Spatial Distribution and Driving Mechanisms of Rural Settlements in the Shiyang River Basin, Western China

Abstract

In the context of the rural revitalization strategy, an accurate grasp of the spatial differentiation characteristics and influencing factors of rural settlements in underdeveloped arid inland river basins is urgently required. Taking the Shiyang River Basin as an example, rural settlements from 2000 to 2019 were examined via visual interpretation using satellite remote sensing data and official statistical data. Following the logic of “state characteristics—evolutionary patterns—influence factors—layout optimization”, the average nearest index, the gravity-center migration model, spatial statistical analysis, and other methods were used in combination with GIS. The spatiotemporal pattern evolution characteristics of rural settlements in the past 20 years were analyzed. The results revealed the following: (1) The distribution pattern of rural settlements in the study area is sheet-like and strip-shaped. The projects in the southeast are mostly distributed in a patchy pattern with high density, while the characteristics in the west and north are exactly opposite. The objects in the south are distributed in an alluvial area of rivers, while settlements in the north are located in an oasis area. (2) From 2000 to 2019, the scale of rural settlements in the Shiyang River Basin, where there is a simultaneous occurrence of newly-built and disappearing phenomena, shows an expansion trend that first increased quickly and then slowed down. Spatially, rural settlements in the basin show a clustering trend toward the southwest. (3) The distribution characteristics of rural settlements are close to water and roads, and they are greatly influenced by urban–rural integration and ecological migration. The results will provide a scientific basis for accelerating the modernization of rural areas and the construction of new rural areas according to local conditions.

1. Introduction

Rural areas refer to settlements that are dominated by agricultural populations [1]. The spatial distribution of rural settlement reflects the residential activities of rural residents in their distribution areas under certain natural, economic, and social development environments [2]. The spatial pattern and morphological characteristics of rural settlements continue to evolve with the development of rural areas, especially in developing countries. The evolution trend of human–land relations, with respect to the process of social development and its progress, has become an important factor affecting regional economic development [3,4,5].

Based on the importance of the spatial distribution and evolution of rural settlements for the sustainable development of rural economies and societies, it is of great significance to study the spatial pattern’s evolutionary characteristics and the driving factors of rural settlements for integrating rural land resources, in order to improve the level of land resource conservation and its intensive use to service the rural revitalization strategy [6]. Starting from the 19th century, research studies began with the topic of spatial forms and changes in rural settlements lands, and these were primarily qualitative in nature [7]. In the 20th and 21st century, scholars were interested in combining qualitative and quantitative methods with respect to studying spatial structures, morphological types, and development trends [8,9,10,11,12]. With the rapid development of GIS and RS, as well as the integration and application of theoretical landscape ecology and related approaches, the representation of rural settlements’ spatial patterns was gradually affected by multi-disciplinary integration and development with rich research content [13,14,15]. Currently, there are diverse research studies on the spatial pattern evolution of rural settlements. The research content involves spatial distribution characteristics, driving factors, layout planning, new rural construction, and the renovation of hollow villages [16,17,18,19]. In terms of research methods, point–axis theory, spatial analysis, gravity models, weighted Voronoi diagrams, and spatial autocorrelation are popularly applied [19,20,21]. Regarding research areas, many studies focus on rapidly urbanizing areas, metropolitan suburbs, the Loess Plateau region, and low mountainous and hilly areas [6,22,23]. Aimed at the driving factors, the existing literature has mainly explored how topographic conditions, transportation locations, and socio-economic situations impact the spatial pattern evolution of rural settlements [24]. However, these studies rely primarily on published land-use data, lacking high-precision extraction and validation methods [6,25]. In addition, static research based on data from a single period cannot reflect the dynamic changes in rural settlements, and there is difficulty with respect to accurately assessing their regularity characteristics. The development of RS and GIS methods has made them powerful tools for monitoring the dynamic analysis of rural settlement expansion in space and time.

A rural settlement is a comprehensive representation of rural society, economy, and culture, which are crucial for biodiversity and resource conservation [26]. In China, the traditional oasis settlements in the Shiyang River Basin formed under the influence of the natural environment [27,28,29,30]. However, in response to urbanization, economic development, and the improvement of regional functions, the spatial pattern of these traditional rural settlements is constantly changing [31,32]. Surprisingly, empirical studies on the evolution of oasis rural settlements with rapid urbanization in the background are quite limited, and they cannot grasp the regularity characteristics at spatiotemporal scales [33]. The way rural development takes place in arid inland river area is even more dependent on oases and water resources, with significant characteristic differences compared to other regions [34]. In particular, there is less research on the spatial characteristics and driving factors of rural settlements in this special geographical environment. In terms of space, the spatial location of the distribution of rural settlements is only described qualitatively, which makes rural residents’ planning and management practices in some areas too superficial with respect to achieving precise management. Against the backdrop of the continuous acceleration of rural agricultural industrialization and urbanization in China, analyzing the spatial distribution characteristics of rural settlements under the comprehensive impact of the regional environment is beneficial for understanding and mastering the distribution characteristics and land-use situation of rural settlements, and this has important practical guidance significance for improving the living environment.

