As discussed in Methodology, the first step in assessing current density was to isolate the Residential Land Area (RLA), excluding all the land-uses which could not accommodate current and future populations. The map to the left shows the resulting Residential Land Area in the City of Vancouver. Clearly a large proportion of areas outside the central area are principally residential land which will be investigated in its potential and appropriateness in absorbing more density. Parkland has led to large areas being excluded from current or future population accommodation, notably Stanley Park north of the West End and the parkland in the University Endowment Lands (UEL), as well as, several areas in the south and southwest of the City. A large proportion of Downtown and areas to its east and south are not currently zoned for residential as a lot of this land is either commercial office space or industrial land. Although these areas could be converted to residential in the future, they are importantly excluded since the stated goals of Smart Growth and EcoDensity is to ensure that jobs provided by these areas are maintained in the city and in close proximity to where people live, instead of moving industry into the suburbs which would defeat the purpose of creating less auto-dependent live-work neighbourhoods. The Residential Land Area is shown in divided blocks since all the roadways were excluded as well withh the assumption that these would be important to preserve, even if they are converted in the future to accommodate less cars such as into more transit lines, or even into green space.
There are of course various potential sources of error in the measurement of the RLA. The first one is based on the actual DMTI land-use data used which contained some apparent misclassification. Although there was correction for this by combining aspects of the 2001 and 2006 data, some errors could remain. Laneways were also not included in removed roadways as this data was not available in the GVRD street layers. Furthermore, the actual widths of the roads were averages based on road class and therefore these could be variable between different neighbourhoods and create certain small calculation biases. Finally, in terms of the actual significance of the RLA, it is impossible to determine whether all lands that are residential will be maintained that way in the future and that industrial and parkland won’t be converted to housing and, therefore, one could argue those lands also hold potential for densification. However, since the purpose of the project is to analyze density patterns with the assumption of wanting to preserve a similar land-use and zoning structure as existing today, it is assumed that looking at the selected RLA provides the most useful analysis method.
The map on the right shows the overall population density map by DA, but is based on people per Residential Hectare (RHA) rather than the entire DA land area. The difference from the standard census density data is that this map shows the population according to only the land area that is potentially residential land and excludes all other land-uses which have been determined important to preserve. The most striking differences can be seen in areas such as Stanley Park (north of the West End) which is shown in this map as having a very 'high' density. The reason for this is because this Stanley Park DA only includes a very small strip of Residential Land on the edge of the West End and although there is also a very small population there, in terms of population per RHA, the DA is very dense. In other words, this helps us read the population distribution and see that, although there are not many people in the Stanley Park DA, there is also not any available residential land either and therefore the density is basically ‘maxed out'. This is similar in DAs along the industrial and port lands northeast of the Downtown which have a relatively small population but also a relatively small residential land base.
The overall pattern of current population density per RHA in Vancouver shows a clear concentration of density in the Downtown peninsula and the surrounding ring neighbourhoods around the central core, however, with the notable exception of the areas adjacent and directly to the southeast of the Downtown, which represents the still undeveloped areas of Northeast False Creek and the False Creek Flats, as well as the Southeast False Creek area currently under development. These areas are appropriately shown to have a large potential of residential and undeveloped land that could hold more population but are currently empty, which is unusual considering how close they are to the downtown core.
major pattern to be seen is the characteristic east-west divide which
is apparent in the population density on both
the city. Looking at the map, the Eastside
clearly has a
higher population density than the Westside of the city. However, there
few areas of relatively higher density on the Westside other than along
Broadway Corridor/Kitsilano area, most notably in the Kerrisdale area
and Arbutus. The Marpole and Oakridge areas also have a large
amount of density compared to their relative locations.
Other areas of the Westside have generally a very low relative density
area that exemplifies this the most is Shaughnessy. Shaughnessy is
relatively close to
the downtown area and, although the Fairview neighbourhood directly to
almost entirely high density, the population per RHA
abruptly drops at the Shaughnessy border. This is partly to do with the
was built there when the neighbourhood was first established,
comprising of large wealthy
residences, and also to do with current
protective zoning which doesn’t allow much redevelopment in the area.
However, in comparison, the Riley Park-Little Mountain neighbourhood,
on the Eastside and roughly at the same distance from the Downtown as
Shaughnessy, is much more dense and becomes less dense gradually moving
the centre in what appears to be a much more 'organic' pattern. Due to
the nature of the
neighbourhood and the protective
Shaughnessy seems to act as an actual 'block' to development, which
explain the difficulty for density to continue beyond Shaughnessy into
Kerrisdale and Dunbar neighbourhoods as there isn’t a continuous or
outward intensity from the central core.
