Modeling and Analysis of Ecological Urban Landscape

Research ArticleModeling and Analysis of Ecological Urban Landscape PatternEvolution Based on Multisource Remote Sensing Data

Zhang Min1 Wang Xuejie 1 and Liu Yun2

1Academy of Art and Design Anhui University of Technology Marsquoanshan China2School of Information Engineering Chaohu University Chaohu China

Correspondence should be addressed to Wang Xuejie zm7773ahuteducn

Received 1 May 2021 Accepted 21 May 2021 Published 30 May 2021

Academic Editor Huihua Chen

Copyright copy 2021 Zhang Min 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

Considering that the development of urbanization cannot be separated from the application of landscape pattern evolution inorder to improve the development level of ecocity a modeling analysis of ecological urban landscape pattern evolution based onmultisource remote sensing data is proposed Taking ecotype city as the research object the remote sensing images of ecologicalurban landscape pattern are screened by using multisource remote sensing data and nonremote sensing data as the basic data CA-Markov model is constructed and the evolution of ecological urban landscape pattern is analyzed )e experimental results showthat from 2005 to 2020 the development level of urbanization process is higher and higher and the area of building land patches isincreasing which reduces the fragmentation of building land patches However the landscape of cultivated land and green space isless and less and the distribution of the patches also causes uneven phenomenon which leads to the gradual decline of ecologicalurban landscape diversity In the ecological urban landscape pattern the degree of fragmentation and diversity of urban landscapeis reduced due to the high connection and coverage of the construction land

1 Introduction

Due to Chinarsquos vast territory and large population energyconsumption is increasing As a result a series of envi-ronmental pollution problems have become more and moreserious )e destruction of the ecological environment hascaused extreme weather such as drought and high tem-perature leading to frequent occurrences of natural disas-ters )erefore the protection of the ecological environmenthas become the primary issue to realize the sustainabledevelopment of mankind [1] At present the ecologicalenvironment in densely populated urban areas is seriouslydamaged and urban green spaces forests and water play adecisive role in the entire ecological environment cycle Itcan not only purify waste but also regenerate the ecologicalenvironment )erefore building an ecological urbanlandscape pattern is an inevitable way to achieve sustainablesocial development [2] )erefore protecting the ecologicalenvironment and building an ecological city are the currentfocus of the Chinese government

In the process of urbanization due to the diversifi-cation of land use in ecological cities the landscapepattern of each region is different )ere are highly ur-banized business districts urbanized suburbs and eco-logical forest land that has not yet been urbanized [3 4])e urban landscape pattern is a witness to the evolutionof urban development It records the traces of urbanconstruction and also reflects the overall style of the city Itis a unique perspective of urban development [5] As thefirst large-scale city developed in China its urban land-scape pattern leads to the urbanization process of othercities )e construction of an ecological urban landscapepattern has an important impact on the sustainable de-velopment of the city and is the development direction offuture urbanization [6]

Based on the above analysis this paper appliesmultisource remote sensing data to the modeling andanalysis of the evolution of the ecological city landscapepattern so as to improve the development level of theecological city

HindawiComplexityVolume 2021 Article ID 8158158 8 pageshttpsdoiorg10115520218158158

2 Basic Theories and Models of Research

21 e Basic eory of the Research )e sustainable de-velopment of a city is complex system engineering It in-volves many aspects such as environment resources andmode )e process of urbanization has a huge impact on thetypes of urban landscapes which must be combined with themigration of population and the adjustment of industrialstructure )e landscape pattern is reasonably evolved )eexisting urban landscape pattern is formed through a longprocess of development and is a manifestation of the har-monious development of nature and mankind )e processof urbanization is the process of transforming ecologicalland into construction land After continuous transforma-tion the urban ecological landscape pattern based on man-made landscape pattern has been formed in the city Ecotypecities rely on their unique geographical advantages as theforerunners of urbanization leading to the development ofthe economy and the process of urbanization has changedfrom fast to slow with a qualitative leap

