Updates

[aseemdeodhar]

Spatially joined the sidewalks data with the vectorized LandScan data. The histograms show a high right skew for both data. Next steps are developing a relationship map with these two variables: mean sidewalk width, and daytime population density.

[aseemdeodhar]

By overlaying the previous two maps with a suitable color-scheme, I created this relationship map to understand how the sidewalk width in an area relates to the daytime population. As we are comparing these two features, we can ascertain that there may be four extreme scenarios:

• High Density & High Sidewalk width
• Low Density & High Sidewalk width
• High Density & Low Sidewalk width
• Low Density & Low Sidewalk width

[vr00n]

@aseemdeodhar this is fantastic. Can you share a shapefile of the relationship map?

[aseemdeodhar]

Crosswalk Files

I now have stress index data for day (boston_day_stress.zip) and night(boston_night_stress.zip) as sourced from the LandScan raster population database. By creating a crosswalk file, I was able to create insights for several administrative land partitions:

• ZIP Codes
• PWD districts
• BPD (Boston Police Department) districts
• BFD (Boston Fire Department) districts
• BTD (Boston Transportation Department) districts
• BPDA (Boston Planning and Development Agency) districts
• Election ward-precincts
• Boston City Council Districts
• Official Boston Neighborhoods
• Census Tracts & Block Groups

Data Source Boston Area Research Initiative, 2018, "Administrative Geographies for the City of Boston", https://doi.org/10.7910/DVN/JZV6ON, Harvard Dataverse, V1

Spatially Constrained Multivariate Clustering

The next step is to generate spatially constrained multivariate clusters for both day and night indices. Taking into account our indices, and other relevant variables, which areas of the city are most distinct in their clustering? In addition to the stress index (developed based on day/night population + sidewalk width), what variables would be suitable in this multivariate clustering analysis?

Day Stress Index:

• Summer land surface temperatures. I am contacting my colleague at BARI, who have been using summer land surface temperature in their parcel explorer Twitter bot.
• Tree canopy size/percent of land covered by tree canopy: sourced from the USDA Forest Service page

Night Stress Index

• Streetlight locations: sourced from the Analyze Boston database

[vr00n]

@aseemdeodhar Could you please update the README with the definition of the Stress Index?