Spatial Pattern Characteristics and Influencing Factors of

Research ArticleSpatial Pattern Characteristics and Influencing Factors of GreenUse Efficiency of Urban Construction Land in Jilin Province

Huisheng Yu 1 Ge Song 1 Tong Li1 and Yanjun Liu2

1School of Humanities and Law Northeastern University Shenyang 110169 China2School of Geographical Sciences Northeastern Normal University Changchun 130024 China

Correspondence should be addressed to Ge Song songgelaoshi163com

Received 20 March 2020 Revised 12 May 2020 Accepted 18 May 2020 Published 18 July 2020

Guest Editor Wen-Ze Yue

Copyright copy 2020 Huisheng Yu et al +is is an open access article distributed under the Creative Commons Attribution Licensewhich permits unrestricted use distribution and reproduction in any medium provided the original work is properly cited

How to explore the allocation and green utilization level of urban construction land resources has an important role in thesustainable development of the city Taking 47 counties and cities in Jilin Province as an example this paper evaluates the greenutilization efficiency of urban construction land (GUEUCL) in 2011 and 2015 by using the unexpected output super-SBM modeland explores the spatial-temporal differentiation characteristics and influencing factors of GUEUCL by using GIS and machinelearning methods +e results show that (1) the GUEUCL in Jilin Province is low mainly distributed in small- and medium-sizedareas with significant positive spatial correlation +e L-L concentration area is mainly distributed in the eastern region but thedegree of spatial concentration is small the spatial structure characteristics of the two periods are different and the spatialheterogeneity is large (2) the internal factor decomposition shows the impact of pure technical efficiency on the comprehensiveefficiency and the restriction ability is stronger than the scale efficiency that is to say the factors such as management andtechnology have a greater impact on the comprehensive efficiency (3) the relative importance of external factors has always beenranked as socioeconomic factors urban development factors and natural science and technology factors +is paper focuses onthe temporal and spatial characteristics of each county and city and the influencing factors which provides a certain valuereference for the pilot of ecological construction and the development of ecoenvironmental benefit economic system inJilin Province

1 Introduction

+ere are many theoretical research studies on urban landuse efficiency at home and abroad Economic geographysocial behavior and political economics all focus on theprogress of land use efficiency research from their respectivefields [1ndash5] In recent years the research focuses on theoptimal allocation intensive utilization and efficiencyevaluation of urban construction land [6ndash9] even theevaluation index driving factor analysis and improvementof management methods [10 11] Evaluation methods suchas analytic hierarchy process principal component analysisand data envelopment analysis have the defects of subjec-tivity and information compression which are not suitablefor efficiency evaluation Traditional data envelopmentanalysis ignores slack variables or unexpected output In

evaluation criteria the unexpected output is only a singlepollution index [6 7] and the driving factors are mostlytraditional statistical analysis [8 9]+e research scale coversnational [7] urban agglomerations [10] provincial [11]municipal [12] county [13] and village [14] +e countyscale is the grassroots unit of Chinarsquos administrative man-agement and the best scale of land use management andplanning [15]

With the rapid development of urbanization and in-dustrialization a large number of people gather in the cityand the urban space sprawl disorderly inefficient expansionproblems are serious +e urban development is also facedwith resource constraints environmental pollution andecological damage problems [16ndash23] ldquoGreenrdquo mainly em-phasizes the importance of green development Green de-velopment is an inevitable choice for countries in the world

HindawiComplexityVolume 2020 Article ID 5637530 12 pageshttpsdoiorg10115520205637530

to achieve sustainable urban development under the con-straints of resources and environment It is also an im-portant guiding ideology and main realization path forChinarsquos current social and economic transformation+erefore the green utilization efficiency of urban con-struction land is a new model of land use in the new periodof ecological civilization construction and is the extensionand expansion of sustainable land use Green utilizationefficiency is a comprehensive index of the effective utiliza-tion of resources in current natural economic technolog-ical and social conditions based on the principle ofminimizing resource and environmental consumption andmaximizing the comprehensive economic social and en-vironmental benefits of production Land is a complexsystem of socioeconomic-natural environment and theefficiency of traditional urban construction land use islimited from the socioeconomic perspective ignoring thegreen development concept of land use efficiency followingthe principle of coordination of economic social andecological benefits +erefore in the current background ofurban sprawl resource waste and environmental pollutionand ecological civilization construction require ldquogreen de-velopmentrdquo and the GUEUCL has attracted wide attentionof the state and society [24ndash33]

Based on theoretical perspective many literature worksstudy the efficiency of urban land use mostly from theinterests of all parties lack of the core elements of urbanland and the single subjectrsquos interest goal orientation isserious According to the theory of human land relationshipland use should not only maximize the economic benefitsbut also minimize the adverse effects on the geographicalenvironment and avoid exceeding the upper limit thresholdof the self-recovery of the geographical environmentAccording to the theory of scale economy resources need toeliminate the ldquoshort board effectrdquo by optimizing the allo-cation and improving the level of production technology toform an ideal scale economy rather than blindly investing inthe number of resources in the theory of production frontierand the most ideal situation for producers is to be able toreasonably allocate various resources and produce the mostvaluable products at a certain level of science and tech-nology the theory of sustainable land use mainly focuses onthe long-term interests of land users that is the economicsocial and ecological sustainability of land use In conclu-sion the concept of green utilization efficiency of urbanconstruction land is put forward which refers to the greensustainable development system of comprehensive evalua-tion of the quality and benefit of urban construction landunder the guidance of economic social and ecologicalbenefits under the factors of urban construction land laborcapital etc invested in a certain period of time

From the perspective of geography and ecology whichattach importance to regional and sustainable developmentthis paper uses pollution load index as the comprehensiveland pollution output evaluation of GUEUCL throughspatial analysis and machine learning methods combinedwith many factors in urban construction land use to identifyand quantify the spatial-temporal pattern and driving factorsof GUEUCL +e purpose of the study is as follows (1) to

interpret multitemporal images extract urban constructionland evaluate GUEUCL by unexpected output super-SBMand solve the problems of input-output variable relaxationeffective differentiation of multiple decision-making unitsand unexpected output (2) to explore the spatial correlationand heterogeneity of GUEUCL by using the methods ofspatial autocorrelation and spatial variation function (3) tomake use of the evaluation model decomposition andsupport vector machine regression model to analyze theinfluencing factors of GUEUCL which provide a valuablereference for formulating effective measures and improvingthe allocation and sustainable development of urban con-struction land resources

2 Data and Methods

Jilin Province is located in the middle of Northeast Chinabetween 121deg38primendash131deg19primeE and 40deg50primendash46deg19primeN covering anarea of 187400km2 (Figure 1) which belongs to the temperatecontinental monsoon climate It is an important grain basegoverning 47 counties and cities In 2015 the GDP was1406313 billion yuan the total population was 26621 millionand the per capita disposable income was 2490086 yuan Withthe rapid development of industrialization and urbanization inJilin Province large-scale urban expansion has begun resultingin low economic benefits of land and the problems of urbanconstruction land resource allocation and green utilizationbegin to appear which have a potential threat to the sustainabledevelopment of Jilin Province and the national food security

Jilin Province is the national ecological province con-struction pilot and 2015 is defined as the node of theecological construction development period in the outline ofthe overall planning of the ecological province constructionof Jilin Province [34] +is paper studies the green use ef-ficiency of urban construction land in Jilin Province whichcan not only test the actual results of the planning but alsoprovide reference experience and practical value for theecological construction improvement period in 2016ndash2030

+e data of this study include Landsat remote sensingimage DEM environmental pollution and socioeconomicdata Landsat remote sensing image is divided into twophases Envi software is used to extract urban constructionland including large medium and small cities and built-upareas above county and town factories and mines largeindustrial areas oil fields salt fields quarries and otherlands as well as transportation roads airports and specialland (since the output value of secondary and tertiary in-dustries and rural residential areas is excluded) +e in-terpretation accuracy of Landsat data extraction forconstruction land is more than 90 +e technical processand basis refer to ldquoNational Ecological EnvironmentMonitoring and Evaluation Technical Planrdquo DEM data useArcGIS software to calculate the slope Environmentalpollution data include wastewater chemical demand am-monia nitrogen sulfur dioxide nitrogen oxide and smokesocial and economic data include fixed asset investment unitemployees GDP total population hospital beds unit em-ployeesrsquo wages professional technicians patents and non-agricultural population Other indicators in this paper are

