NYC Mobility

Start with geography. NYC is split into five boroughs, and each one has a different shape, density, and travel rhythm. That matters because mobility options are not evenly distributed: some areas have many subway lines close together, while others rely more on buses, cars, or longer walks between stations. Use the map below as your orientation layer before reading any trend chart.

The notes on the right summarize what the borough usually means for a visitor. Think of them as quick expectations: where rail is thick and fast, where surface routes carry more of the load, and where distances are simply larger. This helps you interpret later charts more correctly, because the same mode can behave very differently across boroughs.

With boroughs in mind, the next step is access. The map that follows layers stations and stops for each mode so you can see where coverage is dense and where it thins out. Dense clusters usually mean easy transfers and multiple route choices. Wide gaps usually mean longer walks, a bus connection, or a mode switch.

Use this map to answer a practical question: “If I step outside here, how many options do I have nearby?” If the answer is “many,” you can be flexible. If the answer is “few,” you should plan more carefully and expect a longer first/last mile.

Access is one thing, reach is another. The isochrone view estimates how far you can travel from Grand Central in typical travel-time steps. This helps you compare modes on a common scale: time. A mode might have many access points, but if it is slow in traffic, its reach can be smaller than expected. Another mode might have fewer access points, but move quickly once you are on it.

Walking and cycling follow street networks, while bus and subway are approximated using access points and average speeds. Treat the result as a planning baseline, not an exact prediction. The goal is to compare typical reach, not to guarantee a specific minute-by-minute route.

Combine the two maps. First check where the access points are clustered, then check how far the selected mode typically reaches in the same amount of time. This contrast is a strong hint about when to start on rail, when a bus transfer is worth it, and when a bike or a short ride-hail hop is faster. If you are unsure, choose the mode with both good access and good reach for your area.

This section shows how the system behaves at a city-wide level. The ridership dataset is annual, so each point represents a full year of travel. Use these charts to understand scale and direction: which modes carry the most trips, which ones are growing, and how the total system changes over time. This is the “big picture” that explains why some modes feel dominant and why others feel more niche.

Read the visuals in this order: start with the annual bars to see long-run trends, then check the latest-year pie to understand today’s mix, and finish with the total ridership bars to see how the whole system expands and contracts. Looking at both share and totals is important: a mode can keep a similar share while the absolute number of trips changes dramatically.

The annual lines show direction and scale. Subway remains the largest channel of demand, buses move within a tighter band, taxi/FHV can swing more sharply, and Citi Bike has one of the clearest growth trajectories. The 2020 disruption is visible across modes and helps explain later changes in how the city feels during commuting hours.

When a line rises steadily, it suggests that the mode is becoming more common or more available. When it is volatile, it often reflects sensitivity to external conditions (policy, demand shocks, pricing, or changing service supply). Keep these patterns in mind when you choose a mode for a specific kind of trip.

The latest-year mix is your baseline for how trips are distributed now. Subway usually leads, buses and taxi/FHV contribute large shares, and bikes remain a smaller but increasingly visible slice of total trips. Use this as a quick reality check: if a mode has a small share here, it can still be excellent for certain routes, but it is not the main backbone of the whole city.

Total ridership puts the shares in context. Even if the mix stays similar, the absolute number of trips can change a lot. The system peaks before 2020, drops sharply, and only partially rebounds. This helps explain why some routes can feel less crowded while still being essential, and why service patterns may not match older expectations.

The subway is the fastest way to move across the dense core and between boroughs. It is built for high-capacity travel along major corridors, and it usually offers the best speed-to-distance ratio for medium and long trips. If your route crosses Manhattan or connects to Brooklyn/Queens corridors, the subway is often the first option to check.

The annual line highlights long-run demand and the strong 2020 disruption. The share line shows how dominant the subway remains relative to other modes. Use these two views together: the top chart tells you how large the mode is in absolute trips, while the share chart tells you how central it is to the overall system.

For visitors, the practical takeaway is: use the subway when you need predictable cross-city movement, especially during busy hours. Combine it with short walking segments for the last mile, or with a bus when the destination sits far from the nearest station. When you see the subway share stay high, it is a signal that many typical NYC trips still depend on rail as the core connector.

Buses are the surface connectors. They are especially useful where subway coverage is thinner, for crosstown movement, and for short hops that would require multiple subway transfers. Think of buses as the “grid” that fills gaps: they reach many streets, but they are more exposed to traffic conditions.

