Cellular automata (CA) has been frequently used to investigate the logical nature of self-reproducible systems and simulate the evolution of complex geographical phenomena such as urban expansion. The core of cellular automata is to define transition rules. Traditionally, approaches for defining the transition rules of cellular automata had difficulty to balance the interpretability, accuracy, and convenience. This article presents a new cellular automata model for simulating urban expansion based on random forest algorithm. The proposed model extracts CA transition rules of urban expansion by introducing random factors in training samples and candidate spatial variables that split nodes during the multiple decision trees building process. One significant advantage of our approach is that it can be easily adopted for parallel implementation and has high prediction accuracy and tolerance to random factors in urban expansion. Another strength of the proposed approach is that it can estimate out-of-bag errors to obtain model parameters quickly and measure the importance of spatial variables and explain the contribution of each variable in urban expansion. The model was applied to simulate urban expansion in Foshan City, Guangdong Province. We used the urban land change of 1988 and 2000 as the dependent variable and the spatial variables as the independent variables to construct the CA model based on random forests, then simulate and predict urban expansion of 2000 and 2012. The results show that random forest model can improve the simulation and prediction accuracies by 1.7% and 2.6%, respectively, when compared to the logistic regression model commonly used in CA simulation. This suggests that random forest model is superior for modeling complex nonlinear urban evolution. Urban expansion of Foshan City in 2024 was also predicted according to its urban development trend. Through measuring the importance of some spatial variables that affect urban expansion, we found that distance to national roads and to the city center are the two most important spatial variables for urban expansion simulation in Foshan City.
Using space syntax, kernel density estimation, and Pearson correlation analysis methods, this research combines the segment lines model of space syntax and Baidu POI data to explore the traffic network point of services location choice in the central city area of Beijing. The results indicate that (1) the traffic network is well connected, accessible, and dense in east and north, relatively poorly connected and sparse in west and south, and most poorly connected in southwest. The overall network shows the spatial pattern of annular agglomeration in the inner city and polycentric distribution in the periphery. (2) Compared to traffic network with high connectivity NACH(Normalised Angular Choice), traffic network with high level of integration NAIN (Normalised Angular Integration) has stronger level of point of services location choice and is more likely to induce agglomeration of services.(3) Traffic network point of services location choice is clear but the point levels of different services are remarkably different at different scales. Therefore, the services are divided into four types, including point at the neighborhood scale; point at the street scale; point at both scales, and nonpoint. This research provides some reference that facilities the understanding of services agglomeration and helps to improve services and traffic network planning and decision-making for optimizing urban structure of the central city area of Beijing.
Under the new normality of China's economic development, service industries are becoming the strong driver for maintaining and advancing growth. Urban agglomerations are an important spatial carrier of China's service industries. The development level of core cities' service industries is pivotal for the comprehensive competitiveness of these cities and a symbol of modernization degree in urban agglomerations. This article uses the entropy method, location quotient, and external service flow model to analyze the characteristics of services specialization, external function, and development level of 32 core cities of China in 2003 and 2012. Conclusions can be drawn as follows: Territorial division of service industries was clear in these core cities. The division degree of producer services was much higher than consumer and public services. Internal differences of producer services' exhibited an increasing trend. The core cities' external service ability showed strong correlation with city scale. The majority of the core cities were predominantly producer service-oriented, and fewer core cities were consumer service-oriented. Spatial concentration of external function of the core cities' service sectors was clearly different. Overall, spatial concentration of producer service was higher than that of consumer service. The spatial differentiation of services development level of the core cities was significant—core cities of national urban agglomerations generally had higher levels of services development than the regional and local urban agglomerations.
Urban livability is an essential science issue, which is closely related to residents' life qualities. Consequently, studies on the influencing mechanism of residents' livability satisfaction in an urban transitional period are of great practical significance in Beijing, which facilitate the understanding of residents' residential environment demand and promote construction of a livable city. Based on a large-scale questionnaire survey of livable city of Beijing in 2013, factor analysis was applied to extract perception factors of residents' livability satisfaction. Then we used the geographical detector method to reveal the influence mechanism of livability satisfaction from the perspective of livable satisfaction's perception factors and individual and family characteristics. The results show that: (1) Residents' livability satisfaction in Beijing is mainly consisted of seven dimensions, including pollution and noise factor, residential facilities factor, social environment factor, convenient transportation factor, natural environment factor, safety factor and leisure activity factor. (2) Livability satisfaction is jointly affected by perception factors of livability satisfaction and individual and family characteristics, but the former has stronger impact than the latter. Pollution and noise, natural environment, and convenient transportation are key factors among the livability perception factors. (3) With disparities between individuals and families, livability satisfaction is relatively consistent with its perception factors. This means that lower livability satisfaction usually occurs along with lower scores of livability perception. Moreover, residents of outer suburbs, rental housing, having migration experience, male, 30~39 years old of age, with high school degree, family income below 5000, and nonlocal population registration have much lower livable satisfaction.
