Baobab — Christmas in New York

A new way to represent city data. Instead of a chart or a flat map, the data grows a living tree over a 3D model of the city: each group in the data is a branch with its own leaves, its roots run along the real streets to the neighborhoods where its people live, and moving lights carry them through their day, by subway, bus, car, bike and on foot, at the city's own hours. You read the numbers by looking, the way you read a place.

The first city Baobab is built for, and a foundation rather than a finished picture: New York's own open data, organized so more of it can be added, combined and shared. It is meant to be shown (on a phone, a classroom screen or a lobby TV), shared (one link, everyone sees the same moment), and engaged with (fly anywhere, choose anything, compare layers), so the patterns in the data come to light.

Not only. Baobab began with the languages New Yorkers with limited English speak at home, and that dataset was essential in building it; today it is one of three datasets Baobab NYC grows from, beside origin and country of birth. They are examples of what it can do: any data that counts groups of people in the city's boroughs or districts can grow the tree, yours included.

Three ways to grow the branches. Origin (heritage), which it opens on, counts New Yorkers by where their families come from, whether born abroad or here (US Census ACS tables B03001, B02015, B04006 and B03002, each person counted once as far as the tables allow). Languages spoken at home counts New Yorkers with limited English by the language they speak at home (NYC Department of City Planning). Country of birth counts New Yorkers born abroad by the country they were born in (ACS table B05006). Root size shows everyone living in each district (2020 Census); the districts' floor can show limited English or poverty (NYC Health). Around them: every subway train live from the MTA, in its tunnels and on its els, stations and platforms; every building in the five boroughs; streets, sidewalks, parks, playgrounds, piers and bridges as you come close; the city's weather and its yellow cabs in two editions.

Yes, every group in the data, down to the smallest: 135 origins, 111 languages, 146 countries. A leaf and a light each stand for a set number of people (the tree says how many); a group smaller than that still has a leaf of its own, a light on the streets and its roots to the districts where its people live. The legend lists them all, largest first, in a list you can scroll.

A dashboard answers the question it was built for. Baobab lets several datasets meet in one place, the way they meet in a neighborhood: who lives where, how they get around, when they travel, what is around them. Up to three layers show at once, one per visual channel (branches, root size, map colour), and it always says what it is showing: the unit, the year of each dataset, whether a number is an estimate, and that showing things together does not show cause.

Yes. Choose “Try with your data” on Baobab's page, or open Data layers in Baobab and paste the address of a CSV or JSON file: rows by borough or community district, with a group and a number (borough, group, people, for example, or one column per group). Each group in your data grows its own branch, with its own name and colour, and How to use then shows you where to see each of them. Trying it is free and needs no account; Interartive reads the file once and keeps nothing. To save your version and come back to it, sign in with a Motion Data subscription.

Every number comes from a public source named in the panel and the ticker along the bottom. Survey data (the ACS) are estimates with margins of error, larger for small groups and small areas; Baobab rounds small counts and hides the smallest, following each dataset's own privacy rules. One light stands for a group of people (its tooltip says how many); their trips are modeled from when New Yorkers really leave for work and how they travel, not tracked.

Yes. With a subscription, a live screen link puts Baobab full screen on any number of screens at once. You drive it and everyone watching sees what you do, as you do it. Anyone watching can choose Explore on my own to use it themselves (no subscription needed) and Rejoin live to watch again.

How to use, in the menu, walks you through it, and it speaks of the data you have chosen. Drag to turn, scroll or pinch to come close, and click or tap anything (a person, a train, the tree, a building) to choose it; press F, or Fly there, and the camera flies to it and follows it if it moves. Down at street level, a drag turns your view: all the way round, and up into the leaves. W A S D fly freely; H comes home.

Baobab is built so a city is data entry: each dataset is described by a manifest (what it is, where, when, who published it, how to show it, its privacy rules). New York is first; a city with the same kinds of open data can follow.