The New Alison App has just launched Download Now
We'll email you at these times to remind you to study
You can set up to 7 reminders per week
We'll email you at these times to remind you to study
So, today’s lecture is going to cover concepts and the very first one that we are going to talkabout is urban sprawl characterization. Now, the next that we are going to look into is themetrics that can be used to assess the growth pattern in urban areas. We would be looking intothe patterns of urban growth and we can do a urban growth analysis using such landscapemetrics.
(Refer Slide Time: 01:43)
So, let us see first, what are the different types of metrics that can be used to characterize orassess the urban growth pattern. Now, there are different types of metrics which can be usedto characterize the underlying spatial processes and we can determine these spatial processes, Imean by spatial processes we mean that the land use may change, it may evolve as, I meancompact forms into compact forms or it could be broken up into small fragments, what is the Imean your combination of the different types of land use in a given area.
So, all these can be studied. So, first we can assess the complexity or aggregation in a urbanarea, so, using this particular metrics. Then we can assess the centrality, how central thisfeature is or that particular patch is, we can assess the compactness and dispersion of thedifferent patches in the urban area.
We can also look for parameters of porosity which basically gives us whether the your I meanwhether there are any gaps to be in the urban area, where it could be seen as pores. So, in thefirst image you can see the example wherein, you can see in this particular image the form ishighly complexed, it is serrated and the edges are really complex. So, this is an example of acomplex shape of a particular patch, on the other hand you can see this is a particular verysimple shape. So, its shape complexity is low.
Talking about the centrality, you can see that these patches are very close to each other. So, inthis case the centrality measure of centrality is relatively high compared to another patch,wherein the centroidal distance of the patch is relatively on the higher side. So, here in thisparticular case, the measure of centrality as you would calculate using urban metrics would beon the lower side.
Now, if you come back to the next image, we can see an example of compactness in thisparticular figure. So, here you can see that the figures are aggregated and they have the shapecomplexity simple and they are aggregated, they are not dispersed and not fragmented. So,this compactness in this first case is high where is whereas, if we see the next image you cansee the different patches.
So, if we imagine this to be an urban area, you may have different scattered patches. So, in thiscase the compactness is relatively low. The next one is the case of porosity. So, in case ofporosity we may have a particular land use, which is land use or land cover which is given bythis outer boundary and then we may have different classes within this particular class.
So, in this case you can see there are more than one classes or there are too many polygonsinside this particular class or patch. So, in this case the porosity is very high, this particularclass is porous. I mean there are other classes lying within this particular class. So, on theother hand if you see the adjoining image, in this case you see that you have this class which isbounded by this outer line outer edge and you have a internal polygon which belongs toanother class.
So, in this case if you do a comparison between these two images on your left and right forthis porosity assessment, you can see the porosity of this particular class that is given as whiteis much less than the porosity of this particular patch. So, in this case you have more numberof pores of other class in this particular catch, patch.
So, we can do I mean study the spatial processes, depending on the complexity, depending onthe centrality of the patches, depending on the compactness of the patches and also dependingon the porosity of the patches.
(Refer Slide Time: 06:58)
So, if we I mean do a growth analysis, I mean we can use these kind of measures to assess, Imean how the growth is evolving over time. So, let us see an example of metropolitan area ofGranada in Spain and this study has been done by Francisco Aguilera.
So, you can see that this is the your I mean the land use plan of Granada city in Spain. So, inthis particular region you can see that there is a, I mean compact centre and then you haveoutlying areas and you have speckles of growth. So, we can analyse this using the differenttypes of metrics. So, I mean if we take images, you can see that within the image you can seehigh density residential areas, you can see the low density residential areas, you can see thecommercial area and I mean industrial area and the commercial area at the bottom.
So, you can see that all these different images have got characteristic spatial I mean yourdistribution of the different land use patches.
(Refer Slide Time: 08:30)
So, these can be analysed in the using your spatial metrics. So, there are four patterns ofUrban growth in the last image we have seen the land use plan of Granada and there you canidentify four patterns of Urban growth. So, we shall try to see the Mumbai metropolitan
region and find out what are the different types of patterns of urban growth in the Mumbaimetropolitan corporation metropolitan region.
So, first is the aggregated pattern which I mean is gives an increased aggregation and it isgenerally by reduced dispersion of the different patches, you can see the, result outcome asaggregated patches. So, in this the urban growth pattern, I mean it either remains constant or Imean there would be infill development and as a result it would reduce the fragmentation inthe landings landscape. So, we can see in this particular example that we have this aggregatedpattern.
