal structure and system of China’s agricultural zone. Acta Geographica Sinica, 30(1): 14-24. (in Chinese)

Options
Outlines

/

〈 〉