The 28 Panels — What Each Shows, and How to Read It
sitedia draws twenty-two panels (twenty-one in Houston, which has no zoning) for a site in New York City, Los Angeles, San Francisco, Boston, Washington, DC, Chicago, Miami, Dallas, Houston, Denver, Seattle or Minneapolis. They are ordered so that each one explains the next: the ground, then movement, then fabric, then what the fabric adds up to. This is what each panel contains, where the data comes from, and — the part that matters when you put it in a report — where it is thin.
1 · Aerial
USDA NAIP imagery, public domain, clipped to your frame. Its job is not decoration: it is the control. Put it beside the drawn panels and check that they agree with the world. If a building is drawn where the photo shows a parking lot, you have found a vintage mismatch before your reviewer did.
2 · Buildings
Every footprint in the radius, solid, no categories. Source is Overture Maps. This is the figure-ground, and read alone it tells you the density and grain of the fabric faster than any statistic. Look at the white, not the black — the shape of the void is the shape of the public realm.
3 · Building use
Footprints colored by land use, from the city's tax lots — MapPLUTO in New York, the County Assessor's parcels in Los Angeles, the Assessor-Recorder's roll in San Francisco, MassGIS assessor parcels in Boston, Office of Tax and Revenue parcels in Washington, DC, Cook County Assessor parcels in Chicago, Miami-Dade Property Appraiser parcels in Miami, Dallas Central Appraisal District parcels in Dallas, City and County of Denver assessor parcels in Denver, King County Assessor parcels in Seattle, City of Minneapolis assessor parcels in Minneapolis — joined to each footprint by lot (in Houston, the appraisal district's 2022 codes as the City published them). Residential, retail, office, commercial, lodging, cultural, education/welfare, industrial, other.
Known gap: the join is lot-based, and a lot with several buildings matches one of them. Footprints with no matching lot — including everything on the New Jersey side of the frame — stay Unknown. In practice about 70–90% of footprints get a use inside NYC, less near the edges. In Downtown Los Angeles 98.6% get a use, around Union Square in San Francisco 99.8%, in Back Bay, Boston 99.9%, in Downtown Washington, DC 98.0%, in the Loop, Chicago 94.2%, in Brickell, Miami 93.7%, in Downtown Dallas 100.0% (see Data sources).
4 · Building age
Same footprints, colored by decade of construction (year built from MapPLUTO in New York, from the County Assessor in Los Angeles, from the Assessor-Recorder in San Francisco, from MassGIS assessor parcels in Boston, from the Office of Tax and Revenue in Washington, DC, from the Cook County Assessor in Chicago, from the Miami-Dade Property Appraiser in Miami, from the Dallas Central Appraisal District in Dallas, from the Harris Central Appraisal District in Houston, from the City and County of Denver's assessor in Denver, from the King County Assessor in Seattle, from the City of Minneapolis assessor in Minneapolis — the original year built, not an effective year). This is the most underused panel in the set. Age reads development history directly: a block of 1920s fabric next to a 1960s superblock next to 2010s towers is three different planning regimes, visible at a glance.
Known gap: year built is present for roughly two-thirds of footprints citywide. The rest are drawn in light gray as Unknown — do not read them as “old”.
5 · Building height
Height in five bands on one sheet: under 40 ft, 40–80, 80–150, 150–260 and over 260 ft. High-rise and Low-rise split the fabric at a single threshold; this panel shows what lies between — whether the neighborhood is a field of mid-rise buildings or low fabric with towers standing apart. Height comes from the city or county, or from USGS LiDAR measured by sitedia, and otherwise from Overture (see Building heights); footprints without one are drawn in gray as Unknown.
6 · High-rise
Buildings at or above a height threshold you set, highlighted against the rest. Heights are the same as in Building height (city or county data, USGS LiDAR measured by sitedia, otherwise Overture).
Set the threshold deliberately. In Midtown, 150 ft is noise; in a low-rise Brooklyn neighborhood, 75 ft is the story. The default is a starting point, not an answer.
7 · Low-rise
The inverse read — everything at or below a threshold. Worth drawing as its own panel because low-rise fabric is where change pressure shows up. A field of two-story buildings inside a half-mile transit shed is a development story whether you like it or not.
8 · Zoning
Each city's zoning districts, colored by eight classes so that one legend reads in every city: residential, mixed use, commercial, industrial, public and institutional, open space, planned development (PD, PUD) and other. The city's own code — C-MX-8 in Denver, T6-48A-O in Miami, C6-4 in New York — is in the panel details; the class is ours, the code is the official designation. New York is drawn lot by lot from MapPLUTO's zoning district; the other cities from their zoning district maps (see Data sources). Houston has no zoning ordinance, so the panel is not drawn there.
Read it against Building use: where the zoning allows more than what stands, change is possible; where downtown is all planned developments (Chicago's Loop), the rules are in each development's own ordinance, not in the base district. It is not an official zoning determination.