Ecocity is used as the research object and its landscapepattern is modeled and analyzed based on multisource re-mote sensing data Ecocity is a sustainable developmentmodel that uses environmental resources to realize industrialproduction and human life [7] From a narrow perspectivean ecocity is to design the urban landscape pattern accordingto ecological principles to establish a harmonious efficientand healthy human settlement environment [8]

Taking ecological city as the research object whenmodeling and analyzing the landscape pattern evolution ofecological city multisource remote sensing data and non-remote sensing data are mainly used as the basic data [9]Multisource remote sensing data are core data for interpre-tation and translation of remote sensing images Nonremotesensing data are mainly used to assist the interpretation andtranslation of remote sensing images )e structure of theecological urban landscape pattern is shown in Figure 1

After screening the remote sensing images of the eco-logical urban landscape pattern the remote sensing imagesin Table 1 are finally selected as the data source forinterpretation

Nonremote sensing data are used to assist in interpre-tation of remote sensing images It mainly includes soil-typemaps land use maps field survey historical data admin-istrative division maps and DEM maps [10] as shown inTable 2

22 CA-Markov Model CA-Markov model has no afteref-fect and can predict the future time state It is only related tothe current ecological urban landscape pattern not affectedby the past state and future state and the trend is stable Itcan better predict the quantity )e simulation results areonly quantitative changes but not embodied in space )eCA-Markov model can predict the quantitative change ofecological urban landscape pattern generate the suitabilityof urban landscape pattern as the conversion rule accordingto multisource remote sensing data and run the CA moduleto simulate the future land use spatial distribution pattern in

the study area )e CA-Markov model can make the sim-ulation accuracy of ecological urban landscape pattern landevolution higher )e working principle of the CA-Markovmodel is that CA is transferred by converting the image ofarea and probability thereby simulating the evolutionprocess of urban landscape pattern )e specific operationsteps are as follows

Step 1 )e standard for CA model transfer is constructedFor the same area of a city urban land use trans-

portation and water sources are all important factors thatdetermine the evolution of the urban landscape pattern)ese determine the suitability of ecological city construc-tion )erefore the image probability image is used totransform the city suitability image

Step 2 )e simulation transition probability matrix isestablished

)e Markov model is used to simulate the matrix of thetransition probability of the urban landscape pattern and thematrix of the transition area

Ecological urban landscape

Expediency ofroad

Environmentalecology

Peoplersquos psychologicalcomfort

Urbanfunctionalization

Sense of spatialorder

Figure 1 Structure diagram of ecological urban landscape pattern

Table 1 Remote sensing image sources

Landscapepattern number Satellite type Remarks

123-32 Landsat5 TM )e remote sensing image datacome from the geospatial datacloud platform of the computernetwork information centerand the cloud coverage rate is

less than 5

123-32 Landsat7 ETM

123-32 Landsat8 OLI

Table 2 Nonremote sensing data sources

Type of data SourceSoil-type map Agriculture bureauLand use map Bureau of land and resourcesFieldwork historical data Historical dataAdministrative division map Bureau of surveying and mappingDEM diagram Geospatial data cloud

2 Complexity

Step 3 CA model application)e CA model is used to break the original landscape

pattern model Combining the weighting factor to recom-bine the new urban landscape pattern state can be obtained)e weight factor includes the type quantity spatial dis-tribution and configuration of landscape units in urbanlandscape pattern By importing the data of weight factorsinto the Markov model for analysis we can get the situationthat each weight factor may have a certain type of landscapepattern utilization

Step 4 )e evolution process of the future urban landscapepattern is simulated )e specific flowchart is shown inFigure 2

)rough CA to simulate the fragmentation and reor-ganization of the urban landscape pattern the process of theevolution of the urban landscape pattern is simulated

