Mapping Density around St Joseph’s Seminary

3 min readMar 9, 2016

I’m going to be revising based on methodology conversations and clarifications. The numbers below are likely to change. Thanks for your patience while I work through this draft.

I decided to take a closer map analysis look at neighborhood density around St Joseph’s Seminary. I’m sharing this to get feedback and ideas on this work.

EYA, who have proposed the 12th and Allison development on the St. Joseph’s Seminary property, recently held a community focus group to address concerns raised by the neighborhood, and share an updated plan.

I have a lot of thoughts on this, but wanted to start with a study EYA conducted on neighborhood block density. Discrepancy in density between the existing neighborhood and the proposed development feels like the key neighborhood concern, and is the motivation for their potential application for rezoning from R-2 to R-5-A.

According to my analysis, EYA might have over-counted the density in surrounding blocks by between 5% — 44%, and under-counted the density in the area to be developed by 61%. No matter what you’re feelings on the development, we need to base discussion on accurate numbers and clear methods. I’ll explain and illustrate my methodology below. I’d welcome the chance to sit and work with EYA to understand how their methods differ, and figure out where the error has been made, whether in my calculation or theirs.

EYA Neighborhood Block Density Estimate. Interactive Version

My first concern on reviewing this study was that the density of the Josephites’ property was calculated based on the entire property area, including proposed “communal open space”. Realistically, the only area of the property that could be developed under any scheme (whether R-2 density, or R-5-A density, or not at all) would be north of the seminary building, while the remaining space to the west, east, and south of the seminary would be retained as open. Secondly, truly public space in general is not counted towards zoning density of residential blocks. For instance, properties abutting the nearby Turkey Thicket public space do not count those fields as part of their density.

Replication of EYA Neighborhood Density. Interactive Version

So, I redrew the boundary of the 12thandAllison development to lie only north of the seminary building. With the high estimate of 100 new townhouses, and 6 existing structures, over ~5.5 acres, the density is 19.0 units/acre vs 11.8 units / acre in EYA’s map.

I also looked at the density in the surrounding neighborhood blocks delimited by EYA. Surprisingly, the density found is consistently lower than EYA’s map. I can not account for this difference, and would love to resolve this difference with EYA.

One suggestion for the discrepancy has been is that the density was calculated without streets inside the “neighborhood blocks”. There’s actually no concept of “neighborhood blocks” in DC zoning — this concept seems to have been adopted in this study in order to have comparable size residential areas to the proposed development. In zoning in DC, zones are broken down by “squares”, as seen in the above screen of DC’s Zoning Map. Squares include properties and alleys, but not streets; this could explain the difference. So a more accurate measurement would be to measure unit density of each square, along with potential Squares inside the proposed EYA development.

Technical notes

Used to draw the neighborhood block boundaries, and count the number of units per block. Wrote a short turfjs script to calculate and insert area and density for each block. Uploaded to Mapbox Studio for processing and styling.

For the EYA map, took a screenshot and rectified using mapwarper.




Mapper. Coder. Earth Genome. OpenStreetMap Foundation. HOT. former Mapbox / Presidential Innovation Fellow. Views are my own.