An exploration of access issues to Fort Washington Park
My focus for this project was to geographically show the challenges of getting into New York City's Fort Washington Park by bike or on foot. Starting with Chris Whong's Mapbox Storytelling template, I adapted it to function as a "sticky" graphic, imported layers directly from GeoJSON to get past MapBox's 15 layer limit, added and rendered some of my own JavaScript properties, etc.
See also my Columbia Lede class presentation river side, not Riverside for a deeper dive into my process specific to rendering the unofficial trail.
Here's a prototype draft of my story, New York's Isolated Park on the Hudson. At this time (January 2024) I'm in conversation with a New York outlet about developing this story for publication. See also a similar map-based story I published with Streetsblog Chicago; because CMS restrictions confined the map to an iFrame, I also have a self-published version.
Data Source | Description |
---|---|
AllTrails | This free app allowed me to create path maps by recording them, as I did with the unofficial riverfront trail north of the George Washington Bridge, or drawing them, as I did with the Hudson River Greenway south of Fort Washington Park. I then exported these maps as KML files. |
NYC Planimetrics OpenData | Shapes for parks and Amtrak line, which I processed in QGIS and uploaded as layers in my MapBox style. |
Fort Washington GeoJSON | Points and paths supplementing MapBox layers. These include entry points to the park and landmarks such as the Little Red Lighthouse as highlighted in my story. This also includes a path for the official trail following the Henry Hudson Parkway north of the George Washington Bridge. |
To create most paths, I did the following:
- In AllTrails, used the Draw Map feature to create the path, then exported as KML.
- In QGIS, imported KML files and exported as shapefiles.
- In MapBox, imported the shapeFile into a style specific to this project.
- Index.html references my MapBox style.
For the unofficial trail north of the George Washington Bridge, I brought this into GeoJSON.io. Using satellite streets view, I deleted about 2/3rds of the points to reduce file size, and edited the path to correct imprecisions such as weaving over the Amtrak fence or into the river. Then I exported as GeoJSON.
My phone automatically geocoded the images, though I manually pasted these lat and lng coordinates into chapters in my MapBox.
This was my second MapBox storytelling project. It's very quick and easy to get started with the basic template, but I learned a lot through tailoring the template for my work:
- Became familiar with virtually the entire template, deleting sections (e.g. inline map) which I didn't use or need
- Uploaded map layers using map.addSource and map.AddLayer (+ filter) to work past MapBox's limit of 15 layers
- Picked up a lot of techniques which I ultimately stripped out to simplify my story. E.g. georeferencing to overlay an historic map of Fort Washington. I also initially used masking to hide New Jersey which made it look like the GW Bridge dropped off a cliff. I later settled on a more basic base map
- Experimented and revised my workflow for map design using a range of apps (QGIS, MapBox, DataWrapper, GeoJSON.io, AllTrails, Strava, Google Maps, Google MyMap)
- Most importantly, I learned to not be afraid to ask for help. And I got help from so many folks at Columbia- Aaron Reiss, Kelly Waldron, Steph Andrews amongst others. Federico Teijeiro originally found and recommended the MapBox storytelling template to me.
Fortunately I had the time to revise and improve this project after completing my studies at Columbia. Some possible next steps:
- Work with outlets to develop this story or variants based on this framework. Related to this, I'll need to find ways to work around outlets' proprietary content management systems.
- Create my own set of recommendations to help small, CMS-restricted outlets scope out exactly what it would take (e.g. AWS setup) to host this sort of project.
- Develop a scrollytelling story using MapBox's 3-D utilities, e.g. to show Von Humboldt's climb up Chimborazo.
Following is an overview of files in this repository:
- data- not much here beyond my GeoJSON file including some points and path data.
- Jupyter Notebooks- some early experiments in trying to programmatically reduce file size by deleting points. This was not used in the final project.
- images- photos used for this story
- config.js- Most of the text content for the story.
- index.html- Most of the code supporting the story flow.