After the CA simulation the collected data and mod-ularization can be processed for information fusion )eCA-Markov model is shown in Figure 3

3 Specific Implementation Methods

)e landscape pattern index is an important indicator ofurban ecological construction and it is also an importanttool for managing urban landscape pattern and monitoringand evaluating it [11] Based on the landscape pattern indexthrough the description and analysis of the dynamic changesof the urban landscape the impact of urbanization on theecological environment is better understood [12] Multi-source remote sensing data are used to calculate the index ofecological urban landscape pattern )e main content isshown in Figure 4

)e index calculation method and significance of theevolution process of urban landscape pattern are as follows

31 Mean Patch Size (MPS)

MPS Ai

Ni

(1)

In the formula Ai is the total area of landscape greenspace Ni is the number of green plots in the i landscapeMPS mainly reflects the average fragmentation degree of thelandscape

32 Maximum Plaque Index (LPI)

LPI max a1 a2 an( 1113857

Atimes 100 (2)

In the formula a is the green area in a certain area ofthe city A is the total area of the landscape pattern of theurban research area LPI is used to express the percentageof the largest green area of the urban green area to the totalarea and the green index can be measured at differentlevels [13]

33 Plaque Density (PD)

PD 1113944m

i1

Ni

A (3)

In the formula m is the total number of urban landscapepatterns Ni is the number of urban landscape green spacesA is the total landscape area of the urban study area PD isused to represent the average fragmentation status of thelandscape [14]

34 Contagion Index (CONTAG)

CONTAG 1 + 1113936

mi1 1113936

mk1 Pi gik1113936

mk1 gik( 1113857( 1113857ln Pi gik1113936

mk1 gik( 1113857( 11138571113858 1113859

2 lnm1113890 1113891 (4)

In the formula Pi is the proportion of the total greenarea of the urban landscape pattern gik is the number ofadjacent sprawl grids from category i to category kCONTAG represents the development trend between urbangreen space types )e larger the index value the better thegreen space types of the urban landscape pattern On thecontrary the lower the index value the more scattered theurban landscape pattern and the higher the average degree offragmentation [15]

35 Fragmentation Index

DIVISION 1 minus 1113944m

i1

ai

A1113874 1113875⎡⎣ ⎤⎦ (5)

)e DIVISION index represents the actual situat-ion of fragmentation of the urban landscape pat-tern )e fragmentation of the green space in thelandscape becomes more complicated as the index valueincreases

36 Diversity Index

SHDI minus 1113944m

i1pi times lnpi( 1113857 (6)

In the formula pi is the total amount of green space inthe adjacent urban landscape pattern When the publicboundary of the green space in the urban landscape patternbecomes larger and larger the degree of fragmentation of the

Complexity 3

landscape is higher On the basis of the above indexesecological urban landscape pattern renderings are con-structed as shown in Figure 5

Cityscape image

CA-Markovmodel

Simulationfusion

Exportresults

Data 1Information

classification 1

Optimization

Data 2

Data n

Information classification 2

Information classification n

Analysisof fittingresults

Figure 3 Schematic diagram of CA-Markov model

Index of ecological urban

landscape pattern

Average patch area

Maximum plaque index

Patch density

Sprawl index

Fragmentation index

Diversity index

Figure 4 Index structure diagram of ecological urban landscapepattern

Figure 5 Effect picture of ecological urban landscape pattern

Start

Data collection

Build CA modeltransfer standard

Qualified or not

The simulatedtransition probability

matrix is built

CA model application

Dense structure isbuilt

The evolution of the simulated urbanlandscape pattern was completed

EndY

N

Data preprocessing

Forecast the futurestate

Figure 2 Flowchart of the evolution of the simulated urban landscape pattern

4 Complexity

4 Main Experiments and Analysis

41 Overall Characteristics and Changing Trends of EcologicalUrban Landscape Pattern )e distribution of landscapetypes of ecocity in 2005 2015 and 2020 is shown inFigure 6