2 Complexity

calculated and converted with the above data See Table 1 forspecific data sources and descriptions

21 Unexpected Output Super-SBM SBM model is a non-radial and nonparametric model based on relaxation vari-ables which allows input and output variables to change indifferent proportions and measures efficiency by improvedaverage proportion At the same time in order to solve theproblem that the highest efficiency value of multiple deci-sion-making units is 1 a super-efficiency SBM model isproposed and in order to reflect resource saving pollutionreduction and economic growth etc based on the com-prehensive realization degree of the goal the pollution loadindex is introduced as the unexpected output to build theunexpected output super-SBM which is more in line withthe internal requirements of GUEUCL Specific evaluationindex is shown in Table 2 Pollution loadindex 020timesACODtimesCOD emissionstotal regional annualprecipitation+ 020timesANH3 times ammonia nitrogen emissionstotal regional annual precipitation+ 020timesASO2 times SO2emissionsregional area + 010timesASO1 times solid waste disposalregional area + 010timesAYFCtimes smoke (powder) dust emis-sionsregional area + 020timesANOtimes nitrogen oxide emissionsregional area (according to the national environmentalprotection of the Peoplersquos Republic of China Standard)

Suppose that there are nDMUs and each DMU requiresm inputs (x) to produce desirable outputs (yd) and unde-sirable outputs (yu) In this paper they can be denoted by thevectors x isinRm yd isinRr1 and yu isinRr2 then we can define the

matrices X Yd and Yu as X [x1 x2] Yd yd1 yd

n isinRr1xn and Yu yu

1 yun isin R

r2xn and SBM (SSBM) can beexpressed as follows [30]

min ρ 1 minus (1m) 1113936

mi1 wminus

i xik( 1113857

1 + 1 r1 + r2( 11138571113936r1s1w

ds yd

sk1113936r2q1w

uqyu

qk

st xik 1113944n

j1xijλj + w

minusi i 1 m

ydsk 1113944

n

j1y

dsjλj + w

ds i 1 r1

yuqk 1113944

n

j1y

uqjλj + w

us i 1 r2

λj gt 0 j 1 n

wminusi ge 0 i 1 m

wds ge 0 s 1 r1

wuq ge 0 q 1 r2

(1)

When ρ 1 in other words wminus 0 wd 0 and wu 0the DMUk is efficient In this paper we can define that the

46ordmN

44ordmN

42ordmN10ordmN

30ordmN

50ordmN 80ordmE 100ordmE 120ordmE 140ordmE 124ordmE 128ordmE 132ordmE

0 200km0 1000km

Urban construction land

Figure 1 Location of the study area (Jilin China)

Complexity 3

DMUk is efficient +e construction of a super-efficient SBMmodel with undesirable outputs is as follows [30 35]

minφ 1 minus (1m)1113936

mi1 wminus

i xik( 1113857

1 + 1 r1 + r2( 11138571113936r1s1y

dydsk + 1113936

r2q1y

uyuqk

st xge 1113944n

j1nekxijλj i 1 m

yd le 1113944

n

j1neky

dsjλj s 1 r1

yd 1113944n

j1y

uqjλj + w

us i 1 r2

λj gt 0 j 1 n

xge xk i 1 m

y dleydk s 1 r1

yuleyuk q 1 r2

(2)

22 Spatial Autocorrelation According to Toblerrsquos first lawof geography the closer the space distance is the greater thecorrelation between the attribute values is that is thestronger the spatial dependence In order to quantitativelymeasure the spatial dependence of GUEUCL this paperchooses global spatial autocorrelationMoranrsquos I Moranrsquos I isa widely used global spatial autocorrelation statistic and itscalculation formula is as follows [36]

I n 1113936

ni1 1113936

njnei wij xi minus x( 1113857 xj minus x1113872 1113873

1113936ni1 1113936

njnei wij 1113936

ni1 xi minus x( 1113857

2 (3)

where x is the mean of the observations at all n positions(areas) Wij is the spatial weight matrix xi and xj are theobservations at the spatial positions i and j +e range of

Moranrsquos I index is [minus 1 1] A value less than 0 indicatesnegative correlation a value equal to 0 indicates no corre-lation and a value greater than 0 indicates positivecorrelation

23 Spatial Variogram +e spatial variability function isintroduced to analyze the GUEUCL spatial variabilitySpatial variability function also known as semivariogram isa basic means to describe the randomness and structure ofregional variables and is an effective tool for the analysis ofspatial variability and structure +e calculation formula isgiven by [37]

c(h) 1113936

N(h)i1 Z xi( 1113857 minus Z xi + h( 11138571113858 1113859

2

2N(h) (4)

where Z(xi) and Z(xi + h) are the GUEUCL (i 1 2 3 N(h)) of Z (x) on the space units xi and xi + h respectivelyand (h) is the sample size of the segmentation distance h

Fractal dimension is another important parameter tocharacterize the variation function and its value is deter-mined by the relationship between the variation functionc(h) and distance h [38]

2c(h) h(4minus 2D)

(5)

where the fractal dimension D is the slope in the logarithmiclinear regression equation and represents the curvature ofthe variogram +e larger the value the higher the spatialheterogeneity caused by the spatial autocorrelation part thecloser the value is to 2 the more balanced the spatialdistribution

24 Support Vector Machine Regression Model Supportvector machine is a method of machine learning based onstatistical theory It integrates multiple technologies such asmaximum interval plane Mercer kernel convex quadraticprogramming and relaxation variables +e practicalproblems in the classification of small samples nonlinearityhigh digits and local minimum points are well solved +ebasic idea of regression is to map the data x to the high-dimensional feature space F through a nonlinear mappingΦand perform linear regression in this space +e calculationformula is as follows [39ndash43]

f(x) (ωΦ(x)) + b

Φ middot Rn⟶ F ω isin F

(6)

where b is the threshold and the high-dimensional featurespace corresponds to the nonlinear regression of the low-dimensional input space which eliminates the dot product

Table 1 Data sources and descriptions

Data Date Resolution (m) Source

Landsat 8 OLI_TIRS 2011 30 Geospatial Data Cloud2015 30 Geospatial Data Cloud

DEM environmental pollution 20112015 30Geospatial Data Cloud

State Environmental Protection AdministrationSocial economy 20112015 Jilin Statistical Yearbook

Table 2 Evaluation index of green utilization efficiency of urbanconstruction land

Standard layer Index

InputUrban construction landEmployees of the unit

Investment in fixed assets

Expected output Output value of the second and thirdindustries

Unexpectedoutput Pollution load index (PLI)

4 Complexity

calculations in high-dimensional spaces ω andΦ(x) SinceΦis fixed the sum of the empirical risks Remp affecting ω andω2 makes it flat in the high space +en the calculationformula is as follows [39ndash43]

R(ω) Remp + λ ω2 1113944

l

i1e f xi( 1113857 minus yi( 1113857 + λω

2 (7)

where l represents the number of samples e is a lossfunction and λ is an adjusted constant Minimize R(ω) toget ω expressed as data points

ω 1113944l

i1αi minus αlowasti( 1113857Φ xi( 1113857 (8)

where α and αlowast are solutions that minimize R(ω) Con-sidering equations (6) and (8) f(x) can be expressed asfollows [39ndash43]

f(x) 1113944l

i1α minus αlowasti( 1113857 Φ xi( 1113857Φ(x)( 1113857 + b 1113944

l

i1α minus αlowasti( 1113857k xi x( 1113857 + b

(9)

where k(xi x)=Φ(xi)Φ(x) is called the kernel function Itis the dot product of any symmetric kernel function thatsatisfies the Mercer condition corresponding to the featurespace

In order to better illuminate the models and algorithmsused in the research and the purpose of each research partthe method flowchart is drawn (Figure 2)