Now, we have the next pattern we are talking about four patterns of organ urban growth thenext pattern is the linear pattern. So, these patterns generally these kind of developmentgenerally happens along the road networks generally, these type of developments are industrialor mixed industrial residential land use. So, these happen along the major network corridors.
So, we can identify two processes from this linear pattern. The first one is increased or stableaggregation and the second one is decreased compaction. So, you can see this particularimage, wherein it shows a linear pattern.
So, we are talking I mean we are seeing an image of Mumbai metropolitan region, whereinyou can see that in the Navi Mumbai area along the major transportation spine you have mostof the development. So, this is the linear pattern of growth and in the Fort region or theColaba region you can see high dense development wherein the landscape fragmentation hasdecreased over time.
So, this represents the aggregated pattern. So, the different colours gives you an idearegarding when and how the Urban transformation has happened in a temporal way in atimeline.
(Refer Slide Time: 11:23)
So, let us see the next two patterns, the third one is the leapfrogging pattern. So, in thispattern the urban patches of residential use appear as I mean in a dispersed fashion. So, I meanit is low in density and generally, you would see come across a single family houses and it isdominated by process of decreased aggregation and greater dispersion of the urban use. Youwill also see the process of decreased elongation and which is result of the formation of moreor less rounded patches.
So, you would also see the process of increased dispersion in this particular case. So, here youcan see an example of your leapfrogging. So, in this case you can see the patches aredispersed. So, the dispersion is more, the elongation is not there and the, I mean these patchesare more or less rounded in nature and there is a increased dispersion in this patches. The next
and the final pattern of urban growth that we can see which is manifested in different cities isthe nodal pattern of growth.
It reflects the existing commercial or industrial urban growth along the main transportationnodes.
So, this leads to decrease in aggregation along the along with increase in the dispersion andalso your increase in compaction. So, this is how it happens, I mean as a nodal pattern. So,you can see that most of this growth would be in the industrial or the commercial urbangrowth centres. So, we can see this as an example in case of Mumbai metropolitan regionagain. So, here you can see there are some speckles of leapfrogging pattern that we havediscussed earlier and then you have the nodal pattern of growth happening in the new nodes ofthe new Bombay region.
So, again you have the temporal growth which is shown in this particular region. So, we knowthat there are four different patterns of growth and by characterizing these patterns usingspecial metrics we know that whether it is a nodal pattern whether it is leapfrogging whether itis aggregated pattern. So, we can assess what is the nature or the what are the processes thatis going behind in the those Urban areas, I mean which would result in a particular nature orpattern of growth.
(Refer Slide Time: 14:35)
Now, let us see I mean how we can do a growth analysis. So, this particular example is ofKathmandu valley and this is a study by Ishtiaque, Shrestha and Chhetri you can refer to thisparticular research article. So, in this particular research what the authors have done is theyhave tried to identify the urban growth in Kathmandu valley and use your landscape metrics toassess how the growth pattern is varying over time.
So, there are several questions which can be further evaluated based on this particular studyha, which can be further I mean delved into I mean there are some research questions like, Imean the what are the sustainably it implications of urban sprawl in a fragile mountainouslandscape like Kathmandu, then we can also explore the impacts of conversion of agriculturalland to built up areas.
Here, we can also do an analysis of the socio ecological significance of how the open spacesare disappearing. Also, we can study the fragmentation of the habitats which are importantbiological corridors where in it supports the fauna and the flora in this particular region.
(Refer Slide Time: 16:22)
We can also do the metric, I have the metrics to assess the Urban growth patterns. So, the firstone that we see is the percentage of landscape. So, it is given by the proportion of area of apatch to the total landscape area and it is expressed as percentage, then we can count thenumber of patches in of each land use.
So, you can see it is given as ni subscript i then we can calculate the mean patch size which isthe summation of all the areas of different patches divided by the number of patches. Now, wecan find out the mean radius of gyration which is the mean distance between each cell in the
patch and the cell centroid. So, I mean your hijr is the distance between cell ijr that is locatedwithin patch ij and the centroid of the Pi ij which is the average location.
So, based on the cell centre to cell enter distance we can calculate this hijr and your zi trepresents the number of patches in number of cells in patch ij, then we have the mean shapeindex which gives us a measure of the ratio between the perimeter of the patch and theperimeter of the simplest patch in the same area. So, this again is a ratio and does not have anyunit, then the next measure that we come across is the mean Euclidean distance neighbour.
