About this data
This table ranks census tracts in cities (pop. ≥ 50K) east of the
Mississippi as potential 15-minute neighborhoods. Each index is a
percentile rank (0–1): a value of 0.85 means the tract scores
higher than 85% of all tracts. Column headers are abbreviated; hover
for the full field name. Columns marked ⊕ feed into the combined score.
weather (city) is uniform across all tracts in a city; all other indices vary by tract.
- combined
- Weighted average of five ⊕ sub-indices (equal weight by default).
- ⊕ vibe
- Percentile rank of culture density + veg restaurant count + bike repair shop count. Each item strongly boosts the score.
- culture
- Sub-component of vibe. Density of cafes, pubs, bars, bookshops, bakeries, museums, universities, libraries, theatres, cinemas, and community centres. Source: OpenStreetMap via Geofabrik.
- vegetarian
- Number of vegetarian and vegan restaurants in the tract. Source: OpenStreetMap via Overpass API.
- bike_repair
- Number of bicycle repair shops in the tract (
shop=bicycle or amenity=bicycle_repair). Source: OpenStreetMap via Overpass API.
- ⊕ affordability
- Percentile rank of housing cost (lower cost = higher rank). Blends median gross rent and median home value. Source: ACS 5-Year Estimates (2022).
- ⊕ weather_city
- Percentile rank blending two city-level terms: comfortable days (75%), the rank of
nice_days − 2×hot_muggy − frigid_days, and sunshine (25%), the rank of clear_sky_%. A muggy day costs 2 nice days because it is a whole day spent indoors; that is a stated preference, not a measurement. Source: Daymet and gridMET, daily values 2016–2025.
- ⊕ safety
- Percentile rank of safety: 50% violent crime rate (city/agency level, state-median fallback) + 50% incarceration rate (tract-level). Sources: Kaplan UCR 2020 (Jacob Kaplan’s Concatenated UCR Files, ICPSR 100707); Opportunity Atlas (Opportunity Insights).
- ⊕ mobility
- Composite of walkability (30%), transit (30%), and bikeability (40%). Source: OpenStreetMap via Geofabrik.
- walkability
- Sub-component of mobility. Density of footways, pedestrian zones, steps, and crossings.
- bikeability
- Sub-component of mobility. Density of cycleways and bicycle parking.
- transit
- Sub-component of mobility. Density of bus stops, bus stations, and railway stations.
- pop_45min
- Working-age population reachable within a 45-minute transit commute (walk + GTFS schedules, time-decay weighted), averaged across block groups in the tract. Source: EPA Smart Location Database V3.
- 0_cars_%
- Percentage of zero-vehicle households, averaged across block groups. Higher = more car-free viable. Source: EPA Smart Location Database V3.
- rent / home_val
- Median gross rent and median home value. Source: ACS 5-Year Estimates (2022).
- nice_days
- Days per year with daily high between 60–85 °F.
- hot_muggy
- Days per year too hot to be outside, by peak heat index. Heat index already folds in humidity, so no separate dew-point test is needed. A day ramps in rather than switching on: 85 °F counts as nothing, 95 °F as a whole day. Onset sits in the NWS “caution” band (80–90 °F) and a full day inside “extreme caution” (90–103 °F).
- frigid_days
- Days per year too cold to be outside, by lowest wind chill — which folds in wind the way heat index folds in humidity. Ramped the same way: 20 °F counts as nothing, 0 °F as a whole day.
- clear_sky_%
- Share of clear-sky solar radiation the ground actually receives, averaged over the record: measured radiation over what a cloudless sky would deliver at this latitude and date. Dividing out the astronomy is what makes cities comparable --- raw sunshine hours would rank southern cities highest simply for being southern, counting warmth twice. This replaces an earlier proxy that counted non-overcast hours at airport ceilometers, which see only below 12,000 ft and so missed high cirrus and over-reported coastal stratus.
- jul_high / jan_low
- Average daily high in July and average daily low in January (°F), over the full record. Cities with a July high above 93.0 °F are excluded.
- weather record
- All weather columns average daily values over 2016–2025 at the city centroid. That is a decade rather than the conventional 30-year normal, on purpose: the climate is not stationary, and over 1995–2024 every one of fourteen sample cities grew muggier on these thresholds — by 6 days on average, 24 in Miami. A longer mean would dilute the trend with years nobody will live in again; the ranking is nearly unchanged either way (Spearman 0.98). Temperature and humidity come from Daymet at 1 km, fine enough to separate a city from the water beside it; wind and solar radiation come from gridMET, which is published on a 4 km grid but interpolates those two fields from NARR/NLDAS-2 at 32 km native, so treat them as regional rather than local. NLDAS-2 shortwave carries a known positive bias over North America, and measures about 5–12% above published NREL insolation here; that matters little because this column is percentile-ranked, but it does mean a mountain town and the city below it can read alike. Daymet also publishes radiation, but it is modelled from temperature range rather than observed and puts Miami’s annual sunlight level with Boston’s, so gridMET’s assimilated figure is used instead. Neither is a weather station, so no city borrows a distant airport’s climate. Daily values matter: a 30-year hourly normal averages each calendar hour and so smooths away every heat wave, which is why an earlier version of this table showed zero muggy days for Cambridge and Asheville alike. Sources: Daymet V4 R1, ORNL DAAC (Thornton et al. 2022); gridMET, Abatzoglou 2013 (CC0).
- crime/100K city
- Violent crimes per 100,000 residents (city-level).
- incarceration_rate
- Fraction of children from low-income families (25th percentile) who grew up in this tract and were incarcerated on April 1, 2010 (2010 Census), pooled across men and women (tract-level). Lower = safer. Source: Opportunity Atlas, Opportunity Insights (opportunityinsights.org).
- neighborhood
- Nearest neighborhood name; click 🗺 to open Google Maps centered on the tract. Source: Google Maps Geocoding API.
Excluded data
- Cities with pop. < 50,000 are not included.
- Cities with mean July high > 93.0 °F are excluded for excessive heat.
- Census sentinel values are treated as missing:
-666666666 (data not available),
top-coded values (3501 for rent, 2000001 for home value — exceeds reporting cap),
and bottom-coded values (99 for rent, 9999 for home value — below reporting floor).
Tracts missing rent or home value after this cleaning are dropped.
- Zero-amenity tracts — tracts with no OSM features in any culture, veg, walkability, bikeability, or transit category — are dropped.