3 Results

31 Spatial and Temporal Characteristics of Green UtilizationEfficiency of Urban Construction Land In order to analyzethe distribution characteristics of GUEUCL units moreclearly according to the results of unexpected output super-SBM model evaluated the spatial pattern scale of GUEUCLis divided into three categories high medium and low bythe method of natural interruption +e spatial scale dis-tribution of GUEUCL in two time periods in Jilin Province isshown in Figure 3 +e GUEUCL of counties and urbanareas in Jilin Province is relatively low mainly distributed insmall and medium scale In 2011 high-efficiency unitsaccounted for about 851 medium efficiency units forabout 2553 and low-efficiency areas for about 6596 in2015 high-efficiency units accounted for about 1702medium efficiency units for about 3404 and low-effi-ciency areas for about 4894+e average level of GUEUCLin 2011 and 2015 was 036 and 043 respectively it has anoverall growth trend but the spatial scale distribution is stillshowing a significant imbalance

32 Spatial Autocorrelation of Green Use Efficiency of UrbanConstruction Land In order to explore the temporal andspatial characteristics of GUEUCL the spatial autocorre-lation analysis method is used to analyze its spatial de-pendence As can be seen from Figure 4 Moranrsquos I ofGUEUCL in 2011 and 2015 is 023 and 019 respectivelywith P value less than 005 indicating that it has significant

positive spatial correlation but has been at a low value thatis the degree of spatial agglomeration is not large

In order to further reveal the local spatial autocorrelationchange characteristics of GUEUCL the LISA value of greenuse efficiency of urban construction land in two years iscalculated (Figure 5) It can be seen from the figure that thespatial distribution pattern of GUEUCL in the two years isbasically stable L-L (low-low) cluster area is mainly dis-tributed in the eastern part of Jilin Province includingDunhua Antu County and Helong +e spatial distributionof H-H (high-high) cluster area is relatively scattered andthere is no cluster scale

33 Spatial Heterogeneity of Green Use Efficiency of UrbanConstructionLand According to Table 3 from 2011 to 2015the nugget coefficient decreased but all of them are smallvalues and the determination coefficient is relatively largewhich shows that the theoretical and practical situations fitwell which can reflect the interaction and linkage effect ofthe high and low GUEUCL which is basically consistentwith the results of spatial autocorrelation analysis From2011 to 2015 the step length of GUEUCL in Jilin Provinceremained unchanged but the range of variation increasedindicating that the spatial influence range of GUEUCL in-creased +is kind of spatial heterogeneity is influenced byboth structural and random components +e degree ofspatial variation caused by random components (such aspolicy and regulation elements and human factors) is lessthan that of structural components (such as social elementseconomic elements and resource and environment ele-ments) [44] which also lays a foundation for the analysis ofinfluencing factors below

According to Table 4 from 2011 to 2015 the quantile isfar away from 2 in all directions indicating that the spatialheterogeneity of GUEUCL is relatively large and tends to berelatively balanced with the time change but the spatialheterogeneity is still large which is basically consistent withthe results of spatial scale distribution From the quantiles ofall directions from 2011 to 2015 the east-west dimensionvalue is the smallest and the fitting degree is the best whichshows that the spatial heterogeneity of GUEUCL in east-westdirection in Jilin Province is the largest from 2011 to 2015but the spatial heterogeneity is slightly reduced It can beseen from Kriging interpolation 3D fitting chart that from2011 to 2015 (Figure 6) the peak value increased mainlydistributed in the northwest and in the east the structuredistribution is mostly flat In general the characteristics ofspatial heterogeneity and spatial scale distribution are ba-sically the same

34 Decomposition of Structural Factors of Green UtilizationEfficiency of Urban Construction Land +e green utilizationefficiency of urban construction land is the expression ofcomprehensive efficiency Pure technical efficiency is thetechnical efficiency after eliminating the influence of theorganizational scale of the decision-making unit Scale ef-ficiency is a measure of whether the decision-making unit isproducing at the optimal scale Pure technical efficiency and

Complexity 5

scale efficiency are also calculated by the unexpected outputsuper-SBM model In order to reflect the contribution ofpure technical efficiency and scale efficiency to

comprehensive efficiency a scatter plot (Figure 7) of therelationship between comprehensive efficiency and its de-composition efficiency can be drawn based on the DEA

GUEUCL

evaluation

Analysis ofGUEUCL

factors

Unexpected output super-SBM

GUEUCL

Spatial autocorrelation

Spatial variogram

Degree of spatialagglomeration

Spaceheterogeneity

The importance of the factors

Support vector machineregression model

Figure 2 Method flowchart

012ndash034034ndash077077ndash133

0 40 80 160km

N

(a)


0 40 80 160km

N

(b)

Figure 3 Spatial-temporal pattern of green utilization efficiency of urban construction land in different time periods (a) 2011 and (b) 2015

Moranrsquos I 0226775

ndash40

ndash24

ndash08

08

24

40

Lagg

ed G

UEU

CL 2

011

ndash24 ndash08 08 24 40ndash40GUEUCL 2011

(a)



ndash35

ndash21

ndash07

07

21

35

Lagg

ed G

UEU

CL 2

015

(b)

Figure 4 Moran scatter plot of green utilization efficiency of urban construction land at different time intervals

6 Complexity

results +e scatter points in the figure cannot be wellmatched with the 45-degree diagonal indicating that thecomprehensive efficiency is simultaneously affected by thetwo kinds of decomposition efficiency Because only a fewcities have reached the effective state of comprehensiveefficiency compared with pure technical efficiency morescale efficiency has reached effective state and the scaleefficiency is generally greater than pure technology effi-ciency so there are more dispersions determined by scaleefficiency and comprehensive efficiency +e points are lo-cated at the top and upper part of the scatter plot whichmakes these scatter dots deviate from the 45-degree diagonalmore than the deviation of pure technical efficiency and thepoints in the scatter plot constructed by the comprehensiveefficiency and pure technical efficiency are closer to the 45-degree diagonal +e results show that in the decompositionof the comprehensive efficiency of urban construction landgreen utilization the pure technical efficiency has a greaterimpact on the comprehensive efficiency the restriction onproduction capacity is greater than the scale efficiency andfactors such as management and technology have a greaterimpact on the comprehensive efficiency +e results showthat in the decomposition of the comprehensive efficiencyof urban construction land green utilization the puretechnical efficiency has a greater impact on the

comprehensive efficiency the restriction on productioncapacity is greater than the scale efficiency and factors suchas management and technology have a greater impact on thecomprehensive efficiency

35 Driving Factors of Green Utilization Efficiency of UrbanConstruction Land In order to alleviate the contradictionbetween man and earth we must understand the relation-ship between human activities and GUEUCL According tothe existing literature and research this paper studies thedriving factors of spatial differentiation of green use effi-ciency of urban construction land in Jilin Province fromthree aspects social economic factors natural science andtechnology factors and urban development factors [24ndash33]

Before the establishment of SVM regression modelfirstly preprocess the data and use the feature selectionmodel to remove the indicators that have little correlationwith the comprehensive efficiency In order to accuratelyclassify the test data we need to search for the best pa-rameters divide the training sample data into two take 70of them as the training set and 30 as the test set andcontinue cross-validation until the classification effect is thebest then select the RBF kernel function and establish theSVM regression model +e results (Table 5) show therelative importance of all independent variables in the formof percentages +e sum of the relative importance of allthese variables is 100 +e contribution rate of social andeconomic factors in 2011 is 057 which is greater than that ofurban development factors (031) and natural science andtechnology factors (012) +e contribution rate of social andeconomic factors in 2015 is 047 which is greater than that ofurban development factors (031) and that of natural scienceand technology factors (012) and elements of natural scienceand technology (021) +erefore socioeconomic factorshave the greatest impact on the comprehensive efficiencyWith the change of time the relative importance ranking of

Not significantHigh-highLow-lowLow-highHigh-low

N

0 40 80 160km

(a)


N

0 40 80 160km

(b)

Figure 5 LISA spatial agglomeration of green utilization efficiency of urban construction land at different time periods (a) 2011 and (b)2015

Table 3 Variability function fitting parameters of green utilizationefficiency pattern of urban construction land

Index Parameter 2011 2015Range a 291900 818000Nugget value C0 02340 02330Abutment value C0 +C 06150 09320Abutment value nuggetcoefficient C0C0 +C 03804 02500

