6.8      Analysis in ArcView GIS


6.8.1    Introduction

The stages for developing application to carry out demographic spatial analysis in GIS using ArcView GIS as an example; the specifications to carry out demographic spatial using ArcView GIS Avenue are the following documents, which describe what the application will accomplish

·        The required data include: parcel information which is recorded in the lots table, table containing personal information and the building attribute table

·        Read records from the table containing personal information and the building attribute table

·        Then to select the records/fields containing the buildings coordinates (both eastings and northings for each corner) and building identifiers from building attribute table

·        Create a new table where the selected records are stored

·        Join the newly created table with the table containing personal information, using building identifiers as the common key.

·        The queries to carry out in the application about population display include:

*           Location of individuals (Where s/he lives)

*           Who are the neighbours?

*           Who are the members of the same household?

*           Ethnic group

*           Age bracket of an individual

*           Sex category of individual

*           Martial status

*           People living with a selected individual

*           Relationship between members of the household

*           Displaying according to household

*           Display according to building

*           Display all population

*           Clustering of above

*           Combination of some of the above to find common characteristics

*           How the above can be view at different scales

In Prototypes, I begin with customizing of the interface controls and project components so that the above analyses are easy to perform. Identification of the required scripts, extensions, and writing a short description of each then develop the more scripts and extensions required for the application. In structure testing, the aim is to identify application defects the following levels of testing are used:

·        To ensure that each script behaves as expected

·        System testing to determine whether various scripts work together properly

·        The user acceptance testing is carried on the data from the study are to determine if the application meets the needs

 


6.8.1.1      Analysis preparation

It involves getting a base map, in this case the base map used is that of the cadastral GIS of the study area, on base map overlay the building layer on which all the spatial analyses are based.

The population analysis is first carried out in database using Microsoft access as the database management system and also in SPSS as statistical analysis software.

Population data from these packages is either exported from these packages using the export functions or it is imported into ArcView GIS using the accessing tabular data capabilities of ArcView like the SQL connection

After importing population files they are georeferences to the buildings as the reference spatial units using ArcView geocoding styles.

To aid in spatial planning the points are spread randomly within the boundaries of the polygons (buildings or zones). It done using the script random points in the spatial analysis menu when the view is active

When points are spread randomly they may be on top of each other, this does not give good visualisation of the density. To accomplish that use a script displace points, which is attached in the spatial analysis menu when view is active, so that each point appears with a unique spatial reference to aid in spatial planning.

By the aid of script random sample it is possible to vary the number of features in the theme, so that we can test what would look like when some individuals are taken from the place.

When it comes to evaluating alternatives by spatial varying, this is done using script displace point which can change the spatial location of features.

In this point analysis to avoid the problem of points in the display having the same constant attributes after the construction of the database we use One to many script which helps to link the personal attributes to the randomly displaced points created using the random point, random sample or displace point script.

Some times it important to carry out progressive polygon to point analysis, it is done using the transfer convert item in the X-Tool under the view can convert polygons into point to add to another point theme.

In order to add on the capacities in making decision there is need Linking to images and pictures, done using the hot potato extension which has menu tools on the view.


6.8.1.2     Individual display

Accomplishing the steps below enables to displaying population according to individual location

1.      From the building table read the building identifiers (B) and building numbers (Bn)

2.      Read the coordinates of each building, find the maximum (max), minimum (min) and range for eastings (Reast = Emax - Emin) and northings (Rnorth = Nmax - Nmin) for each building.

3.      From the table containing people’s information read building number (Bn), householder (H) and all the rest of the table.

4.      From the data in step 3 for each household find the total number of members (H) of the household (persons with the same householder identifier)

5.      Assign each member a number (Hn) staring from 1 up to H.

6.      Using the range, minimum and maximum for the coordinates of buildings from section 2 calculate the coordinates of each individual according the household. Persons easting (Pe = Emin + (Reast/(H + 1) * Hn)); persons northing (Pn = Nmin + (Rnorth/(H + 1) * Hn)).

7.      Add the coordinates for each person (Pe, Pn) to the table containing people’s information.

8.      Locate each person on the view using coordinates for each person (Pe, Pn) or added as a theme.

This can be accomplished in ArcView GIS in order to aid in spatial planning the points are spread randomly within the boundaries of the polygons (buildings or zones) done using the script random points in the spatial analysis menu in a View document graphic user interface. When points are spread randomly they may be on top of each other, this does not give good visualisation of the density. To accomplish that use a script called displace points, which also in the spatial analysis menu, so that each point appears with a unique spatial reference. In this point analysis to avoid the problem of points in the display having the same constant attributes after the construction of the database we use One to many script which helps to link the personal attributes to the randomly displaced points created using the random point, random sample or displace point script.


