Tag: OpenStreetMap

How to map where I traveled when I went to Gorinchem, NL

On Monday, December 4, 2023, I wanted to ride a line in the Netherlands that I hadn’t yet, which is called the “MerwedeLingelijn” and goes between Dordrecht and Geldermalsen. In the NS journey planner database it’s called “Stoptrein” which distinguishes it from “Sprinter” and “Intercity”. Those names distinguish the service types on the Dutch railway network. (This particular Stoptrein is also a diesel-electric trainset.)

From Rotterdam, where I was staying, it would require at least one transfer to get to Gorinchem. But I wanted to stop in Utrecht to say hi to a friend during his work break – this meant there would be two transfers.

Here’s the itinerary I traveled on Monday

  • Rotterdam Centraal to Utrecht Centraal via Gouda, Intercity (half-hourly service) – 55 km
  • Utrecht Centraal to Geldermalsen, Sprinter (10-20 minute service) – 26 km
  • Geldermalsen to Gorinchem, Stoptrein (half-hourly service) – 27 km
  • [lunch and walk in Gorinchem, distance not recorded]
  • Gorinchem to Dordrecht, Stoptrein (quarter-hourly service) – 24 km
  • Dordrecht to Rotterdam, Waterbus (hourly service) – 21 km

How I drew the map

I wasn’t about to draw the routes by hand (although I did record the Waterbus ride on Strava as a “sail”) so I grabbed the data from OpenStreetMap.

If you want data in bulk from OpenStreetMap a common way to get it is from the HotOSM export tool. But I wanted specific transit routes, for which I could find the “way” IDs and export only those. For that I used Overpass Turbo and wrote the following query:

[out:json][timeout:25];
// gather results
rel(id:324888,13060594,5301520,2785504);
way(r);
// print results
out geom;

Notes

Frequencies refer to the pattern in the hour I used the service. The itinerary doesn’t include a Rotterdam Metro ride or the roundtrip bike ride from the Schiebroek neighborhood to Rotterdam Centraal).

Starting on December 10, the NS (Dutch national railway operator) is adding over 1,800 train services each week.

Traffic calming around Chicago schools and parks

Reducing the number and speed of automobiles near schools and parks is a proven way to reduce the number of traffic crashes involving children, part of a practice known as “traffic calming”. In Chicago a key way to do that has been to “cul-de-sac” (which I’m using as a verb) a street to prevent through traffic (reducing the number) and preventing speeding (reducing the speed).

My study of this was inspired by the above tweet, where the person is accurate when they say, “it works with any street”. Indeed: Full access to the school or park and to every property on the block is maintained, but drivers are not able to go through while pedestrians and bicyclists can (another practice called “modal filtering“).

My favorite example in Chicago is Hadiya Pendleton Park, at 4345 S Calumet Ave. This project created two mid-block cul-de-sacs and a park in the middle of a block, using vacant city-owned land on both sides. Creating new open space is a common corresponding outcome of the cul-de-sac application, which is what occurred at Funston Elementary School (see the before and after aerial photos below).

two black and white aerial photos taken of Funston Elementary School, one in 1998 showing a 4-way intersection of McLean Ave and Central Park Ave and one in 2003 showing a cul-de-sac on McLean Ave west of the school and landscaped area between the cul-de-sac and the three-way intersection of McLean and Central Park.

Through Twitter I solicited additional examples of where the city has created traffic calming near schools and parks using cul-de-sacs. Examples were submitted by RolandEmily (who mentioned Funston), Matt, Steven, and another tweeter.

Using OpenStreetMap, Overpass Query Language, and Overpass Turbo, we can find all of the schools and parks that are within a specified distance of a cul-de-sac. It turns out there are 153 schools and parks in Chicago that are within 150 feet of a cul-de-sac. (This considers only schools, parks, and cul-de-sacs, tagged as “turning circles”, currently mapped in OSM, and I have not verified each of the 153 instances.)

Map showing parks, schools, and adjacent cul-de-sacs, the results of the Overpass Turbo query.
Map showing parks, schools, and adjacent cul-de-sacs, the results of the Overpass Turbo query. An inset map shows a zoomed in portion of the map to illustrate the different types of features captured in the Overpass Turbo query (specifically it shows Graham Elementary School and McInerney Park).

The query below will find all of the cul-de-sacs (mapped as “turning circles” in OSM parlance) that are in Chicago, all of the schools and parks in Chicago, and then all of the two categories of features that are within 45 meters of each other. (Run the query and show the map, which will always grab the latest data.)