Although population per RHA showed the overall density in the DAs, by looking as well at the dwellings per RHA it was possible to see the structural distribution of units in the city. In densification plans, there are various proposed ways of increasing populations in the city. The EcoDensity initiative proposes in their 'draft actions' various measures ranging from rezoning that allows apartment buildings and rowhouses to allowing infill houses on laneways and extra suites within houses. Looking at dwelling units per area, we can see where areas have lower dwelling density even if they perhaps still have high or medium population densities. Although this isn’t as direct and important as looking at overall population density, it can be used as a tool to examine where potential extra units can be added rather than where population is of lower density. Areas of low unit density represent neighbourhoods of generally single family homes and therefore are more appropriate for perhaps infill housing or limited rezoning to allow some apartment buildings. On the other hand, areas of high unit density generally represent smaller units within apartment buildings and might be able to absorb extra density through allowing more building height.
The map shows a very similar pattern to the population density as would be expected, although there are a few slight differences. The overall distribution is more even in a gradual lowering of density as one moves away from the Downtown core (with the notable exceptions mentioned previously) and this might perhaps point to a structurally more controlled building environment where units are built in concentrations around the central area. Actual population distribution is, however, more influenced by specific areas’ desirability as certain neighbourhoods that are more popular or important in terms of amenities and accessibility are more likely to attract people who might be willing to share units even when there are less actually built in the area than the population demands.
This following map uses census data on average bedrooms per dwellings by DA, total dwellings per DA, and population per DA to look at the average bedrooms per capita per DA in Vancouver. This map is able to reveal, beyond the other density maps, the current use of existing structural capacity by area in the city. Using bedrooms as a gauge of structural dwelling potential, one can determine where areas of the city have possibilities of accommodating more within the existing structure. This is relevant to ideas of increasing allowable suites in existing houses in the city as proposed in EcoDensity and, therefore, this map shows where areas of the city are in fact under-populated compared to the available bedrooms that are already built. The map clearly shows that the Downtown and areas directly surrounding the CBD have a much lower number of average bedrooms per capita, often under 0.6 bedrooms per capita meaning there is less than one bedroom per person. This is likely explained by the fact that many couples would probably be sharing one bedroom apartments or some people would be living in bachelor suites. Outside the central core, there are only a few areas of lower bedroom averages such as the Marpole area which again matches more closely to the numbers in the Downtown. In the lower density neighbourhoods on the Westside such as Kerrisdale, Dunbar, Arbutus and Shaughnessy, however, the average bedrooms per capita is often over 1.4 and in Shaughnessy this goes up to almost 2 per person. This is because people are likely living in large houses with many bedrooms and many of the rooms are perhaps 'unused'. Also, there is possibly a higher proportion of empty-nesters who no longer use many of the extra bedrooms since the children have already moved out. Interestingly, however, there isn’t as clear of an absolute east-west divide in bedrooms per capita and it appears that many residential neighbourhoods on the Eastside also have surplus bedrooms. Considering that a lot of couples could be sharing, even when there is a 1.0 average in an area, this might represent an actual 'surplus'. Therefore this map shows that in most of the surrounding residential neighbourhoods there is a large potential in increasing density within existing housing structure without actually building any new dwellings.
many people who do have large houses and extra
bedrooms aren’t necessarily going to immediately rent out part of their
regulations to allow suites
and making it easier to add them, will at
increase some rental units within houses. The EcoDensity Draft
Actions proposed allowing secondary and even tertiary suites in
This map shows that there is at least the existing structure to support
people within many of these areas and through changing regulations and
promoting suites in certain neighbourhoods, there could at least be
increase. Suites also provide more rental and thus potentially more
housing and should be prioritized in meeting goals of affordability.
There are some sources of potential error in examining average bedrooms per population, as the census data itself was only based on a 20% sample per DA and therefore was not as accurate as other data. Also, some of the DAs did not have any data available on bedrooms and therefore were not included in the analysis, as is seen on the map. Regardless of minor uncertainties, the map is effective at showing the overall pattern of bedrooms per capita in the city and showing that potential for densification within existing structures definitely exists.