Land-use change is a dynamic process through which human and natural systems interact and the link between social economic activities and ecological processes, and plays an important role in the understanding of human-environment relationship. It has important implications for land management, therefore exploring the many factors influencing land use and land cover has been the focus of scientific study across multiple disciplines, locations, and scales. Simulation of land use spatial pattern under different scenarios may reveal the driving factors of regional land-use change spatially, which is an important way to analyze future land-use demand. In this study, using Beijing as an example we examined characteristics of land-use change and simulated future land-use demand by a CLUE-S model, and identified related driving factors using a Logistic model combine with land use data extracted from remote sensing images in 1985, 2000, and 2010. We simulated the land use requirements of Beijing in 2020 under different scenarios based on economic and social development, land use planning, resources and ecological protection and other conditions. The results are as follows: (1) different driving factors resulted in some different changes in various land-use types. Transportation and socioeconomic factors may have played an important role in the conversion of land-use types especially through affecting accessibility to the administrative center and level of urbanization. The effect of topographic factors, slope in particular, on land-use types is clear, with higher slope conducive to forest land and lower slope for other land-use types.(2) Consistency between the projected and actual land-use situations is high—the kappa index is 0.87. (3) Comparison of simulated land-use changes under various scenarios shows that future land use differs under different scenarios. Spatial pattern of land use in Beijing is optimized to a certain degree under the ecological protection and water resources conservation scenario, which promotes rational expansion of construction land. Under the natural development scenario land use should be based on the Land Use Plan to ensure the sustainability of land resources by setting urban growth boundary and so on.
Administrative divisions are the basis of local administration of countries and important national system of local governance. Sound administrative division contributes to efficiency of administration and stability of governments. Therefore, administrative division has received wide attention around the world. Administrative regions are divided into regional-type and city-type, where regional-type administrative region covers mainly rural areas while city-type administrative region is predominately urban. The setting-up and adjustment of city-type administrative regions are inevitable requirements of urban development. City-type administrative regions include not only all levels of cities such as municipalities, prefecture-level cities, and county-level cities, but also municipal districts, towns, and neighborhoods. Based on the analysis of the concept and connotation of city-type administrative region, this article focuses on the evolution process, setting-up modes, levels, and jurisdiction of city-type administrative regions. It also summarizes research methods of city-type administrative regions in China and abroad, which gradually changed from qualitative to quantitative. Formed city is the main kind of city-type administrative regions. From a global perspective, modern city originated during the British Industrial Revolution. The emergence of formed city was marked by the "City of Autonomy" enacted in 1835. Formed city developed from one hundred years ago in China and has had an extremely complex process of evolution. Before the reform and opening up of the late 1970s, the speed of development was slow but after that time, the number of cities increased rapidly. In 1997, the State Council called off the conversion of counties to cities, and the number of cities stopped growing. Cities abroad can be as small as having a few thousand residents but in China, given its large population, a city must have at least 80,000 people. Many Chinese towns have more than 100,000 residents. Therefore, some scholars have recommend reform of "town to city" conversion. Cities in Western countries in general have equal legal status, but Chinese cities have a four-level administrative hierarchy: formed cities are divided into provincial-level, sub-provincial-level, prefecture-level, and county-level cities. Jurisdiction area of cities in Western countries is small but in China the area is large. This article reviews the characteristics and shortcomings of city-type administrative division research in China and abroad. Studies suggest that setting up city-type administrative regions based on national and local conditions, innovating the management system in metropolitan areas, and reducing administrative levels of cities should be the direction for future reform. At last, this article examines prospects of future research on city-type administrative division.