Fitting the model Model Gaussian SphericalDecisive factor R2 0701 0600

Complexity 7

factors has not changed+emain reason for the distributionof factor proportions is that population density and landdevelopment intensity are relatively influential +e relativeimportance of socioeconomic factors has decreased and thereason may be that the implementation effect of theldquoOverview of Jilin Province Ecological Province

Construction Master Planrdquo has appeared In addition underthe background of the new economic normal more atten-tion has been paid to the quality of economic developmentwhich has slowed down the speed of social and economicdevelopment in the short term thereby reducing the impacton GUEUCL the relative importance of natural science and

Table 4 Variation dimension of green utilization efficiency pattern of urban construction land

YearsAll-round S-N (0deg) Ne-Sw (45deg) E-W (90deg) Se-Nw (135deg)

D R2 D R2 D R2 D R2 D R2

2011 1844 0677 1846 0146 1983 0002 1589 0821 1878 04062015 1900 0406 1942 0026 1861 0242 1614 0703 1978 0026

0155

0310

0465

0621

Sem

ivar

ianc

e

238179119090 35726900

Separation distance (h)

N-S W-E

ndash0155 ndash2037

(a)

0425

0566

0Separation distance (h)

Sem

ivar

ianc

e

0283

238179119090 3572690

0142 N-S W-E

ndash0353 ndash1762

(b)

Figure 6 Evolutionary graph of variation function of green utilization efficiency of urban construction land (a) 2011 and (b) 2015

00 05 10 15 20 25Comprehensive efficiency

00

05

10

15

20

25

Pure

tech

nica

l effi

cien

cy

(a)

02 04 06 08 10 12 1400Comprehensive efficiency

00

02

04

06

08

10

Scal

e effi

cien

cy

(b)

Figure 7 Analysis of the contribution of decomposition efficiency to total efficiency (a) Comprehensive efficiency-pure technical efficiency(b) Comprehensive efficiency-scale efficiency

8 Complexity

technology factors has increased and the main reason is thatthe land quality has suffered from increased stress and soilerosion and land desertification are becoming more andmore serious +e relative importance of each index haschanged significantly which needs specific analysis

351 Analysis of the Role of Socioeconomic FactorsSocial development is the basis of land use and land use is aspatial expression of human activity needs Populationdensity reflects the uniform distribution of regional spatialpopulation the increase of population density and thebearing of more population per unit of urban constructionland and the development of space utilization mode is morecompact and refined which promotes the growth ofGUEUCL +e relative importance of population densityand public service facilities in 2011 were 038 and 006 theconcentration degree of population distribution is the mostimportant factor influencing GUEUCL Even though therelative importance of population density dropped to 020 in2015 it is still the second only indicator of land developmentintensity and the relative impact of public service facilitieshas increased slightly As more people gather in urban spacethe impact of the compact development model on GUEUCLhas begun to decline

+e quality of economic development is the connotationreflection of land use efficiency and the quality of economicdevelopment is particularly important in the process of landuse GDP per capita is an important indicator for measuringthe level of economic development and the quality ofpeoplersquos lives Under the same technical conditions thegrowth of GDP per capita means that each unit of laborproduces more wealth and increases land output +e in-crease in the wages of employees in the unit means theincrease in peoplersquos remuneration for labor which indirectlyincreases the enthusiasm and efficiency of labor production+e relative importance of per capita GDP in 2011 and 2015was 009 and 007 and the relative importance of unitemployee wages was 004 and 013 respectively

352 Analysis of the Role of Natural Science and TechnologyElements +e natural construction environment is the

precondition of land use which determines the appropriateway and suitability of land use A good natural environmentcan reduce the cost of construction project investment andlater maintenance +e slope is the steepness of the surfaceunit Generally the slope of urban construction land shouldnot be greater than 25 degrees +e greater the slope thehigher the cost of construction and maintenance and therisk of natural disasters and soil erosion+erefore the slopewill directly affect GUEUCL +e land stress index is anevaluation of the extent to which land quality is subject tostress +e decline in land quality will definitely affect theability of land use and land output and stress the efficiency ofurban construction land use +e relative importance of theland stress index in 2015 was 008 indicating that the impactof land quality and land output stress began to appear in2015 but the impact was not significant

Science and technology are the primary productiveforces +e ability of scientific and technological innovationdrives productivity to increase changes the combination offactors and reduces environmental pollution Professionaland technical personnel can be regarded as the accumulationof human capital and the key indicator of technologicalprogress Human capital can create new technology orenhance the ability of existing technology and promote thegrowth of productivity +e number of patents is an im-portant indicator of innovation ability which optimizes theuse of land resources through the research and developmentof new technologies upgrading of production equipmentimprovement of organizational models and managementmethods From 2011 to 2015 the relative importance ofprofessional and technical personnel was 006 and 007 andthe relative importance of the number of patents was 006Science and technology and innovation ability are importantimpact indicators

353 Analysis of the Role of Urban Development Factors+e larger a city is the more it is able to attract own ele-ments and markets and more effectively create valueDifferent sizes of cities have different attractive forces whenattracting external factors Decentralization will also pro-mote the growth of urban construction land by promoting

Table 5 Relative importance of each variable

Element Factor layer Index 2011 2015

Socioeconomic factors

Foundation of socialdevelopment

Population density

057044 038

047027 020

Public facility 006 007

Economic development level Real GDP per capita 013 009 020 007Wages of employees 004 013

Elements of natural science andtechnology

Natural environment support Slope

0120 0

021008 0

Land stress index 0 008

Technological innovation ability Professional skill worker 012 006 013 007Patent number 006 006

Elements of urban development

Scale of urban construction Urbanization rate

031

008 0

031

006 0City size 008 006

Degree of urban development

Land developmentintensity 023

023025

025

Industrial outputstructure 0 0

Complexity 9

competition between local governments and ensuring fiscalrevenue From 2011 to 2015 the relative importance of thecity scale was 008 and 006 respectively +e different levelsof cities and governments have certain effects on attractinginvestment capacity and guiding land development andutilization

Reasonable urban development intensity can promoterational external communication and provide space forcarrying investment +e intensity of land development is aprocess of continuous development of urban constructionland population and economy Reasonable intensity of landdevelopment is conducive to the effective allocation of re-sources and intensive use of land resources From 2011 to2015 the relative importance of land development intensitywas 023 and 025 respectively With the increase of land useand the density of bearing materials the rational growth ofurban space development became more important toGUEUCL

4 Discussion

41 GUEUCLModel Evaluation Urban construction land isa space carrier for resource allocation Since urban space is acomplex socioeconomic-natural ecosystem this process willproduce environmental pollution while generating eco-nomic benefits In the past most of the studies used theconstruction land including rural residential areas and ur-ban built-up area in the statistical yearbook as urban con-struction land As the area of urban built-up areas generallydoes not include high-value construction land such asfactories and mines large-scale industrial areas oil fieldssalt fields and quarries the construction land includingrural residential areas cannot express the output value of thesecondary and tertiary industries well [8ndash10] +erefore it isvery important that the urban construction land interpretedby remote sensing is the basis for the scientific evaluation ofGUEUCL Secondly most scholars only evaluate GUEUCLfrom a single land pollution indicator as the undesiredoutput which is not enough to comprehensively express theundesired output of urban construction land to scientificallyevaluate GUEUCL [2225] +erefore this article compre-hensively considers the wastewater chemical requirementsammonia nitrogen sulfur dioxide nitrogen oxides soot andother pollutants in the land pollution as the pollution loadindex to express the unintended output of urban con-struction land and more scientifically and accurately eval-uates GUEUCL +e evaluation results are of great value tothe construction of ecological pilot and ecological envi-ronmental protection economic system in Jilin Province