In summary, the key issues that need to be addressed urgently for the sustained implementation of the rural revitalization strategy are as follows: What changes have occurred in the rural settlements in the basin since the policy of promoting rural settlements concentration was implemented in the early 21st century, and what are the key factors that affect the spatial pattern changes in rural settlements in the basin? An in-depth analysis of the spatial distribution pattern of rural settlements and its influencing factors is an important method that can be used to study rural settlements and understand and coordinate the relationship between humans and land. This study takes Shiyang River Basin as the research area using GIS, analyzes the spatial distribution characteristics and changes in rural settlements in the study area from 2000 to 2019, and discusses the relationship between them and various environmental factors. This study can provide important information on sustainable land management and regional development for decision makers, and it can also provide a scientific basis for the construction of rural revitalization, which is of great theoretical and practical significance for guiding rural settlements.

2. Study Area, Data and Methods

2.1. Study Area

The Shiyang River originates from the eastern section of the Qilian Mountains in Gansu Province, running across the Wuwei Basin and Minqin Basin, before finally disappearing into the desert (Figure 1). The Shiyang River Basin is one of the three inland river basins of the Hexi Corridor and is situated in its eastern part (101°22′–104°14′ E, 36°57′–39°27′ N). Here, the geographical ecosystem comprises mountains, oases, and deserts. The administrative divisions of the watershed include nine counties (districts) of four cities in Gansu Province: Tianzhu, Gulang, Liangzhou, and Minqin counties (districts) of Wuwei City; the Yongchang and Jinchuan counties (districts) of Jinchang City; portions of the Sunan and Shandan counties of Zhangye City, including the Sunan Huangcheng district and the Shandan military horse farm; and some parts of Jingtai County of Baiyin City, which altogether encompass an area of 41,600 km². The basin’s headwater rivers (from east to west) consist of the Dajing, Gulang, Huangyang, Zamu, Jinta, Xiying, Dongda, and Xida Rivers. The water sources of the rivers include atmospheric precipitation in mountainous areas and melting snow and ice at high elevations [34,35]. Situated at an elevation of 1200–5000 m, the Shiyang River Basin has a continental temperate arid and semi-arid climate and is surrounded by the Tengger Desert and Badain Jaran Desert to the east, north, and west. Thus, the Minqin Oasis in the Shiyang River Basin has become a key green barrier to the joining of these two deserts; its geographical position is unique, although this is typical of other ecological security barriers in western China.

The Shiyang River Basin’s total population was 2.27 million people in 2019, of which the rural population consisted of 1.17 million, and it had an urbanization rate of 48.86%. Wuwei City and Jinchang City are key zones of economic, political, and social development in the Shiyang River Basin; their populations together account for more than 90% of the basin’s total population. Accordingly, both cities are distinguished by concentrated populations and feature the most pronounced conflict with respect to water supply and demand, exhibiting the highest utilization rate of water resources in the Hexi Corridor [36].

2.2. Data Sources and Processing

The social and economic data sets were obtained from the official website of the National Bureau of Statistics of China and the official websites of local governments. The basic geographical data were obtained from the National Geographic Information Resources Directory Service System (http://www.webmap.cn, accessed on 16 June 2022). The rural settlement area was examined via the visual interpretation of remote sensing images and a detailed comparison of land survey data in Gansu Province. Different scholars have various understandings of rural settlements [6,17,23,37,38,39]. According to the Provisions of the Statistical Division of Urban and Rural Areas issued by the National Bureau of Statistics, in principle, the administrative units and division approved by the State Council are taken as the division objects. Therefore, a rural settlement is defined as “rural settlements below the designated town” [38]. Referring to the definition standards provided by studies, rural settlements mainly include land types that are closely related to the production and life of rural people: rural homesteads, commercial land, surrounding gardens, scattered trees, and other basic construction land that is not connected to cities and towns.

Landsat satellite data, ZY-1 02C satellite data, and ZY-1 02D satellite data were used to extract rural settlement information. Information about the derived images is summarized in

Table 1. Basic information on the remote sensing images used in the study.