So, it measures the distance or the average distance between two patches in a landscape. So,again you can find out hij which is the distance from patch i to the nearest neighbouring patchof the same class type. So, based on the patch h to h distance I mean we can compute the cellcentre to centre distance. So, this is summed up for all the different patches and it is divided bythe number of patches of a particular class type. So, we divided by n i that gives us the meanEuclidean distance and it could be this is measured in terms of distance metrics like eithermeters or kilometres or any other units of distance measure.
(Refer Slide Time: 19:14)
Now, we have few other measures to assess the urban growth patterns which are commonlybeing used. So, the first one is the area weighted mean shape index, wherein it measures the, Imean shape complexity of a particular patch, wherein si and pi are the area and perimeter ofpatch i and N capital N is the total number of patches. The second metric is the area weightedmean patch, fractal dimension wherein we try to measure the fractal nature in the landscape,wherein it gets repeated at different scales, si represents the area and pi again represents theperimeter.
So, it is a we take a logarithmic function of the these ratios of perimeter to the log of yourarea and we divide it by the total number of patches and it is further multiplied with the area ofthe individual patch type and it is divided by the aggregated some of the different patch types.So, it gives us that freak, fractal dimension.
So, the next measure is the centrality wherein you guess you can see we find out the centroiddistance of patch i to the centroid of the largest patch. So, I mean it is normalized and dividedby N minus 1 where is N is the total number of patches and your R is the radius of the circle orwith the area S. So, I mean sometimes your patches may not be a regular circle. So, what wecan do is we can I mean take the area S and divide it by pi and take a square root of that.
So, that this gives us a measure of the centrality of a given patch to the biggest patch with ofthe same class type. The next measure is the compactness wherein we find out the area and theperimeter, I mean we do a ratio of that and I mean your pi in this case is the perimeter of thecircle with the area si and again N is the number of patches. So, in the denominator we takesquare of N, because we are I mean I mean taking the ratio of the perimeter of the circle.
So, I mean in this case, you do not have a unit for this particular measure. Next is thecompactness index of the largest patch. So, like in the earlier, I mean index we had seen thecompactness index. It is very similar to it, but we in the denominator we have the yourperimeter of the we divided by the perimeter of the largest patch instead of dividing by thesquare of the number of patches, we divided by the perimeter of the largest patch.
The next metric that we are talking about is the ratio of open space. So, I mean it gives you ameasure of the porosity. So, in this case your S is the summation area of all the holes insidethe urban extracted urban area.
So, if you have a I mean built up area. So, wherever the gaps are there I mean unbuilt areasare there that would integrate to give you the measure of S. So, this S is the summation areaof all the patches so, I mean we do a ratio of S dash by S for finding out the ratio of the openarea and it is a measure of the porosity. Next, we can measure the density which is very simplethat is we take the measure of total population of a city and S is your summarization area of allthe patches. So, I mean we can find out the overall density or we can find out patch wisedensity as well.
So, we can find out the purchasing power parity which is the gross domestic product percapita, we can measure parameters like your number of telephone lines per capita or perthousand persons, we can also find out measures such as vehicle ownership. So, these aresome of the metrics which are used to assess the urban growth pattern.
(Refer Slide Time: 24:17)
Now, let us see another study, wherein the these metrics have been used to assess the growthpattern in Shanghai in China. So, this is a study by Feng, Liu and Tong.
So, I mean this has been published in ecological indicators. So, this particular article studiesthe spatio temporal variation of landscape pattern and the spatial determinants in Shanghairegion in the last twenty years and it uses a exploratory regression and generalized additivemodel. So, we see that this particular region of Shanghai is fragmented in the sub urban areasand in the far suburbs it is more aggregated. We also see that some of the parameters of your
metrics, landscape metrics are multi collinear in nature. There is collinearity between thevariables that we had studied if these people had studied.
So, this collinearity was eliminated, I mean those variables which were collinear and whichwere give me giving similar kind of relationships, some those variables were eliminated and thedominant spatial factor for each landscape metric had been identified in this particular study.
So, the sort order of the factor and the accumulation of the residual deviance where also Imean studied to quantify effects of factors on the different landscape patterns. So, they alsoworked on the distance, I mean this was one of the factors which the assist or that is thedistance to outer ring expressway and subway stations inside outer ring expressways whichwere found to be the most influential to the landscape patterns.