42 Analysis of Influencing Factors Machine learning-basedurban and sustainable development research is rapidlyemerging In the study of the influencing factors of urbanresource allocation we try to use the machine learningmethod to analyze the more abundant driving factors ofGUEUCL which is more difficult and deeper than the case ofthe impact analysis of urban land use efficiency undernormal circumstances +is study only has 47 counties and

districts because of the multidimensionality and complexityof urban systems and the heterogeneity and timing ofGUEUCL machine learning can find unknown relation-ships hidden in complex data and explore deeper problemsthan traditional statistical models Support vector machine isa method of machine learning which has many uniqueadvantages in solving small sample nonlinear and insen-sitive sample dimension +erefore it is more scientific andaccurate to explore the causes of the spatial and temporalpattern characteristics of GUEUCL by using support vectormachine [36ndash40] Population agglomeration economiclevel land quality city scale and land development intensityare the specific driving indicators affecting GUEUCL It hasbeen generally believed that excessive pursuit of develop-ment intensity and scale will lead to the decline of land useefficiency because the dual effects of population agglomer-ation and scale economy are ignored In this paper theimpact of population agglomeration and development in-tensity is the largest among the specific driving indicators ofGUEUCL and under the ldquoceilingrdquo of construction land thegrowth of urban construction land is inevitable whichreasonably controls the disorderly spread of urban spacewhat is needed is a compact urban spatial structure andsmart and rational growth planning means redevelopinginefficient and idle land subjective value changes fromincremental planning to stock and reduction planning andagglomeration of urban population It is an effective way forthe green use of urban construction land in Jilin Province tocontinue to implement the policies and measures to attractpopulation growth and agglomeration in Changchun Jilinbig cities and the central urban agglomeration of JilinProvince and promote the common development of scaleeconomy and population agglomeration by dual meanswhich is also a reference experience and practical value forthe improvement period of ecological construction in2016ndash2030

5 Conclusions

In this study 47 counties and urban areas of Jilin Provinceare taken as the research units Unexpected output super-SBM spatial analysis method and support vector machineregression model are used to evaluate GUEUCL and analyzethe characteristics of spatial-temporal pattern and influ-encing factors

(1) From 2011 to 2015 the research arearsquos GUEUCLvalue is relatively low mainly in small- and medium-sized distribution with an overall growth trendHowever the spatial scale distribution still showsobvious nonequilibrium with significant positivespatial correlation but the degree of spatial ag-glomeration is small +e LL agglomeration area ismainly distributed in the eastern region with dif-ferent spatial structure characteristics and greaterspatial heterogeneity in the two periods

(2) +e internal factor decomposition shows that puretechnical efficiency has an impact on overall effi-ciency and its ability to restrict it is greater than scale

10 Complexity

efficiency Management and technology factors havea greater impact on overall efficiency the relativeimportance of external factors has been ranked associoeconomic factors urban development factorsand natural science and technology factors but therelative importance of socioeconomic factors hasdeclined and the relative importance of naturalscience and technology factors has increased

(3) +e specific performance of the impact is as followspopulation density is an important indicator ofimpact but the impact of urban spatial populationagglomeration on GUEUCL began to decline theeconomic conditions giving people better materialconditions can increasingly affect GUEUCL by in-creasing labor enthusiasm the governmentrsquos abilityto attract investment to guide land development andutilization has certain effects the rational growth ofurban space development has become increasinglyimportant to GUEUCL and has become the mostimportant impact indicator

In view of the research results improving GUEUCL canbe started from several aspects ① continue to play the roleof government policy regulation adhere to the concept ofgreen development put forward compact city and smartgrowth policy recommendations and shift planning to thevalue of stocks and reductions ② according to the func-tional positioning of different cities divide the developmentgoals of different GUEUCL adjust and optimize the spatialpattern of GUEUCL in Jilin Province pay special attentionto the low-lying cities around high-efficiency cities andstrictly prevent them from becoming refuges for the ex-tensive use of low-lying land in policies ③ continue todevelop the management and guidance role of ldquoOutline ofJilin Province Ecological Province Construction MasterPlanrdquo and implement the economic growth model of urbanclusters in central Jilin in-depth thoughts and measures forthe construction of green transformation and developmentzone in eastern Jilin

+ere are two main contributions of this study Firstlythe concept of GUEUCL is proposed and the pollution loadindex is introduced as the unexpected output of urbanconstruction land which is a new method to evaluate andmeasure GUEUCL Secondly based on the support vectormachine regression model the index system of influencingfactors is designed combined with the results of the modelthis paper explores the causes of the spatial-temporal patternof green use of urban construction land in Jilin Province andit is conducive to improving the green utilization level ofurban construction land in Jilin Province alleviating thecontradiction between human and land relations providingan academic basis for the land sector to formulate greendevelopment policies and providing reference for othersimilar regions around the world

Of course the study also has certain limitations It onlyexplores the impact of structural components on GUEUCLand it lacks randomness (policy and other factors) quan-titative research At the same time the support vectormachine regression model was used to analyze the

influencing factors Due to the limitation of data acquisitionthe analysis was performed only for a single time period andthe difference in influence over a long time period was notcompared +erefore the analysis of random componentsand temporal differences combined with panel data toexplore the influencing factors is the focus of subsequentresearch

Data Availability

+e data used to support the findings of this study areavailable from the first author upon request

Conflicts of Interest

+e authors declare that they have no conflicts of interest

References

[1] H Xie W Wang and Y He Research on Green UtilizationEfficiency of Land Resources in China EconomicManagementPress Beijing China 2016 in Chinese

[2] T C Koopmans and M Beckmann ldquoAssignment problemsand the location of economic activitiesrdquo Econometricavol 25 no 1 pp 53ndash76 1957

[3] M M Webber Explorations into Urban Structutre Phila-delphia University Press Philadelphia PA USA 1964

[4] D Harvey ldquo+e urban process under capitalism a frameworkfor analysisrdquo International Journal of Urban and RegionalResearch vol 2 no 1ndash3 pp 101ndash131 1978

[5] O Fisch ldquoOptimal allocation of land to transportation in anon-optimal urban structurerdquo Regional Science and UrbanEconomics vol 12 no 2 pp 235ndash246 1982

[6] Q Yang X Wan and H Ma ldquoAssessing green developmentefficiency of municipalities and provinces in China integratingmodels of super-efficiency DEA and malmquist indexrdquoSustainability vol 7 no 4 pp 4492ndash4510 2015

[7] Y Liu Z Zhang and Y Zhou ldquoEfficiency of constructionland allocation in China an econometric analysis of paneldatardquo Land Use Policy vol 74 pp 261ndash272 2018

[8] X Zhu Y Li P Zhang Y Wei X Zheng and L XieldquoTemporal-spatial characteristics of urban land use efficiencyof Chinarsquos 35mega cities based on DEA decomposing tech-nology and scale efficiencyrdquo Land Use Policy vol 88 ArticleID 104083 2019

[9] J Yang Y Yang and W Tang ldquoDevelopment of evaluationmodel for intensive land use in urban centersrdquo Frontiers ofArchitectural Research vol 1 no 4 pp 405ndash410 2012

[10] J Yu K Zhou and S Yang ldquoLand use efficiency andinfluencing factors of urban agglomerations in Chinardquo LandUse Policy vol 88 Article ID 104143 2019

[11] L Zhang L Zhang Y Xu P Zhou and C-H Yeh ldquoEval-uating urban land use efficiency with interacting criteria anempirical study of cities in Jiangsu Chinardquo Land Use Policyvol 90 p 104292 2020

[12] Z Xinhua P Zhang Y Wei et al ldquoMeasuring the efficiencyand driving factors of urban land use based on the DEAmethod and the PLS-SEMmodelmdasha case study of 35 large andmedium-sized cities in Chinardquo Sustainable Cities and Societyvol 50 Article ID 101646 2019

[13] S Zou and Z Zeng ldquoUrban land use structure characteristicsand efficiency measures in Hubei provinces and citiesrdquo ChinaReal Estate vol 24 pp 27ndash37 2018 in Chinese

Complexity 11

[14] C Shen M Li F Li et al ldquoEvaluation of rural land useefficiency in Fenshui town based on data envelopmentanalysisrdquo China Land Science vol 25 no 01 pp 16ndash21 2011in Chinese

[15] J Wang and X Guo ldquoScientific regulation of sustainable landuse at county level in Chinardquo Progress in Geographical Sci-ences vol 3 pp 216ndash222 2002 in Chinese

[16] L Tong S Hu and A E Frazier ldquoHierarchically measuringurban expansion in fast urbanizing regions using multi-di-mensional metrics a case of Wuhan metropolis ChinardquoHabitat International vol 94 Article ID 102070 2019

[17] W CuiOptimal Allocation of Urban Construction Land Basedon Ecological Efficiency Equation Jiangsu University PressZhenjiang China 2015 in Chinese