Satellite	Manufacturer	Sensor	Source	Spatial Resolution/m	Time Resolution/Day	Number of Images
Landsat	NASA (Washington D.C, America) & USGS (Reston, Virginia, America)	TM	USGS (https://earthexplorer.usgs.gov/, accessed on 2 June 2022)	30	16	4
ZY-1 02C	Ministry of Natural Resources of the PRC (Beijing, China)	PMS	Natural Resources Satellite Remote Sensing Cloud Service Platform (http://sasclouds.com/chinese/normal, accessed on 8 June 2022)	5	3	8
ZY-1 02D		VNIC		2.5	3	8

. To ensure the quality of image data, large image coverage areas, few mosaic images, and low cloud content were used. The downloaded image product’s level is Level 1, which comprises image data that have undergone data analyses, denoising, registration, and other processing. Preprocessing was essentially concerned with stack, correction, mosaic, clip, and projection.

Layer Stack: Since the downloaded TM data contain independent bands (GeoTiff) and comprise grayscale images, it is necessary to use the layer stacking tool for band combinations in order to better reflect the true characteristics of the ground’s surface. Mosaic: Taking the image stitching of a certain period as an example, an image with good imaging quality is selected as the reference, and other images are stitched to it. During the stitching process, histogram matching and color matching are carried out, and the feather tool is used to achieve better stitching results. Finally, this method is also used complete the stitching of other images. Clip: The vector boundary data of the study area are used to clip the image so that it only includes the Shiyang River Basin area. All graphic data were transformed into a unified Albers equal-area conic projection using ArcGIS software (Version 10.6). Based on ArcGIS 10.6 software, three phases of rural settlement distribution maps for 2000, 2011, and 2019 were obtained by carrying out manual visual interpretation.

2.3. Methodology

The research framework of this paper consists of two parts: rural settlement identification and spatiotemporal pattern analyses using remote sensing images, and analysis of influencing factors for rural settlements (Figure 2). Based on ArcGIS, spatial technologies were used to comprehensively analyze the following: (1) the evolutionary characteristics of rural settlements, (2) the gravity center of rural settlements, and (3) their relationship with the factors driving their evolution.

2.3.1. Extraction of Rural Settlements

Using the visual interpretation method with respect to remote sensing images, this paper processed the remote sensing images of the Shiyang River basin in 2000, 2011, and 2019. First, the images were imported into ArcGIS. Then, the residential patches were identified according to the location, shape, layout, and other characteristics of the elements in each image. A new layer in the software was created; it was edited according to the contour of the settlements; and, finally, the maps of the residential patch were obtained for a three-year duration (Figure 3).

2.3.2. Changes in the Rural Settlement Area and Its Annual Change Index

The area of land in the rural settlement makes up the cornerstone of exploring how rural settlements have evolved. The spatial changes in the area can reflect the degree of rural settlement evolution over a period of time. As shown below, ∆RSA indicates the change in the area of a given rural settlement patch within a certain period:

$$∆RSA={RSA}_{n+i}-{RSA}_i$$

(1)

where RSA_i denotes the area of rural settlements in the early stage of the study, RSA_n+i denotes the area of rural settlements in the later stage of the study, and ∆RSA denotes the area of rural settlements during a certain period of study in the study area.

To obtain the yearly trend and its variation, an annual growth index of rural settlement area (RSAI) is introduced:

$$RSAI=\frac{RS{A}_{n+i}-RS{A}_i}{n\cdot RS{A}_i}\times 100\%$$

(2)

where RSAI denotes the annual percentage of change in the area of a given rural settlement patch. In both equations above, RSA_n+i and RSA_i represent the rural settlement area in the years of n + i and i, respectively; n denotes the time interval (in years).

The average nearest neighbor ratio (NNI) is obtained by comparing the observed values of the nearest neighbor distance with the expected values in a random mode [6,40]:

$$NNI=\frac{d\left(NN\right)}{d\left(ran\right)}$$

(3)

$$d(NN)={\sum }_i^n\frac{Min\left(d_{ij}\right)}{N}$$

(4)

$$d\left(ran\right)=0.5\sqrt{\frac{A}{n}}$$

(5)

where, NNI denotes the average nearest neighbor ratio; d(NN) denotes the nearest neighbor distance; N denotes the number of samples; D_ij denotes the distance from point i to j; Min(d_ij) denotes the distance from point i to the nearest neighbor point; d(ran) denotes the theoretical value of the average distance; A denotes the area of the research area. The nearest neighbor coefficient is used to determine the distribution pattern of rural settlement clusters, with NNI > 1 indicating dispersed distributions and NNI < 1 indicating aggregated distributions.

