Spatial
Analysis
- Database
Querying
- Buffering/Boolean
- Fuzzy/Distance
- MCE
1
Database Querying
 |
 |
| The image above shows census tracts
with a medium and high proportion of families with children
(medium is from 50-59% inclusive, and high is 60% or higher).
The database filtering command used is: [%fam_w_children] >=
50 |
This image shows census tracts where
the median household income values fall into the medium and
high income categories which I define as $35,000 to $55,000
for medium income and greater than $55,000 as high. The following
command was used for the filtering: [med_privH_income] >=
35000 |
2
Buffering/Boolean
This image was used to mask out the areas not part
of the municipality of Burnaby in census_tract96.rst. RECLASS was
used to assign all Burnaby census tracts the value of 1.
A simple reclassification where all commercial regions are 1.
A 100 meter buffer was set around the commercial areas to go with
the criteria that restaurants can be allowed to locate at the most
100 m from the commercial zone. This image used later as a constraint
for the MCE.
3
Fuzzy
This image was generated using FUZZY with a linear distance decay
function and control points c = 0 and d = 6000. The regions of highest
suitability are dark green. It cleary shows the high population
density regions and equal concentric rings radiating outwards.
This image generated using FUZZY with a Sigmoidal monotonically
decreasing function and control points c= 100 and d = 1000.
This image generated with FUZZY using a decreasing J-shaped curve
with control points c = 30 and d = 200. The first control point
c was chosen to be 30 m distance since any restaurant located within
this distance of the major roads has optimal access to customers,
but beyond this point suitability begins to decline from the maximum
suitability value.
4
MCE
This image is created by MCE using WLC and having only the constraint
of burnbool which eliminates areas not of Burnaby. This image is
highly informative since it shows suitability based upon all criteria.
The continuous spatial distribution of suitability based upon the
factors can clearly be seen. The green shaded regions in a way defines
the high potential market areas for the new restaurant. One can
see three distinct regions of high suitability. There is the one
in the east which extends into another municipality (Coquitlam),
and then there are the northern central region and the southern
central region.
This final image uses the MCE with the constraints of burnbool.rst
as well as combuff.rst. This knocks out regions that are not within
the desired proximity to commercial centers. Then a RECLASS was
done which set all values above and including 250 to be 1 and the
rest the value 0. This was the more desireable method of presentating
the final image of potential sites since it shows clearly defined
areas of high suitability. This image should also be complemented
with the image above it to present a better understanding of the
market area. For this image I overlayed the major roads and restaurants
layers on top to help others see the spatial association. One can
clearly see that all competitor restaurants are easily accessed
by major roads and are within commercial areas.
|