[18] XWang R Shi and Y Zhou ldquoDynamics of urban sprawl andsustainable development in Chinardquo Socio-Economic PlanningSciences vol 70 Article ID 100736 2020

[19] Z Kovacs Z J Farkas T Egedy et al ldquoUrban sprawl and landconversion in post-socialist cities the case of metropolitanBudapestrdquo Cities vol 92 pp 71ndash81 2019

[20] E J Kaiser D R Godschalk and F S Chapin Urban LandUse Planning University of Illinois press Urbana IL USA1995

[21] Y Feng and X Wang ldquoEffects of urban sprawl on hazepollution in China based on dynamic spatial Durbin modelduring 2003-2016rdquo Journal of Cleaner Production vol 242p 118368 2020

[22] D Chen X Lu X Liu and X Wang ldquoMeasurement of theeco-environmental effects of urban sprawl theoreticalmechanism and spatiotemporal differentiationrdquo EcologicalIndicators vol 105 pp 6ndash15 2019

[23] Y Jun L Weiling L Yonghua et al ldquoSimulating intraurbanland use dynamics under multiple scenarios based on fuzzycellular automata a case study of Jinzhou district DalianrdquoComplexity vol 2018 Article ID 7202985 17 pages 2018

[24] L Liang Y Yong and C Yuan ldquoMeasurement of urban landgreen use efficiency and its spatial differentiation character-istics an empirical study based on 284 citiesrdquo China LandScience vol 33 no 6 p 80 2019 in Chinese

[25] D Wang and X Pang ldquoResearch on green land use efficiencyof Beijing-Tianjin-Hebei urban agglomerationrdquo China Pop-ulation Resources and Environment vol 29 no 04 pp 68ndash762019 in Chinese

[26] B Hu J Li and B Kuang ldquoEvolution characteristics andinfluencing factors of urban land use efficiency differencesunder the concept of green developmentrdquo Economic Geog-raphy vol 38 no 12 pp 183ndash189 2018 in Chinese

[27] L Ma H Long K Chen S Tu Y Zhang and L Liao ldquoGreengrowth efficiency of Chinese cities and its spatio-temporalpatternrdquo Resources Conservation and Recycling vol 146pp 441ndash451 2019

[28] X Lu B Kuang and J Li ldquoRegional difference decompositionand policy implications of Chinarsquos urban land use efficiencyunder the environmental restrictionrdquo Habitat Internationalvol 77 pp 32ndash39 2018

[29] H Xie F Q Chen and G W YaoWang ldquoSpatial-temporaldisparities and influencing factors of total-factor green useefficiency of industrial land in Chinardquo Journal of CleanerProduction vol 207 pp 1047ndash1058 2019

[30] H Yu and W Wang ldquoExploring the spatial-temporal dis-parities of urban land use economic efficiency in China and itsinfluencing factors under environmental constraints based ona sequential slacks-based modelrdquo Sustainability vol 7 no 8pp 10171ndash10190 2015

[31] D Z Su ldquoGIS-based urban modelling practices problemsand prospectsrdquo International Journal of Geographical Infor-mation Science vol 12 no 7 pp 651ndash671 1998

[32] F Qiu Y Chen J Tan J Liu Z Zheng and X ZhangldquoSpatial-temporal heterogeneity of green development effi-ciency and its influencing factors in growing metropolitanarea a case study for the Xuzhou metropolitan areardquo ChineseGeographical Science vol 30 no 2 pp 352ndash365 2020

[33] S Li Z Ying H Zhang G Ge and Q Liu ldquoComprehensiveassessment of urbanization coordination a case study ofJiangxi province Chinardquo Chinese Geographical Sciencevol 29 no 3 pp 488ndash502 2019

[34] H Cheng Research on Land Remediation in Jilin ProvinceGeological Publishing House Beijing China 2018 inChinese

[35] W Lin B Chen L Xie and H Pan ldquoEstimating energyconsumption of transport modes in China using DEArdquoSustainability vol 7 no 4 pp 4225ndash4239 2015

[36] W R Tobler ldquoA computer movie simulating urban growth inthe detroit regionrdquo Economic Geography vol 46 no 1pp 234ndash240 1970

[37] A Getis and J K Ord ldquo+e analysis of spatial association byuse of distance statisticsrdquo in Perspectives on Spatial DataAnalysis pp 127ndash145 Springer Berlin Germany 2010

[38] C A Gambardella T B Moorman J M Novak et al ldquoField-scale variability of soil properties in central low a soilsrdquo SoilSci Soc Am J vol 58 pp 1501ndash1511 1994

[39] V VapnikFe Nature of Statistical LearningFeory SpringerScience amp Business Media Berlin Germany 2013

[40] V N Vapnik and A Y Chervonenkis ldquoOn the uniformconvergence of relative frequencies of events to their prob-abilitiesrdquo in Measures of Complexity pp 11ndash30 SpringerCham Switzerland 2015

[41] Y A Awad P Koutrakis B A Coull et al ldquoA spatio-tem-poral prediction model based on support vector machineregression ambient Black Carbon in three New EnglandStatesrdquo Environmental Research vol 159 pp 427ndash434 2017

[42] Z Huang H Chen C-J Hsu W-H Chen and S WuldquoCredit rating analysis with support vector machines andneural networks a market comparative studyrdquo DecisionSupport Systems vol 37 no 4 pp 543ndash558 2004

[43] Y-C Lee ldquoApplication of support vector machines to cor-porate credit rating predictionrdquo Expert Systems with Appli-cations vol 33 no 1 pp 67ndash74 2007

[44] G Song and Y Wang ldquoSpatiotemporal differentiation of landuse patterns in the Songnen High Plainrdquo Journal of Agri-cultural Engineering vol 32 no 18 pp 225ndash233 2016 inChinese

12 Complexity

to achieve sustainable urban development under the con-straints of resources and environment It is also an im-portant guiding ideology and main realization path forChinarsquos current social and economic transformation+erefore the green utilization efficiency of urban con-struction land is a new model of land use in the new periodof ecological civilization construction and is the extensionand expansion of sustainable land use Green utilizationefficiency is a comprehensive index of the effective utiliza-tion of resources in current natural economic technolog-ical and social conditions based on the principle ofminimizing resource and environmental consumption andmaximizing the comprehensive economic social and en-vironmental benefits of production Land is a complexsystem of socioeconomic-natural environment and theefficiency of traditional urban construction land use islimited from the socioeconomic perspective ignoring thegreen development concept of land use efficiency followingthe principle of coordination of economic social andecological benefits +erefore in the current background ofurban sprawl resource waste and environmental pollutionand ecological civilization construction require ldquogreen de-velopmentrdquo and the GUEUCL has attracted wide attentionof the state and society [24ndash33]

Based on theoretical perspective many literature worksstudy the efficiency of urban land use mostly from theinterests of all parties lack of the core elements of urbanland and the single subjectrsquos interest goal orientation isserious According to the theory of human land relationshipland use should not only maximize the economic benefitsbut also minimize the adverse effects on the geographicalenvironment and avoid exceeding the upper limit thresholdof the self-recovery of the geographical environmentAccording to the theory of scale economy resources need toeliminate the ldquoshort board effectrdquo by optimizing the allo-cation and improving the level of production technology toform an ideal scale economy rather than blindly investing inthe number of resources in the theory of production frontierand the most ideal situation for producers is to be able toreasonably allocate various resources and produce the mostvaluable products at a certain level of science and tech-nology the theory of sustainable land use mainly focuses onthe long-term interests of land users that is the economicsocial and ecological sustainability of land use In conclu-sion the concept of green utilization efficiency of urbanconstruction land is put forward which refers to the greensustainable development system of comprehensive evalua-tion of the quality and benefit of urban construction landunder the guidance of economic social and ecologicalbenefits under the factors of urban construction land laborcapital etc invested in a certain period of time