According to the evolution of the ecological urbanlandscape pattern in Figure 6 the detailed data and trend of

the landscape pattern evolution are formulated as shown inTable 3

From the data in Table 3 it can be seen that the com-position ratio of the ecological city landscape fluctuatesgreatly During the fifteen years from 2005 to 2020 thewoodland area of the ecological city exceeds 42 which hasgreat advantages With the development of urbanization andthe rapid increase of population arable land and grassland have

Arable landGrasslandWetlandsConstruction land

WoodlandBushWaterBare land

2005

N

(a)



2015

(b)



2020

(c)

Figure 6 Distribution map of landscape types

Complexity 5

become the main sources of ecological urban constructionland so the above three types of landscape changes are thebiggest However the proportion of construction land hasexceeded 10 and its growth rate is getting faster and faster inthe later period )e proportion of cultivated land andgrassland pattern has gradually declined especially the declineof cultivated land In terms of water area pattern the pro-portion has declined slightly but the total area is almost un-changed which has little impact on the ecological urbanlandscape Other patterns can be ignored due to the small base

42 Individual Characteristics and Changing Trends of Eco-logical Urban Landscape )emaximum patch index is usedto measure the dominance of the ecological urban landscapeat the patch type level and the results are shown in Figure 7

It can be seen from the results in Figures 7 and 8 thatwoodland has always been a relatively dominant landscapein ecological cities and this result is completely consistentwith the natural environment of ecological cities )emaximum patch index of construction land has risen sig-nificantly with an increase of 828 )erefore it can beseen that in ecological cities construction land has shown asignificant growth trend and the advantages of otherlandscape types are relatively stable )ere is basically norelatively large fluctuation

Table 3 )e overall composition and changes of the ecologicalurban landscape pattern

Type of plaquePatch area

2005 (km2) 2015 (km2) 2020 (km2)Arable land 586227 554165 452617Woodland 696274 706215 735663Grassland 159232 126347 57860Bush 026 026 197Wetlands 743 1339 3178Water 26723 15015 17552Construction land 161244 227356 363221Bare ground 016 023 196

Type of plaque Composition ratio2005 () 2015 () 2020 ()

Arable land 359541 339877 277597Woodland 427035 433132 451193Grassland 97659 77490 35486Bush 00016 00016 00121Wetlands 00456 00821 01949Water 16389 09209 10765Construction land 98893 139441 222769Bare ground 00011 00014 00120

Type of plaque Change ratio2005sim2015 () 2015sim2020 ()

Arable land minus197 minus623Woodland 061 181Grassland minus202 minus420Bush 000 001Wetlands 004 011Water minus072 016Construction land 405 833Bare ground 000 001

Type of plaque

Ara

ble l

and

Bush

Wet

land

s

Wat

er b

ody

Cons

truc

tion

land

Bare

land

Woo

dlan

d

Gra

ssla

nd

0

10

20

30

40

Max

imum

pla

que i

ndex

()

200520152020

Figure 7 Changes in patch types of ecological urban landscape

Arableland

Woodlandchanging trend

Construction land

0

10

20

30

40

Max

imum

pla

que i

ndex

()

200520152020

Figure 8 Trend chart of cultivated land forest land and con-struction land

6 Complexity

)e landscape index of the ecological urban landscapelevel is shown in Table 4 and the fragmentation index isshown in Table 5

It can be seen from the results in Table 4 that the totalnumber of patches of ecological urban landscape was 72884in 2005 and it will increase to 76782 in 2020 reflecting thefragmentation trend of ecological urban landscape