The ‘amenities’ that were selected as being positive factors in a neighbourhood absorbing density were, as discussed, parks, schools and community centers, as well as local commercial nodes and transit accessibility. The map to the left shows an overview of the distribution of parks, schools and community centres in the city which were then used in the Multi-Criteria Evaluations.
that was important in an area’s preference for
density over another was its relative proximity to the selected
amenities. Raster maps
therefore created depicting 'straight-line distance' proximity to each
The following four raster representations, show the resulting relative
of the amenities displayed in the above map. It can be seen visually
from the raster
maps that there appears to be a generally even distribution of the
although there are slightly more schools and community centres on the
which is logical in relation to the higher population densities there.
on the other hand, is more extensive on the Westside and the Downtown
Creek areas seem to generally be more lacking in amenities. These
maps provide a visual representation of relative proximities to
however, through the further MCE to be performed the details and
areas that are amenity rich or amenity poor for densification can be
The map to the right is the result of the identification of commercial nodes in the city outside the CBD. The nodes represent only the larger agglomeration of commercial land-use areas that met the set threshold, as discussed in Methodology. The map shows the various smaller commercial areas in the city, which usually are neighbourhood shopping focal points. Areas such as Davie and Denman are seen in the Downtown Peninsula and are considered commercial nodes rather than part of the downtown CBD since they represent a more neighbourhood-like shopping area that people would be more likely to walk to get to rather than necessarily travel across the city to shop and work in, such as in the way that the CBD functions. Broadway Street clearly represents a major commercial node along a lot of its length, as does Hastings and West 4th Avenue. Other smaller nodes are shown in many neighbourhoods and represent areas such as Commercial Drive, Main Street, Kerrisdale Village, Dunbar and Marpole. From the simple spatial pattern, it is clear that commercial nodes cluster in the areas surrounding Downtown such as along Broadway, especially directly south and southwest of Downtown. The Eastside also has a larger concentration of commercial areas than the Westside, at least south of Broadway.
There are certain sources of errors in this analysis, since the interval for polygon agglomeration and the threshold of minimum agglomeration size dictate which areas would be included and, although these were chosen to try to display an accurate picture of zones with enough commercial agglomeration to be considered a ‘node’ around which density could be preferably concentrated, there are no set theories to dictate what thresholds should be set and there is a certain amount of subjectivity involved. Another important point is that, because of the data that was available, these areas represent agglomerations of commercial ‘land-use’ which includes all commercially zoned areas and therefore doesn’t represent the actual intensity and level of commercial activity in each area. However, the intent of the map is to show areas where density could be concentrated and perhaps the future population growth could intensify the use of the commercial land available in these nodes. Areas along Kingsway, for example, have large commercial land areas but these are relatively low-intensity big box businesses, however, with a density concentration in these areas, this available commercial land area could be developed to create neighbourhood-scale commercial nodes.
The adjacent map shows a raster proximity analysis to the CBD and local commercial nodes. This map will be used in the MCE to favour areas close to these commercial nodes as being better for densification (taking into account the other criteria as well). Visually, one can clearly see from the map that the Downtown and areas directly surrounding it to the east, south and southwest are very well ‘serviced’ by commercial nodal areas, especially all along the Broadway and West 4th corridors. More of the Eastside is also within a close proximity to commercial nodes than the Westside as would be expected, although some areas in the northeast, east and south are still quite ‘isolated’ from local nodes.
Accessibility to the CBD
The following is the final map from the complex transit accessibility analysis. The map displays the relative proximity by walking and taking transit to the Central Business District (CBD). In other words, each pixel represents a relative ‘time’ distance value to the Downtown taking the shortest route while walking to the nearest transit and then taking the transit to Downtown. The three Skytrain routes provide the quickest transportation access. Very notable is the influence that the almost completed Canada Line will have on transit accessibility in large areas of Vancouver to the south of the CBD. The Canada Line (represented by the red lines running north-south) will increase accessibility along key points near the station locations. Broadway Station (at the point where the two diagonal Skytrain lines meet) is also a point with increased accessibility to the Downtown relative to its distance from the core. These ares represent good potential densification points as they will promote increased public transit use, since people will easily be able to commute to the Downtown on transit, perhaps even more quickly than by personal automobile. Although some areas in the Eastside such as Broadway Station area have certain levels of commercial development and residential density, other neighbourhoods farther along the older Skytrain lines which still provide high accessibility to the Downtown are still quite low density. More importantly, the Canada Line will clearly bring very increased accessibility to areas that, as seen in the primary density maps, are currently very low density. Therefore there is a lot of potential in these lower density neighbourhoods to densify, especially in areas that are near to other amenities such as parks and commercial nodes.