Based on the panel data of R&D input and the urbanization level from 2002 to 2012 in 30 provinces, the effect of China's R&D input on the urbanization level was analyzed by using the root test, co-integration test, and error correction model. Results show that: (1) There is a long-term stable relationship between China's R&D input and urbanization level. The impact of technological innovation on the level of urbanization, however, is still weak. In addition, there exist great gaps between the input-output elasticity of R&D in different provinces. (2) R&D personnel input-output elasticity is generally higher than that in R&D funds, which indicates that R&D personnel plays more important roles in urbanization of China. (3) The effect of R&D input on urbanization is related to the level of urbanization. Meanwhile, the coordination degree of R&D input and urbanization is also the key factor that influences R&D input and output. (4) The results of the error correction model indicate that there is a hysteresis effect in the influence of R&D short-term input on urbanization.
Disaster recovery is a process that constantly changes over time. Using the 2008 Wenchuan Ms8.0 earthquake in Sichuan, China as a case of catastrophe, and the 2009 Yao'an Ms6.0 earthquake in Yunnan, China as a case of small and medium-sized disaster, a random household survey was conducted in the disaster-stricken rural areas. Based on 1094 valid questionnaires and the derived recovery curves and by comparing the effect of different disaster intensities on the process of restoration and reconstruction, this study examined how the earthquake-stricken families recovered over time. The findings of the research show that the process of household recovery clearly had different phases, and can be divided, according to the recovery rate, into emergency phase, early phase, middle phase, and late phase. A catastrophe differed from a small or medium-sized disaster in terms of recovery duration and recovery process: for a small or medium-sized disaster, the process of household recovery lasted 20 months for a severely affected family and 18 months for a moderately affected family, which means household recovery was relatively short. For a catastrophe, the process of household recovery lasted relatively long. The effect of the degree of influence on recovery process was increasingly more prominent. Recovery lasted 68 months for a moderately affected family from a catastrophe, while it takes one year longer for a severely affected family is about one year more, and thus become more difficult over time.
Rural transformation is a leading human process accompanying urban-rural development and the transformation of economic development mode, under the background of rapid urbanization in the past few decades in China. In the study of rural transformation process, spatial reconstruction mechanism and optimal rural spatial reconstruction are of great significant. Using the theory of human-nature relationship in the territorial system, the process of rural transformation and interaction mechanism between this process and land use transformation should be analyzed in future studies to examine the mechanism of rural spatial reconstruction and the pattern of optimal spatial reorganization that affect the process of rural transformation. The integration of remote sensing and GIS and sampling survey and modeling methods should be used to: (1) analyze the process and dynamic patterns of rural transformation and its geographical types; (2) explore the collaborative mechanism between rural transformation and land use transformation; (3) establish models for optimal spatial reorganization that affects rural transformation; (4) identify typical patterns of rural transformation and land use allocation that affect rural spatial reconstruction indifferent functional types; and (5) systematically analyze the mechanism of different rural development patterns and the main regional geographical features. Combined with empirical studies, this research compares and analyzes typical development patterns in terms of the evolution process and interaction mechanism among various factors, to explore geographical patterns and methods for rural spatial reconstruction.
Measurement of stable water isotopes (δ2H and δ18O) at the watershed scale can provide a diagnostic indication for hydrological processes and water cycling of the watershed. Kinetic isotope fractionation associated with below-cloud secondary evaporation can significantly affect the observed precipitation isotopic composition and local meteoric water line. Stable isotope composition of δ2H and δ18O in precipitation was investigated in the Haihe River Basin, northern China based on observations at seven stations from July 2012 to January 2013. Linear regression between δ2H and δ18O values of rain samples yielded a correlation equation of δ2H=7.19δ18O-0.74, which is significantly different from that based on snow samples (δ2H=8.42δ18O+15.88). Due to the influence of below-cloud secondary evaporation on rain isotopes, the slope and intercept of correlation between δ2H and δ18O for small amount rainfall samples (<5 mm) progressively decreased with decreasing precipitation, varying from 6.73 to 7.61 and 5.28‰ to -11.04‰, respectively. Correlation of isotope values with local temperature, relative humidity, and precipitation amount provides evidence that small amount rainfall samples underwent secondary evaporation accompanied by mass dependent isotope fractionation during their descent from the cloud base to the ground. Compared to sites in the plain areas, rain stable isotopes at sites in the mountainous areas tended to be influenced by below-cloud secondary evaporation due to the dry atmosphere caused by the rain shadow effect. Since about half of the precipitation events in the observation period were rain samples with amount less than 5 mm, below-cloud secondary evaporation had an important influence on isotope composition of precipitation in the Haihe River Basin. Our study suggests that analysis of isotope composition of individual rainfall events can provide some valuable insight into below-cloud secondary evaporation effect, which is masked to a large extent by analysis of monthly precipitation isotope data.