From the above results it can be seen that the totalnumber of patches in the ecological urban landscape in 2005was 72884 which increased to 76782 by 2020 reflecting thetrend of fragmentation in the ecological urban landscapeCultivated land patch is a landscape with the most obviouschanges Among all landscape patches it has increased byabout 50 and the average patch area has decreased byabout 58 resulting in a continuous increase in patchdensity which indicates that the cultivated land patches arefragmented In the ecocity the environmental carryingcapacity has dropped significantly From 2005 to 2020 thenumber of patches of construction land has been on the rise

and the average patch area has shown a gradual increasingtrend However with the advancement of urbanizationconstruction land has been distributed in a series of land-scapes )e patch density of ecocity construction land has asmall trend and the overall urban diffusion and agglom-eration are gradually increasing indicating that the grasspatch density is basically unchanged only the average area isdecreasing indicating that the number of grass patches isincreasing However it is relatively scattered in the spatialdimension showing a star-like distribution

5 Conclusion

)is paper proposes a modeling analysis of the evolution ofthe ecological urban landscape pattern based on multisourceremote sensing data Taking the ecological city as the re-search object the landscape pattern index and the CA-Markov model are used to quantitatively analyze the evo-lution of the ecological urban landscape pattern to improvethe development level of ecological cities In ecological citiesthe landscape pattern of increasingly higher building landand declining ecological land has a direct impact on theecological environment of the ecological city )e reason isthat the increase in the area of building land has led toincreasing drainage pressure in the ecological city )ehigher it is the frequent the occurrence of water loggingdisasters is the construction land will also affect the re-plenishment of groundwater in ecological cities leading tothe emergence of ecological problems Reducing the eco-logical land in ecological cities increases the frequency ofnatural disasters and at the same time affects urban safety

Data Availability

)e simulation data are used and the model and relatedhyperparameters are provided in our paper

Conflicts of Interest

)e authors declare that they have no conflicts of interest

Acknowledgments

)is work was supported by the Key Research Projects ofHumanities and Social Sciences in Colleges and Universitiesof Anhui Province (Subject no RZ2000003385) and also bythe Key Research Project of Natural Science in AnhuiProvince (KJ2019 A0681)

References

[1] M A Hua Y Wang Y Ning et al ldquoDynamic analysis oflandscape pattern of the plateau wetland of Zoige CountyrdquoForest Resources Management vol 33 no 1 pp 109ndash115 2019

[2] C Chulin H Wenming S Lei et al ldquoEvolution analysis offorest wetland landscape patterns of rural settlements in theYangshan area of Haikoucityrdquo Journal of Central SouthUniversity of Forestry amp Technology vol 40 no 2 pp 131ndash1412020

Table 5 Fragmentation index

Type of plaqueNumber of plaques

2005 2015 2020Arable land 858 1300 1562Woodland 12613 13634 14823Grassland 54275 56617 55120Bush 236 236 251Wetlands 67 40 58Water 2496 1733 1972Construction land 2288 2546 2925Bare ground 51 56 71

Type of plaque Average patch area2005 (km2) 2015 (km2) 2020 (km2)


Type of plaque Patch density2005 (kmminus2) 2015 (kmminus2) 2020 (kmminus2)


Table 4 Landscape index of ecological city landscape level

Years 2005 2015 2020Number of plaques 72884 76162 76782Average patch area 02237 km2 02141 km2 02124 km2

Sprawl index 645128 647397 652860Separation index 08488 08488 08468Polymerization index 948362 948392 957733

Complexity 7

[3] C T Li and C L Yunjian ldquoChanges in landscape pattern ofbuilt-up land and its driving factors during urban sprawlrdquoActa Ecologica Sinica vol 40 no 10 pp 133ndash144 2020

[4] L U Leting J Zhang Q Peng et al ldquoLandscape patternanalysis and prediction in the Dongjiang river basinrdquo ActaEcologica Sinica vol 39 no 18 pp 6850ndash6859 2019

[5] J Luo Z Sun and X Zhang ldquoAnalysis of the characteristicsand changes of landscape pattern of oasisin Ganzhou districtof Zhangye city based on ripleyrsquos K functionrdquo Research of Soiland Water Conservation vol 26 no 4 pp 224ndash231 2019