various sources of potential
errors in examining transit accessibility to the CBD. The transit map
as a relative proximity map and therefore it isn’t extremely precise in
of exact times by transit to the centre. The relative cost friction
was created to calculate the 'relative proximity' used data on average
each transit type (i.e. diesel, trolleys, rapid bus, Skytrain) from a
report and therefore does not take into account actual differences
different lines due to traffic congestion. Another major point of
potential erros is the fact
the map does not take into account transit frequencies which could
and accessibility, however, it is assumed that it is much easier to
frequencies (i.e. adding more buses) than creating new lines or
overall speed. One last source of error, involves actual stop
for all the bus lines the speed was measured for the whole line and
into account the actual location of the stops. For the buses, the stops
to be close enough together to not make it important to specify
location. For the Skytrain, however, a block was put in place (see
to only allow access to the Skytrain at stop locations. Lastly, the
cost frictions do not consider transfer times between different transit
lines whihc would undoubtedly have some effect. Although there
sources of potential error, the map is a useful and fairly accurate
looking at relative proximity to the CBD by transit and is useful in a
prioritize locations with the best accessibility by transit or walking
concentrations of jobs and services.
The Multi-Criteria Evaluation was successfully able to combine the various criteria examined and provide maps showing the overall compatibility of areas according to how well they met the entire roster of criteria and also based on specific weightings assigned to each factor according to its importance. The relative weightings for the various amenities were as follows (see Methodology for more details):
Proximity to commercial nodes
Proximity to parks
Proximity to community centres
Proximity to secondary schools
Proximity to elementary schools
As expected, the resulting map shows the highest matching areas to be in the downtown core which is the result of the importance put on proximity to the CBD in the weighting distribution of the analysis. Areas surrounding the Downtown are also shown as matching the amenity criteria very well such as along Broadway and in the Downtown Eastside and Hastings areas. In neighbourhoods outside the centre, many areas along the north-south streets seem to match the amenity criteria relatively well. The most notable compatible pattern outside of the centre, are along the areas going south of the centre into the Cambie, Riley Park–Little Mountain neighbourhoods and even as far as Oakridge. These areas are well serviced by the various built amenities and parks, as well as being very accessible by transit due to the new Canada Line to be put in place. Main Street which runs through the centre of Mount Pleasant and Riley Park seems to be the focus of an area that fits very well all the criteria. Considering the density map, this area is only at medium density but it would be appear in terms of proximity to many amenities it matches very well the criteria and therefore could be argued for density. Many areas of the East and Westside south of Broadway do not seem to match very well the amenity criteria and similarly both sides of the city seem to have scored relatively low, except for areas along Kingsway and in the Kerrisdale Village node. Although both sides would be considered relatively amenity poor according to criteria for densification, the current density numbers show that the Eastside has a much higher density and therefore proportionally seems to be at a deficit in meeting the criteria. The actual relationships between the areas’ amenity proximity and current density levels will be explored in further detail in the following Multi-Criteria Evaluations.
The next step was to combine the MCE of
Amenities with the actual current densities, which enabled one to see
which areas would both meet the amenities criteria well and currently
have a lower density. The map on the right is the result of this
combined MCE of amenities and current population densities. The map
shows some key differences to the simple MCE of amenities that hadn't
looked at actual current patterns. What this map does, in other words,
is compare the areas that match the amenities best (or the 'ideal'
densification pattern) with the current actual density distribution. In
the map to the left, it is clear that the downtown core has lost a lot
of importance, which is simply because, although it matches the
amenities criteria best, it already is an area of high densification
therefore does not provide that much opportunity relative to other
areas of Vancouver to densify more. However, there are several areas of
the Downtown Peninsula that do match well, such as Coal Harbour along
the shore on the North, Yaletown at the southern end of the Peninsula
and the East False Creek area. The census data in this analysis is of
course from 2001 and these three areas have indeed been the locations
of a lot of recent and continuing redevelopment and growth, which seems
to validate the analysis as at least identifying areas which are
also considered important redevelopment locations currently
by developers and the
City. Outside the Downtown, there are even more areas which
match the criteria and are relatively lower density and therefore could
provide potential for densification. The residential neighbourhoods
which form a ring around the Downtown at a distance of around two to
kilometres could be described to form the 'inner suburbs' in the sense
of what were the first 'suburban' residential areas when the city was
developing. It is these areas which seem to meet the criteria the best
since they are the areas with a close proximity to the CBD and,
especially, on the Westside, are of lower population densities. The map
results show that a lot of areas in these residential neighbourhoods,
especially following along arterial routes and transit lines, match the
well and this is clearly because they have a better access to the CBD
by transit and often form local commercial nodes, which would fit the
criteria of Smart Growth theory and many of the EcoDensity proclaimed
goals well. Northern Shaughnessy stands out as an area that fits the
analysis very well since it is an area of close proximity to the
Downtown, has an abundance of amenities and is currently a very low
density neighbourhood. It must be noted that many of these inner
suburbs are older and therefore more resistant to change and there is
importance in preserving the heritage character, especially in areas
Shaughnessy which currently has restrictive zoning to meet that goal.
