To do this, I'll generate a hex-grid to cover the entire city, where each cell is about one city block, find the center of each cell that intersects the current network, generate an isochrone for each location, then join all those isochrones into a single multi-polygon.

This is very similar to the method for generating the #WalkPotential metric I created, and I'll likely be copy/paste/modifying some of that code.

https://mark.stosberg.com/new-software-to-calculate-walk-potential-for-cities/
🧵

Here's #WalkPotential for #RichmondIN

Lighter color indicates more categories of amenities within a 10 minute walk.

#urbanplanning #openstreetmap #mapping

Since I had all my tooling and context at hand, I went ahead and calculated Walk Potential for a second city that expressed interest, #GoshenIN

Lighter color indicates more categories of amenities within a 10 minute walk.

For a city this size of about 40,000 people, each of the data processing steps takes just about a minute or less.

#walkpotential #urbanplanning
/home/mark/Documents/Mapping/Goshen/2024-05-25-goshen-in-hexgrid.geojson

Here's #WalkPotential for #Minneapolis. Lighter color indicates more categories of amenities within a 10-minute walk. Darker is more car-dependent... or a lake.

Behind this is a hex grid with detailed scores for over 10,000 block-level locations which can be used for precise planning analysis.

#openstreetmap #mapping #urbanplanning

That's a wrap for my live-blog thread! 🧵

The code I'm running today, #WalkPotential is #OpenSource. You can look at the "amenities.json" to see what amenity categories we use and how we query for them.

https://gitlab.com/markstos/walk-potential

You can search for any of those categories individually and confirm (and improve!) the accuracy of the data.

Walk Score is likely querying incomplete and some inaccurate OSM data as well, but there's no way to find and fix those errors.

Some other cities have used "Walk Score" for planning, but that's problematic both because it's expensive and also because it's a black box, which is bad for equity.

Do Walk Score factory in liquor stores but not wine? Community centers but not churches? Salons but not barber shops? Art museums but not brothels? It's a secret. Those choices contain some bias, but you can't see them and you can't adjust for them. 🧵

#urbanplanning #walkpotential

I'm back on track. I downloaded the official city boundary of #minneapolis from their open data portal and used my hex grid generation tool against that, and it worked. Now to plug this into a second run of #WalkPotential calculations.

#OpenStreetMap #mapping

Trivia: Calculating #WalkPotential for #MinneapolisMN includes calculating the 10-minute walk radius from 3,633 bus stops in the city.

My laptop is currently grinding through merging those 3,633 polygons into a single multi-polygon.

I will not be surprised if it crashes at this step, because it has before.

Only 300 polygons left to merge! 🧵