/*
example from OSM wiki: https://wiki.openstreetmap.org/wiki/Overpass_API/Overpass_API_by_Example
*/
[out:json][timeout:25];
// fetch area “Chicago” to search in
{{geocodeArea:Chicago}}->.searchArea;

// get cul-de-sacs
(
  node["highway"="turning_circle"](area.searchArea);
)->.turning_circles;

// get parks and schools
(
  way["amenity"="school"](area.searchArea);
  way["leisure"="park"](area.searchArea);
)->.schoolsParks;

// find parks and schools near cul-de-sacs
(
  way.schoolsParks(around.turning_circles:45);
)->.matchingSchoolsParks;

// find cul-de-sacs near parks and schools
(
  node.turning_circles(around.schoolsParks:45);
)->.matchingTurningCircles;

// output results to the map
(.matchingSchoolsParks; .matchingTurningCircles;);
out geom;

Read the three other blog posts I’ve written about using Overpass Turbo to quickly sift through and extract desired mapping data from OpenStreetMap.

Chicago’s massive parking footprint – as measured on December 30, 2022

It’s been three years since I last measured how much of Chicago’s land area is occupied by parking lots and parking garages. On December 25, 2019, using data drawn into OpenStreetMap by volunteers including myself, 2.5 percent of Chicago was for car parking.

Based on additional data since then, the land area of Chicago occupied by already-mapped parking lots and garages is 176,973,866.57 square feet, or about 2.7 percent of Chicago’s area.

This means that 0.52 additional square miles have been drawn into OpenStreetMap. If it hasn’t been drawn there, we can’t measure it. This means this number is a *minimum* of the land area devoted to car parking in Chicago.

Between 2019 and 2022, 173 more of these parking lots were drawn into OpenStreetMap in Chicago. There are still so many more parking areas that are not mapped!

That converts to:

  • 4,063.32 acres
  • 7.08 mi^2 (square miles)
  • 15.93km^2 (square kilometers)
  • 2.7% area of Chicago is parking (Chicago’s land area is ~589.56 km^2 )

There are some future parking -> building conversions coming soon. The buildings will be providing parking, but it will be integrated into a mixed-use development. The parking lot in the image, for example, is slated to become an office tower.

What is this place? Des Plaines “park” edition

Screenshot shows the “park” in Des Plaines, Illinois.

I was methodically reviewing features on OpenStreetMap that could benefit from additional attributes, namely missing names and cities. The information on OpenStreetMap feeds into Chicago Cityscape so that the real estate information service I created can show people looking up addresses the locations of nearby amenities, including parks.

This particularly large park in Des Plaines had no name and no city, so I started investigating. From the overhead imagery view, it looked to me like a landfill.

Since there was a forest preserve to the east, I looked in the Forest Preserve District of Cook County’s nice interactive map for any properties near Beck Lake. Nothing found.

Overhead imagery of the so-called “park”. Source: Mapbox

I looked on Google Maps Street View for some insight and there was no name. (Note that data from Google Maps cannot be used to amend OpenStreetMap because of Google Maps having a data license that isn’t compatible with OSM’s openness requirements).

Next I looked at the Cook County parcel map that Chicago Cityscape has to identify its “Property Index Number” (PIN). All of it fits within 04-31-300-003-0000 – the database pointed to “Catholic Cemeteries” as the property owner.

I looked up that PIN on the Cook County Recorder of Deeds to try and find more information. Lo and behold, the most recent document that was recorded against the PIN was for a “Memorandum of Option Agreement” that identified the Catholic Bishop of Chicago as the landlord and Patriot Acres, LLC, as the tenant.

Patriot Acres has a website that describes a new commercial composting facility being built at this location.

I re-tagged the feature on OSM with “amenity=waste_disposal” and “waste=organic” to properly describe it as a composting facility, and not a park. Case closed.

Chicagoland’s massive parking footprint – as measured on September 16, 2018

Using the footprints of parking lots and garages drawn into OpenStreetMap as a data source, the area of land in Chicagoland occupied by parking lots and garages is 247,539,968 square feet. (The data was exported using HOT Export Tool; you can replicate my export.)

That converts to:

  • 5,682.71 acres
  • 8.88 mi^2 (square miles)
  • 22.99 km^2 (square kilometers)
  • ≈ 0.26 × area of Manhattan (≈ 87 km^2 )
  • 3.9% area of Chicago is parking (Chicago is ~589.56 km^2 )

(I forgot to measure the portion of this within Chicago, and now the data snapshot is gone. I fixed this in the 2019 report.)