Project 2 — Transit Desert Analysis¶

Goal¶

"Which neighborhoods in [your city] are 'transit deserts' — places with high transit need but low transit supply?"

A transit desert is a place where the demand for transit (low-income, no-car, transit-dependent populations) exceeds the supply (nearby stops, frequent service).

This project tells a powerful equity story.

Data needed¶

Tools used¶

• ArcGIS Pro: Buffer, Spatial Join, Calculate Field
• Possibly Network Analyst for walking time
• SQL for filtering tracts and queries
• Layout, multi-panel map

Step-by-step workflow¶

Step 1 — Define your indices¶

You'll create two scores per tract:

Transit Desert = high TN, low TS.

Step 2 — Set up¶

1. Project: transit_desert_<city>.
2. Project all data to a local State Plane CRS.
3. Clip the city boundary out of the state.

Step 3 — Compute Transit Supply¶

1. Spatial Join: target = census tracts, join = transit stops, match = INTERSECT, ONE_TO_ONE.
2. Field map: keep Join_Count (stops per tract).
3. Calculate: STOPS_PER_KM2 = Join_Count / AREA_KM2.

For a more sophisticated TS, run a 0.25-mile buffer around stops, then use Tabulate Intersection to compute the % of each tract within walking distance of a stop.

Step 4 — Compute Transit Need¶

In the tract layer's attribute table:

1. Add field NEED_SCORE (Double).
2. Calculate as a normalized sum:

def need(pov, no_veh, elderly, kids):
    # standardize to z-scores or simple weighted sum
    return (
        (pov or 0) * 0.4 +
        (no_veh or 0) * 0.3 +
        (elderly or 0) * 0.15 +
        (kids or 0) * 0.15
    )

Expression: need(!PCT_POVERTY!, !PCT_NO_VEHICLE!, !PCT_ELDERLY!, !PCT_UNDER_18!).

Step 5 — Classify each tract¶

For each of TS and TN, classify into 5 quantile bins (1 = lowest, 5 = highest).

Then categorize:

def category(tn_bin, ts_bin):
    if tn_bin >= 4 and ts_bin <= 2: return "Transit Desert"
    if tn_bin >= 4 and ts_bin >= 4: return "Well-served high-need"
    if tn_bin <= 2 and ts_bin <= 2: return "Low-need low-supply"
    if tn_bin <= 2 and ts_bin >= 4: return "Over-served low-need"
    return "Mixed"

Step 6 — Symbology¶

Categorical symbology by CATEGORY. Use a bivariate choropleth if your version supports it; otherwise unique values.

• Transit Desert → bright red
• Well-served high-need → green
• Over-served → light gray
• Low-need low-supply → light blue
• Mixed → mid gray

Step 7 — Layout¶

A 2-panel layout works best:

• Left panel: Transit Desert map (the headline).
• Right panel: a small bivariate scatter chart (TN vs TS) and a tracker showing % of city population in transit deserts.

Or use a multi-panel Story Map in ArcGIS Online to walk through the methodology, then the result.

Step 8 — Write the story¶

Your README / Story Map should include:

1. The question.
2. The methodology (TN, TS, classification rules, sources).
3. The map.
4. Top 5 tracts identified as transit deserts.
5. A 1-paragraph "so what" — policy implication.

Skills learned¶

• Buffers + Spatial Join
• Multi-criteria scoring (a mini suitability model)
• Quantile classification
• Categorical symbology (potentially bivariate)
• Equity / accessibility narrative

Portfolio value¶

This is the project that gets attention for transportation, planning, and policy roles. The methodology is publishable and the equity angle resonates.

Stretch goals¶

• Add walking-time service areas from transit stops (Network Analyst).
• Add frequency weighting (high-frequency stops count more).
• Compare your map to the city's existing planning documents.
• Publish as a public ArcGIS Online dashboard.

→ Next: Food Desert Analysis.