megomechomeconceptdataspatialoperationalref

Methodology

A flow chart may be used as an aid in order to illustrate the following processes of the site selection analysis:
image


Preparation Step:

From the census data tables that I aquire it is determined that the most useful one in our analysis included the information of population density and average household income in 2001, therefore, preparation processes include joining census data tables to the Kelowna Dissemination Area (DA) layer. A new shapefile is also created in order to digitize various locations of competitors. In addition, It is necessary to digitize various points in the landuse layer with reference to a hard copy map and research for the locations of competitors located in the Kelowna region. 


Query:

        Queries offer a method of data retrieval, and can be performed on data that are part of GIS database, or on new data produced as a result of data analysis. Two types of queries I have used for this analysis: spatial and aspatial.
        Aspatial queries are questions about the attributes of features. For example, ‘how many topographic coverage areas are over 118801 populations?’ on map 1 is an aspatial query neither the question nor the answer involves analysis of the spatial component of data.
        Spatial queries are the question or the answer requires information about “where”. For example, where are the existing competition locations? Furthermore, in this project, queries are made more complex by combination with questions about distances, areas and perimeters. For example, I have asked “where are the regions contain high average household income with high population density in labour force?" on map 10.
   

        Boolean operators are often used to combine queries to identify entities in a database that satisfy two or more spatial and aspatial criteria. In GIS, Boolean expressions are AND, OR, NOT and XOR. In this analysis, I have asked these following Boolean operator questions:

A And B (intersect)

intersectWhich the regions are in the category “potential sites”
and regions which are within 100 meters distance
 of major roads and highways?



A OR B (union)

union

Which the regions are in the categories
of vegetation and park & recreation?


A NOT B

Xor
Which regions are in the category
open area but do not have vegetation cover?

 

Buffering:

        Besides queries, buffering functions that are used in this project allow a spatial entity to influence the character of an entity. Buffering is the common method to create a zone of interest around an entity.

        The question “which potential areas are within 1000 meters buffer zone of the existing competition stores?” is being asked for the map 11. However, the first step of this approach is to produce a buffer zone identify all land up to 1000 meters from seven existing competition stores.  A circular five rings buffer zone is produced and each ring indicates 200 meters buffer distance. The second step is to find out which potential areas fall within the buffer zone. In addition, the suggested locations of the new MEC retail store are based on the catchment concept that its buffer zone must not be overlapped on other competitions’ buffer zone.

The next step is to specify the land use in the area, with particular interest in potential areas to locate our new store within. Seven landuse categories, which are Commercial, are ready for query (Map 4).

The fundamental concept of the query for landuse is to separate potential and non-potential areas based on figure 3.


figure3

Figure3

Non-potential areas include the landuse categories of Government & Institutional, Industrial & Resource, Park & Recreation, and Vegetation. A union of vegetation, park and recreation is performed becasue they have the same function. In addition, existing commercial areas are counted as non-potential site because the land price is relatively high that will increase the fixed cost. Finally, a union of open areas and residential areas is performed in order to create a new potential layer that is more useful in my analysis.

In order to refine the potential area, a query is performed to select potential areas only outside 100 meters of selected BC railway (Map 6). Also, another query is preformed in order to select potential areas within 100 metres of major roads and highways (Map7). The main reason is that major roads and highways link to the rest of B.C. in all directions, therefore a location that is visible from those roads and highways is crucial, as many MEC members are from outlying areas.

In addition to land use, major roads and highways, and the competition layer, dissemination areas are added along with the census data that have been previously joined (average household and total population density in the labour force). A query for high average household income is made, resulting in the new layer. This is followed by an intersection of the potential site selection layer and the high population density in the labour force layer, creating a high-density household income layer.

It is necessary to perform another intersect of this new layer with the 1000 meters buffer of competition so that selected polygons within the buffer are excluded. The result of the buffer further refines the potential sites by excluding areas too close to competition locations. As well, another 1000 meters buffer distance from the existing competition buffer is added in order to avoid the overlap of the new MEC store's 1000 buffer distance. The final layer is titled Ideal Locations for New Store, as it is the result of the queries, selections and analysis of the previous maps. 

For the final map I need to decide on a location for the new MEC store from the 8 remaining ideal locations based on a visual inspection analysis. The first criteria I use is to eliminate one of the possible sites included any areas that are surrounded by low population density in the labour force. This subsequently eliminates the ideal polygons at the intersection of Rutland Road and Highway 33 W.
 
    The second criteria that I consider included areas with close proximity to commercial polygons. Therefore the store will remain located in a region where local population normally purchase their products at these areas.  After the visual inspection of Map 11, the final condition consists of an individual area that has an ideal physical size allowing MEC to choose an exact location within the specified area.  If this area is not preferred by MEC, the adjacent smaller polygon will also be suitable based on the above criteria. The suggested location of the new store is represented on Map 12 with a star and has a buffer of 500 metres that indicates the parameters described above with reference to competitors.


back to top

_____________________________________________________________________________________________________________________________
Site Design by Benjamin Lee, Department of Geography
Simon Fraser University
8888 University Drive
Burnaby, B.C.
Canada. V5A 1S6