[6] Y Hang X Cai Y Chao et al ldquoDriving force analysis oflandscape pattern changes in Honghu wetland nature reservein recent 40 yearsrdquo Journal of Lake Sciences vol 31 no 1pp 171ndash182 2019

[7] Y Wan-Ying L Yan-Fang L Yao-Lin et al ldquoInvestigatingthe effect of urban landscape pattern on PM 25 concentrationbased on lur model a chang-zhu-tan urban agglomerationcase studyrdquo Resources and Environment in the Yangtze Basinvol 28 no 9 pp 235ndash245 2019

[8] J Yuan W Chen B Du et al ldquoAnalysis of landscape patternon urban land use based on GF-5 hyperspectral datardquo Journalof Remote Sensing vol 24 no 4 pp 465ndash478 2020

[9] J-J Wang and B Gong ldquoEvolution of landscape pattern andecological risk in Xixian new areardquo Journal of NorthwestForestry University vol 34 no 2 pp 256ndash262 2019

[10] L Ming-Zhen Y-B Li and R Cai-Hong ldquoEvolution of rurallandscape pattern under the background of land use trans-formation based on the transect analysis of Caotangxi wa-tershedrdquo Journal of Natural Resources vol 35 no 9pp 257ndash272 2020

[11] J Lei Z Chen Y Chen et al ldquoLandscape pattern changesanddriving factors analysis of wetland in Hainan island during1990ndash2018rdquo Ecology and Environment Sciences vol 29 no 1pp 63ndash74 2020

[12] Z Luo X Hu B Wei et al ldquoUrban landscape pattern evo-lution and prediction based on multi-criteriaCA-Markovmodeltake Shanghang county as an examplerdquo EconomicGeography vol 40 no 10 pp 60ndash68 2020

[13] Y-Y Jia X-L Tang Y Yang et al ldquoLandscape patternchanges and ecological service values in Wuhu section alongthe yangtze riverrdquo Journal of Northwest Forestry Universityvol 34 no 6 pp 249ndash258 2019

[14] H Ma X Xu T Deng et al ldquoSpatial and temporal evolutionof landscape pattern in Jiangdong new area of Haikou cityfrom 1999 to 2018rdquo Journal of Southwest Forestry Universityvol 40 no 1 pp 116ndash123 2020

[15] L Ke-Jun F Lu-Ming X-B He et al ldquoRelationship betweenforestcity landscape pattern and thermal environment a casestudy of Longquan city Chinardquo Chinese Journal of AppliedEcology vol 30 no 9 pp 186ndash194 2019

8 Complexity

2 Basic Theories and Models of Research

21 e Basic eory of the Research )e sustainable de-velopment of a city is complex system engineering It in-volves many aspects such as environment resources andmode )e process of urbanization has a huge impact on thetypes of urban landscapes which must be combined with themigration of population and the adjustment of industrialstructure )e landscape pattern is reasonably evolved )eexisting urban landscape pattern is formed through a longprocess of development and is a manifestation of the har-monious development of nature and mankind )e processof urbanization is the process of transforming ecologicalland into construction land After continuous transforma-tion the urban ecological landscape pattern based on man-made landscape pattern has been formed in the city Ecotypecities rely on their unique geographical advantages as theforerunners of urbanization leading to the development ofthe economy and the process of urbanization has changedfrom fast to slow with a qualitative leap

Ecocity is used as the research object and its landscapepattern is modeled and analyzed based on multisource re-mote sensing data Ecocity is a sustainable developmentmodel that uses environmental resources to realize industrialproduction and human life [7] From a narrow perspectivean ecocity is to design the urban landscape pattern accordingto ecological principles to establish a harmonious efficientand healthy human settlement environment [8]