The annual line tends to stay in a narrower band than the subway, reflecting the steady role of buses in local mobility. The share view shows how buses remain a substantial slice of travel even when other modes fluctuate. If your destination is not near a subway station, a bus segment can be the simplest bridge.

For visitors, buses are best when you want a direct street-level route, when you prefer fewer stairs, or when you are moving within a neighborhood. They are also useful as a “last-mile” solution: subway to a corridor, then bus to the final stop. When street traffic is heavy (see the traffic chapter), bus travel can slow down, so timing matters more than with rail.

Citi Bike represents the flexible, short-range layer of NYC mobility. It works best when stations are close together and streets are bike-friendly, which is often true in the densest parts of Manhattan and in many Brooklyn corridors. Bikes can be faster than cars for short trips, especially when traffic is slow and when you want to avoid transfers.

The map below shows the most common station-to-station flows. These lines highlight where bikes are used repeatedly, not just occasionally. Strong clusters usually indicate areas with both high demand and high station turnover, which often means better availability throughout the day.

The strongest flows concentrate around the core, especially in lower Manhattan and nearby waterfronts. These are typical “short corridor” trips: commuting links, last-mile connections, and short cross-neighborhood moves. If your itinerary stays within these areas, bikes can be a reliable option for quick point-to-point travel.

The usage and share lines put bikes in context: they are smaller than subway and bus in absolute volume, but their growth trajectory is strong. This usually reflects network expansion, more stations, and wider adoption for short trips. For visitors, it means bikes may be more available and more practical now than older guidebooks suggest.

Hourly demand separates commuter behavior from leisure behavior. Member rides typically spike during morning and evening windows, while casual rides rise later in the day and on weekends. This matters because availability can change quickly: at commute peaks, popular stations can empty out, and in the evening they can fill up again. If you want the easiest pickup, plan slightly outside the sharpest peaks.

When demand spikes, a handful of stations carry a large share of departures. The ranking below shows where most trips begin. These hotspots are useful for visitors: they often sit near major destinations, scenic corridors, or strong transit transfer points. They are also locations where bikes can run out faster, so checking nearby alternative stations can save time.

The next charts explain what a “typical” bike trip looks like. Trip length tells you whether bikes are mostly used for quick hops or longer rides. Rider mix explains how much demand comes from frequent users versus one-time visitors. Vehicle type helps you understand how much electric assistance is present in the fleet.

Duration bins usually show a strong short-trip bias, which matches the idea of bikes as a quick connector. The membership split tells you how much of the system is powered by routine travel versus occasional riding. The bike type share highlights the role of e-bikes: if electric share is high, it supports longer trips and makes bridges and hills more approachable.

Practical summary: bikes are excellent for short distances in dense areas, especially when you want direct travel with no transfers. They become less convenient where station density drops or where the route requires long gaps between access points. Use the access map and the flow hotspots together to decide where bike travel is most reliable.

Taxis and FHVs are the door-to-door layer. They are useful when you want a direct trip with minimal walking, when you are carrying luggage, when weather makes walking unpleasant, or when you are traveling late at night. Unlike fixed-route modes, their availability and travel time depend more on demand peaks and street traffic.

The monthly series compares services over time and shows how the market composition changes. The following charts then explain typical timing (when pickups peak), trip length (distance and duration), and geography (where pickups concentrate). Together, these help you understand when ride-hail is most common and where it is most active.

Monthly lines reveal seasonality and long-run shifts between services. You can often see a gradual move toward high-volume ride-hail, while traditional yellow taxi volumes change relative to that growth. Use this chart as context: it explains why you may see more app-based vehicles than classic street-hail taxis in certain periods and areas.

The heatmap shows the pickup pulse by hour and weekday. Weekdays often feature commute peaks, while weekend demand shifts later into the day. This is a practical hint for pricing and waiting: when the heatmap is hottest, demand is high and trips can be slower due to traffic. When it is cooler, pickups may be faster and streets may be clearer.

Top origin-destination pairs show repeated door-to-door routes. These often connect hubs to hubs: airports, major stations, dense centers, and key neighborhoods. Overall, taxi/FHV travel is most useful when you want directness, when you want fewer transfers, or when the time saved is worth the cost. If traffic is heavy, consider rail for cross-city movement and reserve car trips for shorter hops or late-night convenience.

After transit modes, this section focuses on the street layer. Automated traffic counters track volume on many road segments, allowing comparisons of when and where surface traffic peaks across boroughs. This matters even if you do not plan to drive: traffic affects buses, taxis, and travel-time expectations on the street.