2.3.3. Rural Settlement Gravity Center Migration Model

A regional gravity center model was used to explore the temporal migration of rural settlements on the landscape. Conceptually, it relies on the principles of mechanics to express the regional center of gravity coordinates [41,42]:

$$X=\frac{{\displaystyle \sum _{i=1}^n(C_i\cdot X_i)}}{{\displaystyle \sum _{i=1}^n{C}_i}}$$

(6)

$$Y=\frac{{\displaystyle \sum _{i=1}^n(C_i\cdot Y_i)}}{{\displaystyle \sum _{i=1}^n{C}_i}}$$

(7)

where X and Y denote the longitudinal and latitudinal coordinates of the distribution centers of rural settlement patches, respectively; C_i denotes the area of the i-th rural settlement patch; n denotes the number of rural settlement patches; X_i and Y_i are longitudinal and latitudinal coordinates of the center of gravity of the i-th plate of rural convergence. The rural settlement center of gravity’s shift angle is as follows:

$$a_{i+1}=\left\{\begin{array}{c}\mathit{arctan}\left(\frac{y_{i+1} - y_i}{x_{i+1} - x_i}\right), x_{i+1}\ge x_i\\ \pi - \mathit{arctan}\left(\frac{y_{i+1} - y_i}{x_{i+1} - x_i}\right), x_{i+1}<x_i\end{array}\right.$$

(8)

where a_i+1 denotes the gravity shift angle of the rural settlement, and x_i and y_i denote the longitude and latitude of the rural settlement in the i-th year, while x_i+1 and y_i+1, respectively, denote the longitude and latitude of the rural settlement in the (i + 1)-th year. Hence, the distance that the rural settlement center of gravity has moved can be calculated as follows:

$$L_{i+1}=\sqrt{{(x_{i+1}-x_i)}^2+{(y_{i+1}-y_i)}^2}$$

(9)

Buffer analyses can reflect the proximity or influence degree of a geographical element or space object and its surroundings. The location factors selected in this study mainly include roads (linear elements) and rivers (linear elements). By establishing buffer zones for roads and rivers, the distributions of rural settlements and their relationships are reflected. The specific operations are as follows: Roads and rivers are used as objects, and the buffer analysis tool is used to establish concentric rings with different radii as buffer zones. Then, the relevant attribute data of each buffer zone and rural residential land are overlaid, and the scale and evolution characteristics of rural settlements within each buffer zone are calculated and statistically analyzed.

3. Results

3.1. The Spatiotemporal Distribution Pattern of Rural Settlements

As shown in Figure 4, the rural settlements in the Shiyang River Basin are concentrated in the oasis areas. They are mainly distributed in the Minqin Oasis, Liangzhou Oasis, Jinchuan Oasis, and Yongchang Oasis, as well as the valleys of Gula County, showing a pattern of patchy and strip distributions. The rural settlements in the Minqin Oasis are distributed along both sides of the highway between the urban area and Xiqu Town, while the rural settlements in Liangzhou District are distributed in patches with relatively high densities. The number of rural settlements in Jinchuan District is relatively small and distributed in strip patterns along both sides of the river and the northern oasis. The rural settlements in the eastern part of Yongchang County are distributed in a fan-shaped pattern in the river’s impact zone, and the rural settlements in the valleys are distributed in strips along both sides of the river. The distribution density of rural settlements in Gula County is relatively small, and most are located in mountainous and gully areas.

Figure 5 shows the spatial changes in the rural settlements in the Shiyang River Basin from 2000 to 2019. In the period from 2000 to 2011, the disappearance of rural settlements decreased, while the number of newly added objects increased (Figure 5a). It can be inferred that rural settlements experienced an expansion trend during the period from 2000 to 2011. The main areas where rural settlements increased were Liangzhou District, Gula County, and Minqin County. The newly added rural settlements in Liangzhou District were mainly distributed in the northeastern part of Liangzhou and the eastern oasis near the river. The same trend was observed in Gula County, and the rural settlements were mainly distributed in the northern part. Moreover, the newly established rural settlements in Minqin County were mainly distributed in the southern oasis. After these phenomena, many old rural settlements disappeared, while other new rural settlements were added from 2011 to 2019. The rural settlements showed a coexistence pattern between disappearance and expansion, but the former outnumbered the latter. The lost rural settlements were mainly concentrated in the southern part of Gula County and the northeastern part of Minqin County. The great increase in rural settlements was mainly clustered in Liangzhou District and Minqin County (Figure 5b). The disappearance of rural settlements was observed in the northern oasis of Minqin County and in the valleys and gully areas in Gula County, and this phenomenon was relatively obvious.

3.2. Area Evolution Characteristics of Rural Settlements

This study is based on the spatial data of rural settlements in the Shiyang River Basin from 2000 to 2019, and it calculates the area and annual growth rate of rural settlements in different regions, as shown in

Table 2. Area and annual growth rate of rural settlements in the Shiyang River Basin from 2000 to 2019.

Region	Increased Area/km2
	2000	2011	2019	2000–2011	2011–2019	2000–2019
Gulang	4.25	4.33	4.44	0.08	0.11	0.19
Jinchuan	0.59	0.6	0.66	0.01	0.06	0.07
Liangzhou	10.32	10.41	10.71	0.09	0.3	0.39
Minqin	7.61	8.48	8.62	0.87	0.14	1.01
Yongchang	3.67	3.69	3.76	0.02	0.07	0.09
Total	26.44	27.51	28.19	1.07	0.68	1.75

.