Taking ecological city as the research object whenmodeling and analyzing the landscape pattern evolution ofecological city multisource remote sensing data and non-remote sensing data are mainly used as the basic data [9]Multisource remote sensing data are core data for interpre-tation and translation of remote sensing images Nonremotesensing data are mainly used to assist the interpretation andtranslation of remote sensing images )e structure of theecological urban landscape pattern is shown in Figure 1

After screening the remote sensing images of the eco-logical urban landscape pattern the remote sensing imagesin Table 1 are finally selected as the data source forinterpretation

Nonremote sensing data are used to assist in interpre-tation of remote sensing images It mainly includes soil-typemaps land use maps field survey historical data admin-istrative division maps and DEM maps [10] as shown inTable 2

22 CA-Markov Model CA-Markov model has no afteref-fect and can predict the future time state It is only related tothe current ecological urban landscape pattern not affectedby the past state and future state and the trend is stable Itcan better predict the quantity )e simulation results areonly quantitative changes but not embodied in space )eCA-Markov model can predict the quantitative change ofecological urban landscape pattern generate the suitabilityof urban landscape pattern as the conversion rule accordingto multisource remote sensing data and run the CA moduleto simulate the future land use spatial distribution pattern in

the study area )e CA-Markov model can make the sim-ulation accuracy of ecological urban landscape pattern landevolution higher )e working principle of the CA-Markovmodel is that CA is transferred by converting the image ofarea and probability thereby simulating the evolutionprocess of urban landscape pattern )e specific operationsteps are as follows

Step 1 )e standard for CA model transfer is constructedFor the same area of a city urban land use trans-

portation and water sources are all important factors thatdetermine the evolution of the urban landscape pattern)ese determine the suitability of ecological city construc-tion )erefore the image probability image is used totransform the city suitability image

Step 2 )e simulation transition probability matrix isestablished

)e Markov model is used to simulate the matrix of thetransition probability of the urban landscape pattern and thematrix of the transition area

Ecological urban landscape

Expediency ofroad

Environmentalecology

Peoplersquos psychologicalcomfort

Urbanfunctionalization

Sense of spatialorder

Figure 1 Structure diagram of ecological urban landscape pattern

Table 1 Remote sensing image sources

Landscapepattern number Satellite type Remarks

123-32 Landsat5 TM )e remote sensing image datacome from the geospatial datacloud platform of the computernetwork information centerand the cloud coverage rate is

less than 5

123-32 Landsat7 ETM

123-32 Landsat8 OLI

Table 2 Nonremote sensing data sources

Type of data SourceSoil-type map Agriculture bureauLand use map Bureau of land and resourcesFieldwork historical data Historical dataAdministrative division map Bureau of surveying and mappingDEM diagram Geospatial data cloud

2 Complexity










MPS Ai

Ni

(1)




Atimes 100 (2)



PD 1113944m

i1

Ni

A (3)



CONTAG 1 + 1113936

mi1 1113936

mk1 Pi gik1113936

mk1 gik( 1113857( 1113857ln Pi gik1113936

mk1 gik( 1113857( 11138571113858 1113859

2 lnm1113890 1113891 (4)




i1

ai

A1113874 1113875⎡⎣ ⎤⎦ (5)


36 Diversity Index

SHDI minus 1113944m



Complexity 3


Cityscape image

CA-Markovmodel

Simulationfusion

Exportresults

Data 1Information

classification 1

Optimization

Data 2

Data n






landscape pattern

Average patch area


Patch density

Sprawl index

Fragmentation index

Diversity index



Start

Data collection


Qualified or not


matrix is built




EndY

N

Data preprocessing



4 Complexity








2005

N

(a)



2015

(b)



2020

(c)


Complexity 5











Type of plaque

Ara

ble l

and

Bush

Wet

land

s

Wat

er b

ody

Cons

truc

tion

land

Bare

land

Woo

dlan

d

Gra

ssla

nd

0

10

20

30

40

Max

imum

pla

que i

ndex

()