Read the charts as “when streets are busiest.” The hourly view shows the daily pulse, the weekday/weekend split shows lifestyle differences, and the corridor ranking highlights the routes that consistently carry high volumes. If you are planning a car or bus trip, this section helps you choose a better time window.

Hourly curves typically show morning and evening rush periods with a midday plateau. When the curve rises, buses and taxis tend to slow down, and travel times become less predictable. When it falls, streets are easier and surface modes become more efficient. Use this as a quick guide for scheduling: small changes in departure time can produce noticeably different travel experiences.

Practical takeaway: if you see strong peaks, expect slower surface travel at those times and consider rail or bikes when possible. If your trip must be on the street, aim for off-peak windows and avoid corridors that consistently rank at the top. This is especially useful for airport connections, cross-borough car trips, and long bus routes.

This page connects three practical dimensions: coverage (where options exist), scale (which modes carry the city), and timing (when each mode is strongest or most constrained). Subway dominates long cross-city movement, buses fill neighborhood gaps, bikes serve short corridors, taxi/FHV offers flexible door-to-door travel, and street traffic explains when surface travel slows down.

If you want a simple way to use the information, follow this routine: first identify your borough and the nearest access points; then choose a mode that matches your distance and expected time window; finally, use the mode-specific charts to anticipate peaks, hotspots, and typical trip patterns. This approach makes trips more predictable and reduces the chance of choosing a mode that looks convenient on a map but performs poorly at that hour.

In short: rail for speed across corridors, bus for surface coverage and local links, bike for quick direct hops in dense areas, and taxi/FHV for directness when you value convenience. When in doubt, compare reach and access density, and choose the option that keeps your route simple.

This story uses three primary datasets: annual ridership by mode, collision records by borough, and automated traffic volume counts. The maps use public GeoJSON layers for borough boundaries and service access points.

Core datasets

• Annual ridership by mode (Subway, Bus, Citi Bike, Taxi/FHV)
• Collision records by borough with injured/killed by user type
• Automated traffic volume counts by borough, hour, and corridor
• Citi Bike tripdata (station flows, timing, and rider mix)
• GeoJSON layers for boroughs and service points

Reference projects and source links (last accessed)

• https://catalog.data.gov/dataset/automated-traffic-volume-counts — Last accessed 2025-01-15 13:20
• https://www.kaggle.com/datasets/microize/nyc-taxi-dataset?resource=download — Last accessed 2025-01-14 16:40
• https://uclab.fh-potsdam.de/cf/ — Last accessed 2024-11-08 19:12
• https://senseable.mit.edu/bike-trafficking/ — Last accessed 2024-11-16 20:43
• https://cityflow-project.github.io/#cta — Last accessed 2024-11-23 18:27
• https://a816-dohbesp.nyc.gov/IndicatorPublic/data-explorer/walking-driving-and-cycling/?id=2415#display=summary — Last accessed 2024-12-02 21:05
• https://ir-datasets.com/nyt.html — Last accessed 2024-12-07 16:49
• https://catalog.data.gov/dataset/mta-subway-customer-journey-focused-metrics-beginning-2015 (2015-2019) — Last accessed 2024-12-11 20:18
• https://catalog.data.gov/dataset/mta-subway-customer-journey-focused-metrics-beginning-2020 (2020-2024) — Last accessed 2024-12-14 17:31
• https://data.ny.gov/Transportation/MTA-Daily-Ridership-Data-2020-2025/vxuj-8kew/about_data — Last accessed 2024-12-18 19:02
• https://data.ny.gov/Transportation/MTA-Customer-Feedback-Data-2014-2019/tppa-s6t6/about_data — Last accessed 2024-12-22 14:40
• https://d3blocks.github.io/d3blocks/pages/html/index.html — Last accessed 2024-12-28 09:55
• https://nyc-transit-maps.preview.emergentagent.com — Last accessed 2025-01-03 18:10
• https://research.google/blog/introducing-mobility-ai-advancing-urban-transportation/ — Last accessed 2025-01-06 21:14
• https://c2smart.engineering.nyu.edu/c2smart-data-dashboard/ — Last accessed 2025-01-09 17:38
• https://public.tableau.com/app/search/vizzes/nyc%20transportation — Last accessed 2025-01-11 20:51
• https://willgeary.github.io/portfolio/projects/transitflow/ — Last accessed 2025-01-12 19:27

SUPSI 2025/2026 - Bachelor in Data Science and Artificial Intelligence
Data Visualization Course M-D32023