Regionally, there are significant differences in the expansion area of rural settlements in the Shiyang River Basin from 2000 to 2019. The largest and smallest areas are Minqin County and Jinchuan District, with expansion areas of 1.01 km² and 0.07 km², respectively. In terms of different periods, most counties had higher annual growth rates with respect to rural settlements from 2011 to 2019 than from 2000 to 2011. After 2011, except for the slowdown in the expansion rate and the reduction in expansion areas of the rural settlements in Minqin County, the expansion speed of rural settlements in other regions increased significantly, and the expansion area increased year by year.

3.3. Density Evolution Characteristics of Rural Settlements

The purpose of this section is to explore the distribution pattern of rural settlements in the Shiyang River Basin. Nearest neighbor ratios are calculated to observe the clustering pattern of newly added rural settlements in the Shiyang River Basin in two periods (Figure 6). The results show that the average nearest neighbor ratios of newly added rural settlements in the basin in the two periods of 2000–2011 and 2011–2019 are 0.39 and 0.34, respectively. The fact that the ratios are all less than 1 indicates that the newly added rural settlements are clustered. The z-values are −11.52 and −16.84, respectively, and both passed the 99% significance test. The overall distribution of rural settlements tended to be clustered, and the scale tended to be regular from 2000 to 2019. However, in the early stage, the expansion was relatively obvious. Until 2019, the density and scale of rural settlements in the Shiyang River Basin were still relatively small.

3.4. Evolution Characteristics of Rural Settlements’ Gravity Center

The changes in the gravity center of the rural settlements in the study area depend on the calculation of gravity center coordinates (Figure 7). Over the past 20 years, the rural settlements in the basin have exhibited a contraction trend. The gravity centers of counties in the northern basin have migrated southwest. Overall, there is a clear trend of westward migration of with respect to rural settlements. In the period from 2000 to 2011, overall, the gravity-center coordinates migrated to the northwest, and the gravity-center coordinates shifted to the west from 2011 to 2019. Minqin County had the greatest change in gravity center at 12,775.13 m (

Table 3. Gravity center of rural settlements in the Shiyang River Basin.

Region	2000		2011		2019
	X (m)	Y (m)	X (m)	Y (m)	X (m)	Y (m)
Liangzhou	818,780.40	4,202,954.74	819,337.81	4,202,812.45	819,717.85	4,202,745.22
Jinchuan	789,199.48	4,269,550.63	788,945.43	4,269,448.90	788,473.99	4,269,343.13
Minqin	878,549.00	4,301,874.09	876,054.53	4,298,522.33	870,622.77	4,291,855.16
Gulang	880,655.51	4,154,497.60	875,948.48	4,158,946.55	876,583.38	4,158,549.94
Yongchang	768,804.37	4,237,775.52	768,434.02	4,238,573.62	768,961.17	4,237,577.37
Total	845,513.25	4,225,374.97	844,573.88	4,226,507.26	844,429.09	4,226,321.74

and

Table 4. Moving distance for gravity center rural settlements.

Region	2000–2011		2011–2019		2000–2019
	Radian	Distance (m)	Radian	Distance (m)	Radian	Distance (m)
Liangzhou	−0.250	575.29	−0.175	385.94	−0.220	960.58
Jinchuan	2.761	273.67	2.921	483.16	2.863	754.58
Minqin	2.211	4178.11	2.254	8599.72	2.240	12,775.13
Gulang	3.899	6476.83	−0.558	748.61	3.925	5744.88
Yongchang	4.278	879.85	−1.084	1127.13	−0.901	252.69
Total	4.020	1471.22	0.908	235.34	−0.718	1439.37

). From 2011 to 2019, the distance was more than double the distance in the previous period. This was followed by Gula County, and its migration change mainly occurred from 2000 to 2011 at 5744.88 m. Yongchang County had the smallest change in gravity center migration, at 252.69 m. Liangzhou District and Jinchuan District were relatively stable during the two time periods.

4. Discussion

4.1. Development Model of Rural Settlements in the Shiyang River Basin

From 2011 to 2019, although many new rural settlements appeared, more established ones disappeared (Figure 8, single and piece). We investigated the changed coverage of rural settlement areas in the Shiyang River Basin, and our results show that, overall, it increased from 2000 to 2019. This increase can be explained by the increased establishment of rural settlements from 2000 to 2011. From 2011 to 2019, with relatively fewer newer rural settlements, the land area occupied per rural settlement unit increased, indicating that the scale of individual rural settlements is increasing. These findings suggest that the development mode of rural settlements in the Shiyang River Basin has undergone significant changes in only 20 years, shifting from being greater in number during 2000–2011 to greater in size in 2011–2019.