9 · Special districts
Areas where the city adds its own rules on top of base zoning, from each city's official maps, in four colors: special plans — special purpose and special use districts, specific plans, special area plans, planned developments (PD, PUD) and institutional master plans; local historic districts; overlay districts such as height limits, design review, station areas and inclusionary housing areas; and redevelopment, urban renewal and tax increment (TIF, TIRZ) areas. In New York, Hudson Yards is the Special Hudson Yards District, with its subdistricts A to F drawn inside it. Areas that only restrict a kind of business (liquor sales, check cashing, signs) are left out. The legend lists only the kinds drawn on the sheet; sources are on Data sources.
Read it before the zoning numbers: inside a special district the district's own chapter of the code can change the base limits. It is not an official zoning determination.
10 · Green spaces
Parks, other green, and water, from Overture. Water gets its own treatment because in New York it structures everything — the waterfront edge is the single strongest organizing element in most Manhattan and Brooklyn sites.
Note: this shows mapped open space, which is not the same as accessible open space. A green polygon may be a private courtyard, a cemetery, or a highway median. Check it against the aerial.
11 · Public facilities
Schools, civic and cultural buildings (city halls, courts, police and fire stations, libraries, community centers, theaters, museums, stadiums), markets, and parking lots and rail land — the same four groups as the Korean edition's planned facilities panel. The US has no designation of «planned facilities», so these are the sites as they are mapped in OpenStreetMap, not designated land. Parking lots under 1,600 sq ft (150 m²) are left out; parks are on the green spaces panel.
Read it for what serves the site and what fills the blocks around it — in many US downtowns the parking color is the biggest one on the sheet.
12 · Roads
Street centerlines weighted by hierarchy — arterials heavier, local streets lighter. Source is Overture.
Read it for the armature. Where does through-traffic go, where does the grid break, which edge of your site is a front and which is a back? The weighting is what makes that legible; an unweighted street map is just a net.
13 · Subway & bus
Rail and bus stops from OpenStreetMap, as points — in Houston the bus stops come from METRO's own stop list (Route and arrival data provided by permission of METRO), and in Denver stops and stations come from RTD's schedule data. Combined with the radius, this is your transit accessibility read — and it is why the ½ mile frame exists.
Note: stop locations are well mapped in New York, Los Angeles, San Francisco, Boston, Washington, DC, Chicago, Miami, Dallas, Houston, Denver, Seattle and Minneapolis, but OSM does not carry service frequency. Two stops are not equal if one is a 4/5/6 express station and the other a local bus with 30-minute headways. The panel shows access, not service.
14 · Walking distance
How far you get on foot from the site in 5, 10 and 15 minutes, measured along streets and footpaths rather than as a circle. The network is OpenStreetMap; motorways and expressways are left out because nobody walks them. Walking speed is about 3 mph (262 ft a minute).
Note: signal waits, grades and stairs are not modeled, so the edge is an estimate. Where the shed stops short at a highway, a rail line or the waterfront, that edge is usually the most useful thing on the panel.
15 · City block
Block polygons — the shapes enclosed by the street network. This is grain made explicit, and it is the panel that most often carries an argument on its own. Fine blocks mean more corners, more frontage, more route choice; superblocks mean the opposite.
Comparing block sizes across a seam in the fabric is usually a stronger statement than any density number.
16 · Density
A heat field derived from the buildings, by total floor area or by building count — your choice. It is an interpretation, not a measurement, which is why it comes after the fabric panels rather than before.
Read it for pattern and gradient, not for absolute values. “Density falls off sharply west of the avenue” is a defensible reading; “this block is at 8.2 FAR” is not something this panel can tell you.
17 · Terrain
Contours and elevation bands from USGS 3DEP bare-earth data, with water flattened to its own surface so the harbor does not sprout hills. Contour intervals are chosen in feet — 10, 20, 50 — so the lines land on round elevations the way a survey drawing does.
Note: New York is mostly flat, and on a flat site this panel says “flat”, which is real information. Where it earns its place is upper Manhattan, the Bronx and Staten Island, where relief genuinely shapes the street pattern. Los Angeles is the opposite case — the hills from Silver Lake to the Santa Monica Mountains make this panel and Slope central to many sites.
18 · Slope
Terrain answers “how high”; slope answers “how steep.” The elevation grid is differentiated and banded as percent grade — 5, 10, 15, 25% — the units American site work already uses. For reference, 5% is the ADA limit for a walking surface without handrails and 8.33% (1:12) is the steepest a ramp may run.
Note: slope is the derivative of elevation, so elevation noise is amplified. The grid is smoothed over roughly 164 ft before differentiating, which means features narrower than that — retaining walls, single cuts — do not appear. The caption says so. Read it for landform, not for a cut-and-fill estimate.