From the perspective of geography and ecology whichattach importance to regional and sustainable developmentthis paper uses pollution load index as the comprehensiveland pollution output evaluation of GUEUCL throughspatial analysis and machine learning methods combinedwith many factors in urban construction land use to identifyand quantify the spatial-temporal pattern and driving factorsof GUEUCL +e purpose of the study is as follows (1) to

interpret multitemporal images extract urban constructionland evaluate GUEUCL by unexpected output super-SBMand solve the problems of input-output variable relaxationeffective differentiation of multiple decision-making unitsand unexpected output (2) to explore the spatial correlationand heterogeneity of GUEUCL by using the methods ofspatial autocorrelation and spatial variation function (3) tomake use of the evaluation model decomposition andsupport vector machine regression model to analyze theinfluencing factors of GUEUCL which provide a valuablereference for formulating effective measures and improvingthe allocation and sustainable development of urban con-struction land resources

2 Data and Methods

Jilin Province is located in the middle of Northeast Chinabetween 121deg38primendash131deg19primeE and 40deg50primendash46deg19primeN covering anarea of 187400km2 (Figure 1) which belongs to the temperatecontinental monsoon climate It is an important grain basegoverning 47 counties and cities In 2015 the GDP was1406313 billion yuan the total population was 26621 millionand the per capita disposable income was 2490086 yuan Withthe rapid development of industrialization and urbanization inJilin Province large-scale urban expansion has begun resultingin low economic benefits of land and the problems of urbanconstruction land resource allocation and green utilizationbegin to appear which have a potential threat to the sustainabledevelopment of Jilin Province and the national food security

Jilin Province is the national ecological province con-struction pilot and 2015 is defined as the node of theecological construction development period in the outline ofthe overall planning of the ecological province constructionof Jilin Province [34] +is paper studies the green use ef-ficiency of urban construction land in Jilin Province whichcan not only test the actual results of the planning but alsoprovide reference experience and practical value for theecological construction improvement period in 2016ndash2030

+e data of this study include Landsat remote sensingimage DEM environmental pollution and socioeconomicdata Landsat remote sensing image is divided into twophases Envi software is used to extract urban constructionland including large medium and small cities and built-upareas above county and town factories and mines largeindustrial areas oil fields salt fields quarries and otherlands as well as transportation roads airports and specialland (since the output value of secondary and tertiary in-dustries and rural residential areas is excluded) +e in-terpretation accuracy of Landsat data extraction forconstruction land is more than 90 +e technical processand basis refer to ldquoNational Ecological EnvironmentMonitoring and Evaluation Technical Planrdquo DEM data useArcGIS software to calculate the slope Environmentalpollution data include wastewater chemical demand am-monia nitrogen sulfur dioxide nitrogen oxide and smokesocial and economic data include fixed asset investment unitemployees GDP total population hospital beds unit em-ployeesrsquo wages professional technicians patents and non-agricultural population Other indicators in this paper are

2 Complexity






1 yun isin R


min ρ 1 minus (1m) 1113936

mi1 wminus

i xik( 1113857

1 + 1 r1 + r2( 11138571113936r1s1w

ds yd

sk1113936r2q1w

uqyu

qk

st xik 1113944n

j1xijλj + w

minusi i 1 m

ydsk 1113944

n

j1y

dsjλj + w

ds i 1 r1

yuqk 1113944

n

j1y

uqjλj + w

us i 1 r2

λj gt 0 j 1 n

wminusi ge 0 i 1 m

wds ge 0 s 1 r1

wuq ge 0 q 1 r2

(1)


46ordmN

44ordmN

42ordmN10ordmN

30ordmN


0 200km0 1000km



Complexity 3



mi1 wminus

i xik( 1113857

1 + 1 r1 + r2( 11138571113936r1s1y

dydsk + 1113936

r2q1y

uyuqk

st xge 1113944n

j1nekxijλj i 1 m

yd le 1113944

n

j1neky

dsjλj s 1 r1

yd 1113944n

j1y

uqjλj + w

us i 1 r2

λj gt 0 j 1 n

xge xk i 1 m

y dleydk s 1 r1

yuleyuk q 1 r2

(2)


I n 1113936

ni1 1113936


1113936ni1 1113936

njnei wij 1113936


2 (3)




c(h) 1113936


2

2N(h) (4)



2c(h) h(4minus 2D)

(5)



f(x) (ωΦ(x)) + b


(6)













4 Complexity


R(ω) Remp + λ ω2 1113944

l

i1e f xi( 1113857 minus yi( 1113857 + λω

2 (7)


ω 1113944l



f(x) 1113944l


l


(9)



3 Results








Complexity 5



GUEUCL

evaluation

Analysis ofGUEUCL

factors


GUEUCL


Spatial variogram


Spaceheterogeneity





0 40 80 160km

N

(a)


0 40 80 160km

N

(b)



ndash40

ndash24

ndash08

08

24

40

Lagg

ed G

UEU

CL 2

011


(a)



ndash35

ndash21

ndash07

07

21

35

Lagg

ed G

UEU

CL 2

015

(b)


6 Complexity






N

0 40 80 160km

(a)


N

0 40 80 160km

(b)





Complexity 7






2011 1844 0677 1846 0146 1983 0002 1589 0821 1878 04062015 1900 0406 1942 0026 1861 0242 1614 0703 1978 0026

0155

0310

0465

0621

Sem

ivar

ianc

e

238179119090 35726900


N-S W-E

ndash0155 ndash2037

(a)

0425

0566


Sem

ivar

ianc

e

0283

238179119090 3572690

0142 N-S W-E

ndash0353 ndash1762

(b)



00

05

10

15

20

25

Pure

tech

nica

l effi

cien

cy

(a)


00

02

04

06

08

10

Scal

e effi

cien

cy

(b)


8 Complexity












Population density

057044 038

047027 020





0120 0

021008 0





031

008 0

031

006 0City size 008 006



023025

025


Complexity 9



4 Discussion




5 Conclusions




10 Complexity







Data Availability




References














Complexity 11
































12 Complexity






1 yun isin R


min ρ 1 minus (1m) 1113936

mi1 wminus

i xik( 1113857

1 + 1 r1 + r2( 11138571113936r1s1w

ds yd

sk1113936r2q1w

uqyu

qk

st xik 1113944n

j1xijλj + w

minusi i 1 m

ydsk 1113944

n

j1y

dsjλj + w

ds i 1 r1

yuqk 1113944

n

j1y

uqjλj + w

us i 1 r2

λj gt 0 j 1 n

wminusi ge 0 i 1 m

wds ge 0 s 1 r1

wuq ge 0 q 1 r2

(1)


46ordmN

44ordmN

42ordmN10ordmN

30ordmN


0 200km0 1000km



Complexity 3



mi1 wminus

i xik( 1113857

1 + 1 r1 + r2( 11138571113936r1s1y

dydsk + 1113936

r2q1y

uyuqk

st xge 1113944n

j1nekxijλj i 1 m

yd le 1113944

n

j1neky

dsjλj s 1 r1

yd 1113944n

j1y

uqjλj + w

us i 1 r2

λj gt 0 j 1 n

xge xk i 1 m

y dleydk s 1 r1

yuleyuk q 1 r2

(2)


I n 1113936

ni1 1113936


1113936ni1 1113936

njnei wij 1113936


2 (3)




c(h) 1113936


2

2N(h) (4)



2c(h) h(4minus 2D)

(5)



f(x) (ωΦ(x)) + b


(6)













4 Complexity


R(ω) Remp + λ ω2 1113944

l

i1e f xi( 1113857 minus yi( 1113857 + λω

2 (7)


ω 1113944l



f(x) 1113944l


l


(9)



3 Results








Complexity 5



GUEUCL

evaluation

Analysis ofGUEUCL

factors


GUEUCL


Spatial variogram


Spaceheterogeneity





0 40 80 160km

N

(a)


0 40 80 160km

N

(b)



ndash40

ndash24

ndash08

08

24

40

Lagg

ed G

UEU

CL 2

011


(a)



ndash35

ndash21

ndash07

07

21

35

Lagg

ed G

UEU

CL 2

015

(b)


6 Complexity






N

0 40 80 160km

(a)


N

0 40 80 160km

(b)





Complexity 7






2011 1844 0677 1846 0146 1983 0002 1589 0821 1878 04062015 1900 0406 1942 0026 1861 0242 1614 0703 1978 0026

0155

0310

0465

0621

Sem

ivar

ianc

e

238179119090 35726900


N-S W-E

ndash0155 ndash2037

(a)

0425

0566


Sem

ivar

ianc

e

0283

238179119090 3572690

0142 N-S W-E

ndash0353 ndash1762

(b)