The spatial pattern changes in rural settlements obtained in the previous section can be used as a basis for identifying the development patterns of settlements. There are three main evolutionary patterns with respect to rural settlement size in the spatial distribution of the Shiyang River Basin, namely, the shrinkage mode, diffusion mode, and conversion mode [43,44]. The shrinkage mode refers to the transformation of rural settlements into urban or other types of land. This phenomenon mainly occurs in the outskirts or surrounding towns of the government’s jurisdiction. For example, from 2007 to 2010, Huaer village in the Xindun town of Zhangye city was transformed into an urban area after being requisitioned by government agencies as the city continued to expand. In particular, with the rapid development of urbanization, rural settlements near towns with larger urban scales and better economic conditions began to transform into urban land. The diffusion mode refers to the expansion of rural settlements outward from their original locations. For example, from 2012 to 2017, Dongwan Village near Ningyuanbao town followed the approach of “reclaiming land from the mudflat, demolishing the olds and building the news”, and built thousands of high-standard farmhouses on large areas of a barren beach, bringing the settlement closer to the city. In addition, the government implemented immigration policies, such as the “Dingxi and Hexi” immigration policy and the “Ecological immigration and poverty alleviation development in Huanghuatan”, and promoted the expansion of rural settlements with the introduction of external populations. The conversion mode comprises changing the structure and morphology of settlements via renovation and reconstruction. After the transformation of idle villages in Xiejiawan Village, Wujiang Town, and Ganzhou District, the villages’ styles were maintained with a shared farm. The development of the tourist check-in industry is in full swing. The increasing efforts in the construction of new rural areas have improved the appearance of villages, and some dilapidated and scattered settlements have been converted from construction land to farmland.

4.2. Factors Influencing the Evolution of the Spatial Pattern of Rural Settlements

4.2.1. Natural Factors

Topography is one of the main factors in the selection of sites for rural settlements. By carrying out an analysis of the DEM within the study area, it is observed that the terrain in the area where rural settlements are located does not have significant undulations and changes. The main changes are in the Qilian Mountains with a few settlements, and they rely on the piedmont fans of the mountains for development. Therefore, factors such as topography, slope, and their aspects do not have a significant impact on the development and changes in the settlements.

(1)

Distance between rural settlements and rivers

Specifically, the source of water is a pivotal consideration, as it is the most vital resource for production and life in arid regions. By overlaying the river distance layer with a layer of rural settlements, it can be observed that the rivers are distributed in all towns and villages in the Shiyang River Basin, with the most rivers being distributed in Liangzhou District. The distance from each rural settlement to the nearest river was categorized into one of four levels: <1000 m, 1000–2500 m, 2500–4000 m, or >4000 m. As observed in Figure 9, rural settlements in the Shiyang River Basin are mainly distributed in the proximity of rivers. The rural settlements are located more than 4000 m away from the rivers that are mainly distributed in the valley areas of Gula County and the northwest part of Liangzhou District. The number of rural settlements within 1 km of a river in 2000, 2011, and 2019 accounted for 38.88%, 40.23%, and 39.83%, respectively, of the totals in those years. By contrast, the number of rural settlements situated more than 4 km from a river in 2000, 2011, and 2019 accounted for substantially lower proportions (7.32%, 7.54%, and 8.59%, respectively) of the same totals (

Table 5. The relationship between the number of rural settlements and distance from the nearest river.

Year	<1000 (m)		1000–2500 (m)		2500–4000 (m)		>4000 (m)
	Number	Percentage	Number	Percentage	Number	Percentage	Number	Percentage
2000	2475	38.88	2038	32.01	1387	21.79	466	7.32
2011	2839	40.23	2235	31.67	1451	20.56	532	7.54
2019	2885	39.83	2265	31.27	1471	20.31	622	8.59

). There is a clear inverse relationship between the distance of the region from the river and the number of rural settlements. From 2000 to 2019, the number of rural settlements in all regions increased, but within a distance of <1000 m from the rivers, objects increased significantly.

(2)

Relationship between rural settlements and the oasis margin

We investigated the relationship in the location between rural settlements and oases in 2000, 2011, and 2019. In the Shiyang River Basin, except for its southeastern part, most rural settlements were distributed within 4 km of an oasis, although they were mainly concentrated in zones that are farther from the margins of oases (Figure 10). The rural settlements in the research area are greatly limited by the conditions of the oasis, and their trend towards the central area of the oasis is significant.

4.2.2. Socio-Economic Factors

(1)

Relationship between rural settlements and population

This paper statistically analyzes the changes in rural population size from 2000 to 2019 in each county and region within the basin. From 2000 to 2019, the urbanization rate of the Shiyang River Basin increased from 21.94% to 48.86%. It is observed that the registered population in rural areas was 1.6864 million, while the rural resident population was 1.1676 million in 2019. Urbanization promotes the unprecedented expansion of urban population size and spatial scale, industrial structure optimization drives labor flow, and higher education popularization pulls the population flow, which has an important impact on rural settlements. When affected by urbanization, the overall trend of changes in villager groups exhibits an increasing trend.