19 · Cadastral
The same tax lots as Plot size, in one color with white lot lines — the lot pattern without the size bands. On a grid of similar lots the two panels look alike; where lots have been assembled or split, Plot size shows the difference.
20 · Plot size
Tax lots from MapPLUTO in New York, the County Assessor in Los Angeles, DataSF parcels in San Francisco, MassGIS parcels in Boston, Office of Tax and Revenue parcels in Washington, DC, Cook County Assessor parcels in Chicago, Miami-Dade Property Appraiser parcels in Miami Dallas Central Appraisal District parcels in Dallas, Harris Central Appraisal District parcels in Houston, City and County of Denver assessor parcels in Denver King County Assessor parcels in Seattle and City of Minneapolis assessor parcels in Minneapolis, colored by area. Bands are in square feet, anchored on the New York standard lot — 25 × 100 ft = 2,500 sqft; the common Los Angeles residential lot, 50 × 150 ft = 7,500 sqft, falls one band up. Where a run of brownstones survives it shows up as a striped band; an assembled site reads as one large field against it.
This is the panel to bring to a feasibility conversation. Small lots under single ownership are a different project from small lots under forty owners, and the map cannot tell you which — but it tells you the question is there.
21 · Trees
Canopy from NAIP aerial imagery. NAIP carries a near-infrared band, and leaves reflect near-infrared strongly while asphalt and roofs do not, so canopy separates far more cleanly than color alone allows.
We checked it against New York City's own street tree census: 86% of censused trees fall inside the detected canopy (3,497 trees across seven neighborhoods), with Times Square at 1.6% and the Hudson at 0.0%. In Los Angeles the same check against the city's street tree inventory gives 84.5% (2,224 trees across eight neighborhoods), in San Francisco 89.2% (4,144 trees across eight neighborhoods) and in Boston 86.4% (2,440 trees across eight neighborhoods). The panel is on in those four cities only — the threshold is validated there and nowhere else. Details in Trees.
22 · Sun hours
How many hours of direct sun each point receives across a whole day. Shadows are solved every ten minutes from sunrise to sunset and summed. Buildings are extruded from recorded height; anything with no height is not in the model.
The sheet reports daylight, mean sun hours and the share of ground under one hour. The Korean edition's sunlight criteria are not shown here — there is no US equivalent, and New York's CEQR shadow assessment, for one, asks a different question. Use the map to compare positions, and your jurisdiction's rule for the number that matters. See Sun hours.
23 · Sky view
How much of the sky is visible from each point on the ground, from 0 to 100% (the sky view factor). Surrounding building heights are sampled in 16 directions out to about 1,000 ft. Enclosed ground — narrow streets, courtyards — is dark blue; open ground is pale. It uses the same height model as Sun hours and, like it, is drawn up to a 3,280 ft radius.
Note: buildings only. Trees and terrain are not in the model, so a street under heavy canopy reads more open than it is.
24 · Wind rose
Direction and speed frequency for the year, from the nearest weather station. This is observation, not simulation — it tells you where wind comes from and how often, which is what orientation decisions actually need. The airflow simulation is a separate tab, not a panel.
25 · Choice
Angular choice from space syntax: how often each street segment lies on the least-angle routes between all pairs of segments within a 3,940 ft (1.2 km) network radius. The darker the street, the more often it lies on routes through the area. It's computed from the street layout alone, not from counts of people or cars. Divided roads drawn as two one-way lines are merged into one.
Use: picking the street for the main entrance or ground-floor shops.
26 · Visibility by height
Where a building on the lot would be seen from the street, in height bands. Dark areas see it even if it's low; light areas only if it's tall. Unshaded areas don't see it even at the top height band. The lot is the picked parcels, or the parcel under the site point.
Use: height studies and view analysis — how tall to build to be seen from where.
27 · Street enclosure
Street width (D) ÷ height of the buildings on both sides (H), measured along each street. Darker streets are narrow and lined with tall buildings; lighter ones are open. Streets with no buildings close enough to measure are drawn as a thin gray line.
Use: street sections (building height, setbacks) — spotting streets that feel too closed or too open.
28 · Aspect
Which way the slopes face, in eight directions, shaded by steepness. South-facing slopes get more sun; north-facing ones are shadier. Strong colors are steep, pale colors gentle, and unshaded areas flat.
Use: layout and planting on sloping sites.
What is not here, and why
Every panel of the Korean edition is here. Houston has no zoning ordinance, so the zoning panel is not drawn there (27 panels).
Using them as a set
The panels are built to be composed. The Sheet tab lays your selection onto one board — A4 through A1 or 16:9 — with titles, legend, scale bar and border, at print resolution. The Layer stack tab lays them out in axonometric, stacked, which is a good report cover and a good way to show that the analysis is systematic rather than assembled.
And every panel carries its source note in the exported file. Leave it in.
Next: Finishing the SVG in Illustrator →
← Back to sitedia ·
한국어