200520152020


Arableland


Construction land

0

10

20

30

40

Max

imum

pla

que i

ndex

()

200520152020


6 Complexity





5 Conclusion


Data Availability




Acknowledgments


References













Complexity 7














8 Complexity










MPS Ai

Ni

(1)




Atimes 100 (2)



PD 1113944m

i1

Ni

A (3)



CONTAG 1 + 1113936

mi1 1113936

mk1 Pi gik1113936

mk1 gik( 1113857( 1113857ln Pi gik1113936

mk1 gik( 1113857( 11138571113858 1113859

2 lnm1113890 1113891 (4)




i1

ai

A1113874 1113875⎡⎣ ⎤⎦ (5)


36 Diversity Index

SHDI minus 1113944m



Complexity 3


Cityscape image

CA-Markovmodel

Simulationfusion

Exportresults

Data 1Information

classification 1

Optimization

Data 2

Data n






landscape pattern

Average patch area


Patch density

Sprawl index

Fragmentation index

Diversity index



Start

Data collection


Qualified or not


matrix is built




EndY

N

Data preprocessing



4 Complexity








2005

N

(a)



2015

(b)



2020

(c)


Complexity 5











Type of plaque

Ara

ble l

and

Bush

Wet

land

s

Wat

er b

ody

Cons

truc

tion

land

Bare

land

Woo

dlan

d

Gra

ssla

nd

0

10

20

30

40

Max

imum

pla

que i

ndex

()

200520152020


Arableland


Construction land

0

10

20

30

40

Max

imum

pla

que i

ndex

()

200520152020


6 Complexity





5 Conclusion


Data Availability




Acknowledgments


References













Complexity 7














8 Complexity


Cityscape image

CA-Markovmodel

Simulationfusion

Exportresults

Data 1Information

classification 1

Optimization

Data 2

Data n






landscape pattern

Average patch area


Patch density

Sprawl index

Fragmentation index

Diversity index



Start

Data collection


Qualified or not


matrix is built




EndY

N

Data preprocessing



4 Complexity








2005

N

(a)



2015

(b)



2020

(c)


Complexity 5











Type of plaque

Ara

ble l

and

Bush

Wet

land

s

Wat

er b

ody

Cons

truc

tion

land

Bare

land

Woo

dlan

d

Gra

ssla

nd

0

10

20

30

40

Max

imum

pla

que i

ndex

()

200520152020


Arableland


Construction land

0

10

20

30

40

Max

imum

pla

que i

ndex

()

200520152020


6 Complexity





5 Conclusion


Data Availability




Acknowledgments


References













Complexity 7














8 Complexity








2005

N

(a)



2015

(b)



2020

(c)


Complexity 5











Type of plaque

Ara

ble l

and

Bush

Wet

land

s

Wat

er b

ody

Cons

truc

tion

land

Bare

land

Woo

dlan

d

Gra

ssla

nd

0

10

20

30

40

Max

imum

pla

que i

ndex

()

200520152020


Arableland


Construction land

0

10

20

30

40

Max

imum

pla

que i

ndex

()

200520152020


6 Complexity





5 Conclusion


Data Availability




Acknowledgments


References













Complexity 7














8 Complexity











Type of plaque

Ara

ble l

and

Bush

Wet

land

s

Wat

er b

ody

Cons

truc

tion

land

Bare

land

Woo

dlan

d

Gra

ssla

nd

0

10

20

30

40

Max

imum

pla

que i

ndex

()

200520152020


Arableland


Construction land

0

10

20

30

40

Max

imum

pla

que i

ndex

()

200520152020


6 Complexity





5 Conclusion


Data Availability




Acknowledgments


References













Complexity 7














8 Complexity





5 Conclusion


Data Availability




Acknowledgments


References













Complexity 7














8 Complexity














8 Complexity

Documents

Modeling and Analysis of Ecological Urban Landscape