6.8.1.3     Spatial nearest neighbour analysis

GIS makes it easy to perform Spatial Nearest Neighbour Analysis, it helps in understand the how close or separate individual are. It is mostly termed proximity analysis; it is one way of analysing locations of features by measuring the distance between them and other features in the area. The distance between point A and point B may be measured as a straight line or by following a networked path, such as a street network e.g. in this study it is used in cluster analysis.

Special with using GIS is that we are able to get both the empirical and visual results simultaneously, in addition to being able to carry out the nearest neighbour analysis interactive with any spatial size according the analyst (planner) choice say one metre or 20 kilometres.

In this study, this method is tested on relationships in demographic data, i.e. to access the spatial/geographical relationship between the members of the same ethnic groups since the persons are geographical represented by point features. Each household is assumed to have members of the same ethnicity and each household to belong to different ethnic group and to be located randomly. Also in another investigation each household is represented by a single point feature, it is this point feature, which is used in the analysis. This test is to check if households of the same ethnic group live together or other wise.

This is demonstrated in ArcView GIS using nearest neighbour analysis script attached a new tool in a View document graphic user interface (ViewDocGUI). It is done by just dragging a rectangle around the features you wish to conduct a spatial nearest neighbour analysis of.  The wait cursor will appear and then a message box will tell you the R-value and how many features were accounted for in the analysis.  R-values relate how clustered or dispersed points (or centroids of polygons and Polylines) are within the rectangle you specified.  An R-value of 0 (zero) indicates an intensely clustered pattern, while an R value of 1 indicates a random distribution, and an R-value of 2 (or higher) indicates strongly dispersed or organized pattern.

The above spatial analysis combined with other tools like the random point, random sample, displace point script and others gives the planner a wide variety of alternatives and to evaluate them interactively by changing the spatial distribution and testing the alternative.


6.8.1.4     Cluster analysis

Cluster analysis is most for determining homogenous areas. There exist many methods of clustering like measure of similarity, Iterative clustering, and agglomerative clustering (Plane, 1994). All these methods are not appropriate for this study as they consider only clustering between areas/regions. The measuring or carrying out cluster analysis used in this study a method I have named spatial progressive similarity.

This is demonstrated in ArcView GIS using nearest neighbour analysis script attached a new tool in a View document graphic user interface (ViewDocGUI). It is done by just dragging a rectangle around the features you wish to conduct clustering upon; a wait cursor will appear and then a message box will tell you the R-value and how many features were accounted for in the analysis. This is also done with other group of features/individuals, by comparing the R-value then decision either the join them into one; the process continues until we reach a level of clustering set.


6.8.2    Decision-making

6.8.2.1     Demographic variation

A prototype in ArcView has been on assumption that two areas will have same aggregate population change but with different demographic composition. After the aggregate population change area 1 and area 2 will have the same increase in population of say 20 persons, but area one population will having high percentage of the aged who are spread through out (figure) as compared to area 2 with low percentage of aged population who are only in one corner.

Figure 6.4. Aged are many and spread through out the area

Figure 6.5. Aged are few and concentrated in one corner

 


6.8.3    Performing demographic aggregation

Various demographic aggregations can be carried depending on the need, thus here I will carry out only a few to provide an insight on carrying out aggregation.

6.8.3.1     Creating Zones

Can create zones (polygons) in ArcView to test the method of zoning used in planning or to create zones basing on the demographic characteristics, which can be used for decision-making and resource allocation. It this done in ArcView using the script called convert points to polygons which is embed in the X-tools extension, which joins points to form polygons, then we can use the GeoProcessing wizard which in under the View menu when the view is active; to carry out spatial aggregation using commands like union, merge, intersect, etc by GeoProcessing the crested polygon (zones).

6.8.3.2     Zonal demographic aggregation

Perform the zonal subdivision or use the set ones according to planning regulations, this is input as zonal layer, the other layer is the demographic layer or layers of demographic characteristics.

·        By overlying the zonal layer over the demographic layer this divides the total population according to zones. It should be noted here that also the demographic layer could be a composition of all the demographic characteristics or only one according to the choice and the aim of the analysis.

·        Then add a new field and calculated the demographic characteristic(s) according to polygon (zones).

6.8.3.3     Zonal aggregation according to demographic characteristics

This is done using the GIS techniques of overlaying and buffering, in ArcView it accomplished by employing ArcView GIS tools of intersecting, clipping, union, etc.

·        By clipping the zonal layer by demographic layer to get only the zones with same population. It should be noted here that also the demographic layer could be a composition of all the demographic characteristics or only one according to the choice and the aim of the analysis.

·        Then intersecting the two layers i.e. the demographic data theme (layer) with the zone theme to obtain a theme with combined features.

·        Add a new field and calculated the area and perimeter of polygon(s) formed as a result of clipping, this gives the new zones.