Water conservation capability of the loess plateau increased significantly since the Grain for Green project started. At the same time, the amount of runoff significantly reduced, resulting in the question of how to balance the water conservation needs and the need for water supply in the region. In order to explore causes of the runoff reduction and provide support for balancing different needs for water at different scales, the variation of precipitation and vegetation as well as the scale effect of its impact on runoff in northern Shaanxi loess hilly-gully region during 2006 to 2011 were examined. The results show that: Precipitation of flood season gradually reduced from southeast to northwest in the study area. During the research period, vegetation restored significantly, nearly 80% of vegetated area was improved, especially in areas with poor vegetation conditions. Precipitation and runoff had significant positive correlation. With the increase of sub-watershed area,the correlation of precipitation and runoff increased, which shows a clear scale effect. The main reason was that land use structure above 15° of slope changed along with the spatial scale—vegetation types at steep slope tended to be homogenous and forest that has significant water conservation function decreased along with the increase in sub-watershed area. The correlation between NDVI and runoff was neither significant nor forming a clear relationship with the area of sub-watersheds, therefore the scale effect of vegetation influence on runoff was not obvious.
Precipitation has been a hot topic in the study of climate change and forecast. Heavy precipitation days reflect the condition of concentrated precipitation that has a great significance. Based on the daily precipitation data of 27 rainfall stations from 1957 to 2013 in Liaoning Province, trend coefficient method was used to calculate the trend of heavy precipitation days. Mutation was tested by the Mann-Kendall method. The cycle of heavy precipitation days was explored using the Morlet wavelet method. With the help of the ArcGIS software, this study analyzed the spatial distribution of heavy precipitation days. Both Kriging and inverse distance weighted (IDW) interpolation method were used in this study. The results show that heavy precipitation days decreased from the southeast to the northwest in Liaoning Province. Maximum days of precipitation were found in Dandong. Inter-annual variation of heavy precipitation showed a cycle of about 38 years. Overall the number of heavy precipitation days decreased, but not significantly. Only a few areas passed the significance test. In central Liaoning Province the change slightly lagged, and changes in western Liaoning Province were complex. Seasonally, a 38 year cycle was detected for the spring and the summer—the same as the annual cycle, and a 15 year cycle was found for the autumn. Inter-annual variation of heavy precipitation days showed a slight decreasing trend, and it slightly differed in different seasons.
Using turbulence data from August 2012 to July 2013 collected through the eddy covariance system over a closed lake in the Badain Jaran Desert, this study analyzed the characteristics of the turbulence variance, meteorological elements, and turbulent heat fluxes, and calculated the evaporation of the lake. The results show that: local circulation on the lake showed a diurnal variation; standard deviation of the three-dimensional wind speed versus stability(Z/L) was in agreement with the "1/3" law, and support the Monin-Obukov similarity theory. Each component of the radiation fluxes has clearseasonal and diurnal variations. Energy distribution differed during ice period and non-ice period. The lake transfers the stored energy to the atmosphere in the winter. Latent heat flux has clearseasonal variation and sensible heat flux did not change significantly, but both showed significant diurnal variations. Sensible heat flux has periodic positive and negative diurnal variation from March to August. It was mainly delivered from the lake to the atmosphere from September to January, and had no significant diurnal variation from December to February. Latent heat flux in the summer was much larger than in the winter. It was lowest in a day at 6:00-8:00 and reached the peak at 15:00-16:00. Sensible and latent heat flux diurnal variationsshow anegative correlation. Effective energy of the lake was mainly allocated for latent heat flux. The average evaporation was 3.97 mm/d and the cumulative evaporation was 1450±10 mm a year. The total evaporation exceeded the cumulative precipitation 20-folds during the same period. The existence of the lake not only relies on rainfall and runoff from the local area, but also groundwater replenishment. The results of this study may provide some references for the lake water circulation and sources in the BadainJaran Desert.