(2)

Relationship between rural settlements and the economic industry

The economic development and urbanization process of the Shiyang River Basin in the three periods are shown in Table 4. From 2000 to 2019, the total economic output of the Shiyang River Basin increased nearly eightfold (

Table 6. Industrial economy and urbanization process in the Shiyang River Basin.

Year	GDP /102 Million RMB	Proportion of Industrial Structure/%			Urbanization Rate/%
		First Industry	Second Industry	Tertiary Industry
2000	98.70	27.9	43.3	28.8	21.94
2011	461.45	15.7	59.3	25.0	35.84
2019	828.76	20.3	35.7	44.0	48.86

), and the level of social and economic development gradually improved. As observed in the changes in the economic industrial structure over the years, the proportion of the primary industry has always been the lowest, while there have been changes in the structure between secondary and tertiary industries. The economic industrial structure of the basin underwent adjustments in the second period, and the tertiary industry began to serve as the economic backbone. The development of the overall economic structure of the basin is bound to affect the adjustment of the rural economic structure. Currently, the land-use types mainly include residential quarters, tourist resorts, etc. In particular, with the development of tourism in Zhangye City and Wuwei City, tourist facilities and entertainment venues in rural areas have increased explosively.

4.2.3. Locational Factor

Based on their distance from the nearest main road—national, provincial, and rural roads—four classes of rural settlements were grouped: those <1000 m, 1000–2500 m, 2500–4000 m, and >4000 m from a road. As Figure 11a shows, in 2019, rural settlements were mainly distributed within 4 km of a main traffic road; however, from 2000 to 2019, it is evident that the number of rural settlements decreased with increasing distances from roads. The number of rural settlements is inversely proportional to the distance from the main road. Among them, the area within 1000 m of the main road has the densest rural settlements, accounting for 52.3%, 52.7%, and 52.8% in the three periods. Exploring the details with respect to the time scale, the number of rural settlements within 1000 m of the main road has increased, while the rest of the area shows a decreasing trend.

In addition, policies are the leading factor that directly affects changes in rural settlements. By checking government gazettes in the region, it was observed that the Shiyang River Basin has caused some expansion of rural settlements due to urban–rural integration, while, at the same time, due to policies such as relocation, some rural settlements have disappeared [45].

(1)

Poverty alleviation and immigration

In 2007, the “Shiyang River Basin Key Governance Plan” proposed ecological migration as an important measure that can solve the deterioration of the ecological environment. The policy of ecological migration, including resettlement outside the county, resettlement within the county, and self-employment resettlement, greatly increased the enthusiasm of villagers with respect to building new houses and caused significant changes in rural settlements. Before 2010, 10,500 people in the northern part of the Minqin Lake area, which has a poor natural environment with no guarantee of production and living conditions, were resettled via ecological migration. From 2011 to 2019, 13,500 people in the Qilianshan Mountain Water Source Forest Reserve were resettled. Since 2011, the resettled residents in settlements have been continuously relocating to the eastern desert area of Liangzhou District and the Huanghuatan Township of Gula County. From 2011 to 2020, poverty alleviation relocation in Zhangye City involved Linze, Gaotai, Shandan, and Minle counties, and it promoted the employment of resettled residents, social governance, and subsequent industrial development, effectively changing the spatial and temporal pattern of rural settlements in the basin.

(2)

Integration of urban and rural development

Wuwei City proposed and implemented the strategy of urban–rural integration development in 2010, implemented new rural construction, and built 51 new rural resettlement sites, such as the new Huanghuatan Ganen and Yangguang villages in Gula County. The pilot urbanization plan was carried out in Huangyang Town of Liangzhou District and Huangyangchuan Town of Gula County, initially building a distinctive urbanization demonstration zone, which greatly changed the infrastructure and village appearance of rural areas. The construction of Jinse Road expanded the construction space of new rural communities, and new rural community communities were built along the road. Driven by government policies, rural settlements continuously evolved while simultaneously disappearing (as shown in Figure 12).

In summary, the temporal and spatial evolution of rural settlements is influenced by a variety of factors. Under the constrained pathway relative to natural geographical constraints, problems related to poor survival environments and underlying resource conditions directly hinder the development of countryside settlements, resulting in abandoned settlements. Under the induced pathway of socio-economic changes, rural industrialization, urbanization, population growth, and improved production technology promote the improvement of the regional economic level, and new constructions and reconstruction operations promote the expansion and evolution of rural settlements. Under the technical pathways relative to policies, preferential policies promote population and capital flows and rural development, curb the emptying of rural settlements and the waste of land resources, and establish an important foundation for the further sustainable development of countryside settlements.