Using the scripts called aggregation to summarize population data according to different demographic characteristics in a way of looking for the all over similarity like what is the dominant race in this building, what is the average, mode, median age in this building or zone

6.8.3.4     Aggregation to obtain population density

This starts by zonal demographic aggregation like in the above section i.e. perform the zonal subdivision or use the set ones according to planning regulations, this is input as zonal layer, the other layer is the demographic layer or layers of demographic characteristics.

·        By overlying the zonal layer over the demographic layer this divides the total population according to zones. It should be noted here that also the demographic layer could be a composition of all the demographic characteristics or only one according to the choice and the aim of the analysis.

·        Then add a new field and calculated the demographic characteristic(s) according to polygon (zones).

·        The obtained demographic characteristic(s) according to polygon (zones) is then divided by the area of polygon to obtain a proportional value, which can then apply to the original demographic data.


6.8.4    Evaluating alternatives

6.8.4.1     Evaluating alternatives using coordinates

When it comes to evaluating alternatives of demographic spatial distribution, some times it is important to use coordinates. ArcView makes the process simple, where by use the script called add xy coordinates whose action I have attached to the menu called spatial analysis, which was added to ArcView user interface under the view. By opening the attribute table and making changes to the coordinates and the geocoding again to the buildings or other polygon features created or adding them as event theme under the view menu, can visualise the effect of the change and make decision basing on that for planning purposes.

6.8.4.2     Evaluating alternatives by editing features

By moving points (people) to different or better position in order to evaluate the alternative for planning purposes; GIS provides tools to quickly do that, by editing features in the view using the hot potato extension in ArcView, and then updating the new locations in table using the script calculate and update coordinates under the spatial analysis menu in view, to reflect the change, this can be exported to packages like SPSS for analysis.

6.8.4.3     Evaluating alternatives by varying the numbers

In planning to make alternatives some times it involves varying numbers to evaluate the effect. In GIS this effect can be visualised by using ArcView script random sample giving the power to the analyst to select the percentage to sample, then displaying it.

To further the analysis by varying numbers the script random point to display points randomly, then by script displace point to avoid points to lay over each other.

 


6.9      Presentation in ArcView GIS

We are able to combine the traditional presentation of using tables, charts, and graphs with spatial location of the event. In this case the location is being tested at individual level.

6.9.1.1     Cursor Query

There is interactive Cursor Query. This enables to access the attributes of what is being displayed in the map. This has been utilitised in this study to be connected to the graphs and charts of the traditional display.

6.9.1.2     Windowing

Within GIS it possible to provide many views at the same time, providing many alternatives at the same time and many perspectives for the planner. This further enhanced in that all the difficult alternatives can be linked, thus by changing one the effect is reflected in all the others.

Another provided by GIS under windowing is Zooming in and out at any time and scale, enabling to check details by zooming in and show at aggregation by zooming out.

6.9.1.3     Layering

By overlaying different demographic characteristics for specific area we can see the spatial lay to help in spatial planning.

6.9.1.4     Relationship

It has been difficult to represent individual population, depicting relationship between them. Various methods have been tried which include showing persons in terms of pictures to show that there young and old (see figure 6.3).


6.9.2    Visual comparison of demographic data

Using ArcView GIS as a means of visualisation, not only to depict the geography but also to explore the correlates, thus as means or form of exploratory data analysis in search for correlates of demographic entities. In this case different colours are used to depict change and variation in demographic characteristics and indicators.

6.9.2.1     Visual comparison of percentage of aged

Select population over 60 years and display on a map to be compared with the total population. This can be done by using either one layer which is obtained by overlaying two layers of the aged and the total population, or by using two independent layer in two view one showing the aged and the other total population.

6.9.2.2     Visualisation of age cohorts

By taking each age cohort, giving it different symbols or colour, display them on a map. This can be done either on the same layer which is obtained by overlaying the different layers of the age cohorts, or by using separate layers each showing specific age cohort in different views.

 


6.9.3    Case studies of demographics

6.9.3.1     Aged Care

Where do you plan to put an Aged Persons Community? Obviously you need to develop Aged Community infrastructure in areas that are close to the facilities and services that aged people require. Some of these facilities would be doctors and hospitals, public transport, retail services and appropriate community services. Another scenario would be to develop infrastructure services in areas where there is an abundance of older or ageing population.

Demographic Data and a GIS can be used to answer both of these questions. Analysing Aged Population Density and then overlaying files containing the location of Hospitals, Doctors and retail services will enable us to understand the comparative results of our query; areas where action needs to be taken, or areas where there are opportunities become clear.

From figure 6.6 showing land use in the study area, it can be seen that commercial takes the biggest percentage followed by religious and cultural; with education, open space, and utilities occupying little of the space. The following graphics figure 6.8 (Aged figure 6.7 overlaid onto land use) show aged persons in relation to land use. The area being good fro aged as there are open spaces, religious and cultural facilities.