The maps, however, simply
show the potential of density in the area and do not dictate the form
it should take which is very much dependent on the desired character of
the resulting neighbourhoods. For example, in character and heritage
neighbourhoods, less intrusive forms of density such as infill housing,
laneway housing, small houses and suites, which were some of the
proposed initiatives in EcoDensity, would likely be more appropriate.
As was discussed, however, these
maps provide decision tools and the focus of this project is not
to look at the actual implementation and form of densifying specific
The Canada Line effect is very noticeable in
the MCE maps since
it brings accessibility in a line running south of the Downtown and
makes areas of neighbourhoods such as South Cambie, Riley Park and
Oakridge very fitting for potential densification. These areas are also
in close proximity to many other of the basic amenities, such as parks,
schools and commercial nodes but currently have a lower density. The
Canada Line will undoubtedly result in some densification of the
neighbourhoods and this above map shows some potential areas which
match important criteria very well.
The map to the left shows the MCE combined with
densities analysis. The pattern is very similar to the population and
amenities MCE. The Downtown does show even more 'saturation' of density
and there seem to be even larger areas that match well the criteria.
However, it was determined that there was not very much surplus data to
be gathered from this analysis beyond what was gathered by looking at
The final MCE combined bedrooms per capita densities with the amenities MCE. The purpose of this step was to show the areas that contain the most current structural potential (measured here by bedrooms per capita) to expand density within existing buildings and compare this to which areas met best the amenities match. The resulting map below shows
that there clearly are some areas that provide this potential. most notably here is the Shaughnessy area which has a very high number of bedrooms per capita and is very well serviced by the selected amenities. Other middle Vancouver neighbourhoods such as Riley Park, Oakridge and South Cambie, also seem to have a fair amount of potential to densify within existing structures. Areas such as Shaughnessy are undoubtedly very resistant to drastic changes but densification through existing houses, such as adding suites, provides a very non-intrusive method of augmenting capacity that if done gradually does not disrupt or change the neighbourhood dramatically. The map also reveals that there are pockets throughout the city which do provide potential for suites such as in parts of Kitsilano, Fairview, West Point Grey, Grandview-Woodlands and Renfrew-Collingwoood.
Best Areas of Potential Densification within the Residential Land AreaThe map below provides a closer look at the major areas that matched the criteria of amenities best and had elatively lower population densities currently. The MCE is overlayed on the layer of the Residential Land Area which thus shows the specific potential to densify according to the analysis' criteria on the actual land that is available for residential development. Areas of Shaughnessy and surrounding the Queen Elizabeth Park (the large gap south of the centre) show the strongest potential. Also some large parcels of land in the False Creek Flats show potential for densification, however this map also reveals that many of these areas directly surrounding Downtown do not in fact have very much land available for residential development. This factor was, of course, included in the previous analises and did not skew the data since the densities were calculated based on only the Residential Land Area.
Norquay AreaThe Norquay Area of East Vancouver as was discussed in the introduction, is a residential neighbourhoods characterized by single-family houses that has been slotted for rezoning and densification. Relating this area to the generated MCEs, it is possible to compare how Norquay fits the criteria of this project's evaluation. The following map, situates the Norquay Area on the final 'MCE of amenities and current population'.
Looking at the above map, it is clear that according to our criteria, the Norquay area is certainly not the most logical area to prioritize densification in the city. The area is relatively poorly served by the amenities as it has a low accessibility to the downtown area, only a minor commercial node and isn't served very well by parks. Furthermore, whereas it is a single-family neighbourhood, the area is also somewhat higher density area than many other neighbourhoods in the city which often offer better amenities.