(Refer Slide Time: 26:37)
So, we can see that these are the different measures, which were used to assess the proximitycharacteristics. So, the distance to the city centre, distance to the district centre, distance tothe main road in 1995 and 2005. So, these measures were used to identify the proximity to theroads then there are measured such as D ssi and sso which gives you a proximity to thesubway station and which are distributed in the city and in its nearest suburban areas. Theyalso measured the distance to the ports which is measured using this particular I mean metricD, I mean variable D pt. So, and then distance to the protected area that is ecologically seeksignificant areas.
And what is the distance in terms of proximity measures, where I mean tabulated using this Dpa measure. So, you can see the how these measures are dispersed in the city of Shanghai. Thefirst one gives you the measure of the distance to the centre of the city which is a buffergradient buffer and we had already talked about creating this different types of buffer maps inour earlier lectures, where we were talking about raster and vector data modelling and I meanyour different operations using raster and data, vector data where being I mean studied.
So, you are already familiar about it. So, these processes can be implemented to study thedistance to the city centre, you have the distance to the district centre in this case, distance tothe road in different time periods, then you have distance to the outer ring road, express road,the suburban express road, then you have the distance to the subway stations so distance tothe ports and the protected areas, distance to the protected areas, all these factors werestudied and then what was what was done was the generalized additive model which is givenby this particular equation, where in your beta in this particular equation is a constant.
(Refer Slide Time: 29:03)
And on the right hand side we have DI which is the ith variable and we have another errorterm which is the model residual which could not be accounted by all these differentparameters is written as delta. So, this g M is a link function that represents the effects ofvariables on the metric, on a specific metric, on the different each of the metrics we canidentify what is this what is the link function that gives you an idea regarding the impact ofthese various variables that D 1, D 2, D 3, D 4, etcetera that we had just seen.
(Refer Slide Time: 30:03)
This has been identified and you can see that it was implemented for 1995 and 2015 inShanghai, where in you can see different land use categories and how they have been evolvingover time and you can see that slowly the low density built up areas have been permeating inthe outskirts around this nodes, where in you had high density built up areas and then youhave the core which is expanding and there is infill of the core which is getting compact. Wehad talked about the four different types of urban growth models.
So, in this you can see the central core is becoming compact and it is evidenced from the studyusing landscape metrics.
(Refer Slide Time: 30:55)
So, we can do this assessment of studying the landscape, different parameters, the differentmetrics. So, for 95s and 2015 the landscape largest patch index, perimeter area fractaldimension, interspersion juxtaposition index, number of patches and the Shannon diversityindex was calculated for 95 as well as 2015 and you can see the results for the different areas.
(Refer Slide Time: 31:36)
So, this led to the relationship between the earlier model kind of a regression model that wehad talked about. So, it gives us the kind of dependence of the function factors that we hadtalked about that is the distance functions with respect to the spatial metrics. So, how it variesbased on a fraction of a factor of distance can be studied over here. So, these are differentranges of this factors and these are the mean values of this parameters, your the landscapemetric parameters with respect to the different variables, the distance function that we hadtalked about earlier.
So, these landscape para patterns were aggregated in the in the shanghai centre as evidencefrom the this particular study. The, if the fragmented in the nearest suburban areas this patchesare fragmented in the suburban areas and it is aggregated in the distant suburbs. So, these is
also explained by the dominant factors such as the built up area or the infrastructure that weare I mean assessing like the outer ring road or the roads and the presence of subway stations.
So, the statistical significance of each of these spatial variables, it estimated the relationshipsbetween these different parameters and it could be used as a measure to address the differentissues in urban planning regarding density control. So, this I mean particular study it providesan insight, I mean using landscape metrics, it would give us an insight into the potential impactof a new infrastructure development like you have an outer ring road or you have a suburban Imean your M R T S corridor. So, you can get an insight on the impact potential impact onyour urban pattern due to the infrastructure that is going to come up. So, it can be used forsimulation.
(Refer Slide Time: 34:11)
So, a recapitulation of what we had covered today. First, we had talked about the Urbansprawl characteristics. In that we talked about metrics to assess the urban growth pattern. Wehad also talked about the patterns of urban growth in this particular lecture and then we didurban growth analysis for two or three different I mean urban regions.
So, you can read on this and there are lot of different studies doing different types of analysisand simulation on growth, urban growth. So, I mean we shall look into, the different othermodelling techniques in our next lecture.
So, thank you so much.
Log in to save your progress and obtain a certificate in Alison’s free Urban Planning Criteria online course
Sign up to save your progress and obtain a certificate in Alison’s free Urban Planning Criteria online course
Please enter you email address and we will mail you a link to reset your password.