As we all slog through another week of social distancing, increasingly my news and social media feeds are recommending that Cities engage in adaptive responses to the "new normal". Of particular interest to me is how can we adapt our cities for massively increased pedestrian space requirements and diminished vehicular travel. More simply, the COVID-19 situation is resulting in a lot more people staying home (or near home) who want to go outside but need to be distant from each other. At the same time, we have a STEEP decline in car travel due to everyone working from home and many of our daily needs being met remotely or via delivery services (or not at all).
The solution, of course, is to create more temporary open space from the increased supply of temporarily unused car space. However, we need to do this in a way that maximises benefits for those who need green space the most (i.e. this is a giant equity issue) and minimizes disruption for goods movement, commerce and emergency services (i.e. cities still need streets to do stuff). Given these requirements, it is actually fairly straight forward to set up a model that allows us to isolate roads for closure while minimizing impact to civic functions. Half of this model owes its genesis to a progression/adaptation of the excellent work that the City of Vancouver Parks Board completed for the their Van Play City Wide Parks Master Plan to build equity initatives zones in order to facilitate and prioritize capital planning decisions using an equity lens.
IDENTIFYING BLOCKS THAT ARE IN NEED OF ADDITIONAL OPEN SAPCE
In light of COVID-19 situation, though, I've modified the Parks Board model and I've landed on three factors that should allow for a proper identification of potential blocks where road closure would be a good idea:
Access to public greenspace is a great way to get out of the home and stay healthy, provided social distancing is maintained. In practice, this means highlighting locations in the city with lower areas of parks available within a ten-minute walk. Areas with park access deficits require more open space during COVID-19
We need to recognize that increased numbers of persons at home results in increased demands on existing greenspace and therefore we need to get a good sense of where there is a large degree of population density in the city. Areas with greater population density require more open space during COVID-19
People who have private greenspace (i.e. backyards, courtyards, rooftop gardens) are considerably more privileged than those who do not have these valuable spaces available to them to find some respite from the monotony of staying at home. Areas with reduced access to private greenspace require more open space during COVID-19
With the above three criteria combined into an overall equity score, we can identify key streets that may benefit from some type of treatment (lane closure, partial closure or full closure) to build up some additional human space out of the very functional, but currently functionally underutilized pavement that covers 50% of our city. I evaluated streets based on the average score of their facing blocks (i.e. street between the even addresses and the odd addresses). That in and of itself was a fun exercise and if anyone needs block data tied to streets, I’ve got it!
The image below shows the interconnected circuitry that I used to establish which two blocks face which streets.
I then picked the lowest scoring streets (the bottom 5%) and mapped them out, as you can see below. You can tell that I did something right, because all of these streets are well distributed around the City (following some obvious socio-demographic trends), instead of being clustered in one neighborhood.
ACCOUNTING FOR ACCESS, EMERGENCY VEHICLES AND GOODS MOVEMENT
Now here comes the fun part: making sure that each and every prioritized road closure does not impede emergency access, goods movement, residential access to the parcel and parking for businesses. To do this I set up a couple of criteria:
A street can only be closed off completely if:
it is a local road;
there are alleyways on both blocks facing the street;
all parcels on the street can be accessed through the alleyway; and
the street has both an entrance and egress
A street can be partially closed off if:
it is a local road but with only one alleyway;
it is classified as an arterial or collector (typically transit or truck routes or with much commerce) and has alleyways on both blocks facing the street; and
all parcels on the street can be accessed through some combination of street and alleyway
Regardless of closure type (full, partial, none), emergency level of service should not be demonstrably impacted.
every parcel should be accessible to its nearest firehall; and
response times should not decrease by more than 0.5% over baseline conditions
To classify road treatments, I essentially picked the top 2.5% and if they were a local street, with two lanes on either block and had an entrance and exit, I classified these all as having potential for road closures. To then test if closing the roads actually makes a difference for access or emergency services, I set up a closest facility problem and routed every parcel to its nearest firehall. In reality this should also involve ambulance dispatch as well… but, well, blog post round 1 here.
I won't bore you with a map showing blocks that have alleyways or road classifications, but the LOS (level-of-service) analysis for emergency response is fairly interesting. In the image below you can see how removing 173 local streets entirely from the street network does not affect individual parcel access or demonstrably increase distances driven (0.3% by my calculations). (Also note that we have a firehall gap at Main and 41st)
So, if we put it all together, we can get a pretty good sense of what we can close off, what we can partially close and what should remain as car-oriented streets… for now. The upshot is that this analysis can be refined, rebuilt and refactored, if we consider new equity considerations (i.e. race, income, gender, social justice), new access considerations (parking availability in the alleyway, lane width, pedestrian traffic volumes etc.) or integration with existing greenways and bikeways (hint: existing greenways and bikeways do not typically have very low equity scores).
It should be noted that this a first cut, and there is much to improve in this analysis. Regardless, it does seem that we can conceivably shut down about 17km of city streets to the great benefit of many Vancouverites. For reference, 34 hectares of open space is about as big as Evert Crowley Park (38ha) and is much bigger than Trout Lake (27ha). For some bang for the buck, every km of closed road area adds approximately 1.5ha to our supply of desperately needed open space. Finally, as requested, the top ten list of potential blocks to close is at the bottom of this blog post. As always, any questions, criticisms and kudos are all welcome in equal measure.