The three paths interfere with and constrain one another to jointly promote the development of rural settlements within the basin (Figure 13). The ultimate goal is to create a reasonable layout of the countryside, improve the infrastructure system, and improve the rural ecological environment and sustainable development of the rural economy.

4.3. Optimization of Spatial Pattern

Obviously, the most important natural factor affecting the spatial distribution of rural settlements in the Shiyang River basin is water resources. Nevertheless, the impact of policies on the spatial distribution of rural settlements cannot be overlooked, especially in 2011–2019. The requirements of the following policies produced impacts within the area: “Guiding the development of rural classification”, “Classification and orderly promotion of village planning and management” in the “Zhangye City’s General Land Space Plan (2021–2035)”, and “Agglomeration improvement, suburban integration, characteristic protection, and relocation”. In combination with the distribution of water resources in the basin, the optimization type of rural settlement spatial layouts in the basin is divided into four categories: urbanization classes, development classes, control classes, and migration classes.

(1) Urbanization classes: Some rural settlements are confronted with new mergers and restructuring processes. In the case of the Shiyang River Basin, it is possible to incorporate the economic and industrial development of the existing urban plan, the density of the population, and rural settlements that are closer to the main urban area as reserve zones for future urban development into the urban system. (2) Development classes: Outward from the city’s center, resources are rich, ecological resilience is high, and areas closer to water sources are settlements. In the future, a part of the rural settlements will be used as an outflow zone for urban economic development, and it will experience the outflow benefits of the economic development of the city; moreover, it can be used to undertake the development of industries that are not suitable for urban development, and it can serve as an area for the settlement and migration of rural settlements. (3) Control classes: As a buffer for rural development and ecological coordination, this part of the future rural residential areas should be carefully developed, retaining its size and controlling the boundaries of a village’s development. (4) Migration classes: Projects located in the Qlian Mountain National Park, which is ecologically sensitive and contains highly sensitive areas, continue to develop and cause immeasurable losses to the ecological environment.

4.4. Limitations and Prospects

The findings of this study can provide a certain reference for the future development of rural settlements in dry inland river basins. Certainly, our ideas are not only applicable to arid inland river basins but also to other regions. At the same time, suggestions for developments could provide assistance with respect to the development and construction of any rural settlement that relies on rivers. The findings provide an empirical and methodological way forward toward the generation of new observations and provide insights into growth dynamics and governance conditions, particularly with respect to rapidly changing rural settlements.

This study could perform better in terms of further improvements and expansion. The data used in this paper vary from year to year, and different resolutions of the image recognition results inevitably produce some spatial deviation. These deviations could lead to minor changes in rural settlements at the edge of the pixel. In addition, only a handful of representative natural and social factors have been selected, with a certain degree of subjectivity. The effects of natural disasters were not considered, especially those in the Shiyang River Basin, which is one of China’s largest sandstorm sources. Moreover, there is a lack of in-depth examinations and quantitative analyses of the driving mechanisms, which was caused by unobtainable detailed data. In the future, we should improve the fine-tuning and dynamics of the indicators, build an evaluation system, and ensure the scientific, rational, and operational feasibility of optimizing the layout of rural settlements.

5. Conclusions

The combination of remote sensing and spatial analysis methods can effectively reveal the relationship between the distribution and changes in rural settlements and various environmental factors, and it is an effective method for exploring the formation mechanism of the spatial distribution pattern of rural settlements. The rural settlements in the Shiyang River Basin are undergoing various degrees of expansion, and the overall pattern is developing toward intensification. The degree of change is more severe in immigrant and urbanized areas than in other areas. Different environmental factors have different effects on the spatial distribution and changes in rural settlements. In terms of the natural environment, topography and water sources are the natural basis of the spatial distribution of rural settlements. In terms of the social and economic environments, improving transportation conditions has a certain attraction and clustering effect on rural settlements. The expansion of towns and cities has a significant radiation effect on the distribution of rural settlements, with a closer relationship between towns and rural settlements. The changes in settlements also have a reverse effect on the environment, changing and damaging the ecological environment while improving the quality of the social and economic environment. The loss of rural settlements in fragile ecological environments not only reduces the development and utilization of water resources, but also introduces new challenges to regional ecological environment protection and desertification control. Therefore, it is necessary to optimize the spatial pattern of settlements in combination with local planning needs within the watershed. We propose some policy suggestions, such as merging and reorganizing residential areas in the reserve zone, developing livable areas to accommodate transferred industries, and controlling and contracting ecological buffer zones. A more comprehensive quantitative evaluation system can ensure the scientific, reasonable, and operational feasibility of optimizing the layout of rural living conditions. This is aimed at improving the relationship between human beings and land, increasing resource utilization rates, promoting long-term ecological security, and promoting the sustainable development of the region.