Figure 6.6. Land use in study area

Figure 6.7. Aged population

6.9.3.2     Overlaying

Figure 6.8. Aged populated overlaid on land use

6.9.3.3     Combined analysis

Figure 6.9. An effective application for Demographics and GIS

This type of output along with tables substantiating the numbers, including graphs and reports can be simply and quickly created with products of this nature.

6.9.3.4     Buffering

By making a 250 meter buffer around the open space it can be seen that every aged person within the study area will access to the open space with a sort walking distance of only 250 meters (figure 6.10).

Figure 6.11. Buffer of open space

By making a 50-meter buffer for each aged person it shows that every aged will have a neighbour with in the same age cohort (figure 6.12)

Figure 6.11. Buffer of aged spatial neighbourhood


6.9.4    Demographic Indicators

In analysis demographics for planning there arises reasons which nee the spatial aspect like in which community do find the dominance of certain race group, in the area what is spatial is age distribution, which households have above the average persons per room, children per woman.

6.9.4.1     Indicators on population

(Malaysian figures obtained from population division and statistics division of United Nations secretariat, 1999; Penang state figures from the state population report, Pulau Pinang, 1991)

 

Total

(In 1000)

Male

(In 1000)

Female

(In 1000)

Sex ratio

(Male per 100 female)

Density

(Per sq. km)

Study area

 

 

 

 

 

Penang state

1116

 

 

98

1083

Malaysia

21830

11065

10765

103

56

6.9.4.2     Indicators on youth and elderly population

(Malaysian figures obtained from population division and statistics division of United Nations secretariat) (1999)

 

% Of total population under 15

% Of male population aged 60+

% Of female population aged 60+

Sex ratio (Male per 100 female) in the population aged 60+

Study area

 

 

 

 

Penang state

 

 

 

 

Malaysia

34

6

7

90

 

6.9.4.3     Persons per household

In this study the procedure is to carry out a query to determine persons living in the same household, summarise to obtain the totals for each household to obtain showing the number persons per household (pop-household), join that table with table of attributes of the building, then using the auto label function in ArcView GIS under the theme menu to display the number persons per each household on a map (figure 6.13).

Figure 6.13. Persons per household

6.9.4.4     Average age

Average age for each household, the procedure is the same like for persons per household replacing number of persons per household by the average age

6.9.4.5     Children per woman

 

Children per women use same procedure like for persons per household replacing number of persons per household by the number of children per women

6.9.4.6     Persons per room

 


6.9.5    Area of polygons

GIS performs several basic geographic operations that simplify the geographical analysis of demographic data, which elementary algorithms for calculation of polygons areas. Examples of application in population analysis that require estimates of the areas of regions include

·        The calculation of demographic characteristic density

·        The derivation of the index of concentration of demographic characteristic

·        The estimate of demographic characteristic in areas for which data are not collected

The estimate of demographic characteristic in areas for which data are not collected (to find information for user-defined geographic territories) this requires some type areal interpolation or extrapolation to transform data from a set of administrative regions to a set of target regions defined by the planner. For example this study a planner may wish to know the number of young persons (between 12 and 30 years) who will be able to use the youth centre, who lie within 100 metres from it.

Using ArcView GIS to make a query to select persons aged 12 to 30 from the population table, geocoding it using their residence i.e. building where they stay, making a buffer of 100 metre from the youth centre, then clipping the geocoded theme of persons aged 12 to 30 with the buffer theme of youth centre to obtain a theme of persons aged 12 to 30 within 100 metres from the youth centre, which is only 204 persons see figure below.

Figure 6.12. Population aged 12 to 30 100 M from youth centre


6.9.6    The estimation of population data for user-defined regions

The estimation of population data for user-defined regions lets the user define the area and GIS directly assisted the area with the attributes in the database. The form dasymetric mapping (Plane, 1994) allows incorporating the knowledge that population must equal zero in water bodies, parks, and other non-residential areas. Whole technique is that the non-residential areas to be eliminated from consideration are subtracted from areal measurements.

Taking the study area if the planner is interested in knowing the density in the total area being used for commercial. Start by creating a land use summary table, then taking the total area minus other land uses to get the area of commercial usage (32547.64 Sq metres). Dividing that area by the total population we get density as 21 persons per sq metre.


6.9.7    ArcView GIS Avenue

Using ArcView GIS Avenue facilities, it has been made possible to create customised button driven front-end system, which provides an environment for grouping together several actions into one selection procedure. It has been done with Avenue where scripts call each and run other programs, display graphics using links. With the use of Avenue message box (Msgbox) allows to input data by user to be verified before proceeding to the next stage. There is pop up window for displaying of test files, tables, graphics, pictures relating to results of analysis, as pop up windows do not totally obstruct the main view.