00

05

10

15

20

25

Pure

tech

nica

l effi

cien

cy

(a)


00

02

04

06

08

10

Scal

e effi

cien

cy

(b)


8 Complexity












Population density

057044 038

047027 020





0120 0

021008 0





031

008 0

031

006 0City size 008 006



023025

025


Complexity 9



4 Discussion




5 Conclusions




10 Complexity







Data Availability




References














Complexity 11
































12 Complexity



mi1 wminus

i xik( 1113857

1 + 1 r1 + r2( 11138571113936r1s1y

dydsk + 1113936

r2q1y

uyuqk

st xge 1113944n

j1nekxijλj i 1 m

yd le 1113944

n

j1neky

dsjλj s 1 r1

yd 1113944n

j1y

uqjλj + w

us i 1 r2

λj gt 0 j 1 n

xge xk i 1 m

y dleydk s 1 r1

yuleyuk q 1 r2

(2)


I n 1113936

ni1 1113936


1113936ni1 1113936

njnei wij 1113936


2 (3)




c(h) 1113936


2

2N(h) (4)



2c(h) h(4minus 2D)

(5)



f(x) (ωΦ(x)) + b


(6)













4 Complexity


R(ω) Remp + λ ω2 1113944

l

i1e f xi( 1113857 minus yi( 1113857 + λω

2 (7)


ω 1113944l



f(x) 1113944l


l


(9)



3 Results








Complexity 5



GUEUCL

evaluation

Analysis ofGUEUCL

factors


GUEUCL


Spatial variogram


Spaceheterogeneity





0 40 80 160km

N

(a)


0 40 80 160km

N

(b)



ndash40

ndash24

ndash08

08

24

40

Lagg

ed G

UEU

CL 2

011


(a)



ndash35

ndash21

ndash07

07

21

35

Lagg

ed G

UEU

CL 2

015

(b)


6 Complexity






N

0 40 80 160km

(a)


N

0 40 80 160km

(b)





Complexity 7






2011 1844 0677 1846 0146 1983 0002 1589 0821 1878 04062015 1900 0406 1942 0026 1861 0242 1614 0703 1978 0026

0155

0310

0465

0621

Sem

ivar

ianc

e

238179119090 35726900


N-S W-E

ndash0155 ndash2037

(a)

0425

0566


Sem

ivar

ianc

e

0283

238179119090 3572690

0142 N-S W-E

ndash0353 ndash1762

(b)



00

05

10

15

20

25

Pure

tech

nica

l effi

cien

cy

(a)


00

02

04

06

08

10

Scal

e effi

cien

cy

(b)


8 Complexity












Population density

057044 038

047027 020





0120 0

021008 0





031

008 0

031

006 0City size 008 006



023025

025


Complexity 9



4 Discussion




5 Conclusions




10 Complexity







Data Availability




References














Complexity 11
































12 Complexity


R(ω) Remp + λ ω2 1113944

l

i1e f xi( 1113857 minus yi( 1113857 + λω

2 (7)


ω 1113944l



f(x) 1113944l


l


(9)



3 Results








Complexity 5



GUEUCL

evaluation

Analysis ofGUEUCL

factors


GUEUCL


Spatial variogram


Spaceheterogeneity





0 40 80 160km

N

(a)


0 40 80 160km

N

(b)



ndash40

ndash24

ndash08

08

24

40

Lagg

ed G

UEU

CL 2

011


(a)



ndash35

ndash21

ndash07

07

21

35

Lagg

ed G

UEU

CL 2

015

(b)


6 Complexity






N

0 40 80 160km

(a)


N

0 40 80 160km

(b)





Complexity 7






2011 1844 0677 1846 0146 1983 0002 1589 0821 1878 04062015 1900 0406 1942 0026 1861 0242 1614 0703 1978 0026

0155

0310

0465

0621

Sem

ivar

ianc

e

238179119090 35726900


N-S W-E

ndash0155 ndash2037

(a)

0425

0566


Sem

ivar

ianc

e

0283

238179119090 3572690

0142 N-S W-E

ndash0353 ndash1762

(b)



00

05

10

15

20

25

Pure

tech

nica

l effi

cien

cy

(a)


00

02

04

06

08

10

Scal

e effi

cien

cy

(b)


8 Complexity












Population density

057044 038

047027 020





0120 0

021008 0





031

008 0

031

006 0City size 008 006



023025

025


Complexity 9



4 Discussion




5 Conclusions




10 Complexity







Data Availability




References














Complexity 11
































12 Complexity



GUEUCL

evaluation

Analysis ofGUEUCL

factors


GUEUCL


Spatial variogram


Spaceheterogeneity





0 40 80 160km

N

(a)


0 40 80 160km

N

(b)



ndash40

ndash24

ndash08

08

24

40

Lagg

ed G

UEU

CL 2

011


(a)



ndash35

ndash21

ndash07

07

21

35

Lagg

ed G

UEU

CL 2

015

(b)


6 Complexity






N

0 40 80 160km

(a)


N

0 40 80 160km

(b)





Complexity 7






2011 1844 0677 1846 0146 1983 0002 1589 0821 1878 04062015 1900 0406 1942 0026 1861 0242 1614 0703 1978 0026

0155

0310

0465

0621

Sem

ivar

ianc

e

238179119090 35726900


N-S W-E

ndash0155 ndash2037

(a)

0425

0566


Sem

ivar

ianc

e

0283

238179119090 3572690

0142 N-S W-E

ndash0353 ndash1762

(b)



00

05

10

15

20

25

Pure

tech

nica

l effi

cien

cy

(a)


00

02

04

06

08

10

Scal

e effi

cien

cy

(b)


8 Complexity












Population density

057044 038

047027 020





0120 0

021008 0





031

008 0

031

006 0City size 008 006



023025

025


Complexity 9



4 Discussion




5 Conclusions




10 Complexity







Data Availability




References














Complexity 11
































12 Complexity






N

0 40 80 160km

(a)


N

0 40 80 160km

(b)





Complexity 7






2011 1844 0677 1846 0146 1983 0002 1589 0821 1878 04062015 1900 0406 1942 0026 1861 0242 1614 0703 1978 0026

0155

0310

0465

0621

Sem

ivar

ianc

e

238179119090 35726900


N-S W-E

ndash0155 ndash2037

(a)

0425

0566


Sem

ivar

ianc

e

0283

238179119090 3572690

0142 N-S W-E

ndash0353 ndash1762

(b)



00

05

10

15

20

25

Pure

tech

nica

l effi

cien

cy

(a)


00

02

04

06

08

10

Scal

e effi

cien

cy

(b)


8 Complexity












Population density

057044 038

047027 020





0120 0

021008 0





031

008 0

031

006 0City size 008 006



023025

025


Complexity 9



4 Discussion




5 Conclusions




10 Complexity







Data Availability




References














Complexity 11
































12 Complexity






2011 1844 0677 1846 0146 1983 0002 1589 0821 1878 04062015 1900 0406 1942 0026 1861 0242 1614 0703 1978 0026

0155

0310

0465

0621

Sem

ivar

ianc

e

238179119090 35726900


N-S W-E

ndash0155 ndash2037

(a)

0425

0566


Sem

ivar

ianc

e

0283

238179119090 3572690

0142 N-S W-E

ndash0353 ndash1762

(b)



00

05

10

15

20

25

Pure

tech

nica

l effi

cien

cy

(a)


00

02

04

06

08

10

Scal

e effi

cien

cy

(b)


8 Complexity












Population density

057044 038

047027 020





0120 0

021008 0





031

008 0

031

006 0City size 008 006



023025

025


Complexity 9



4 Discussion




5 Conclusions




10 Complexity







Data Availability




References














Complexity 11
































12 Complexity












Population density

057044 038

047027 020





0120 0

021008 0





031

008 0

031

006 0City size 008 006



023025

025


Complexity 9



4 Discussion




5 Conclusions




10 Complexity







Data Availability




References














Complexity 11
































12 Complexity



4 Discussion




5 Conclusions




10 Complexity







Data Availability




References














Complexity 11
































12 Complexity







Data Availability




References














Complexity 11
































12 Complexity
































12 Complexity

Documents

Spatial Pattern Characteristics and Influencing Factors of