Depave Fort Lauderdale

This map shows where Fort Lauderdale carries the most non-essential pavement, and which neighborhoods stand to benefit most from removing it. It is a screening tool for turning heat-trapping, runoff-generating asphalt back into living ground.

Data tier 2 of 4 — Imagery + ancillary Fort Lauderdale uses 1-meter NAIP imagery classified in-house, FDOT surveyed road widths, and a lidar-plus-soil topographic stormwater proxy. Pavement detection is strong, but the flood layer is an unvalidated topographic proxy rather than a hydraulic model. Neighborhood-screening reliability. Tiers reflect input-data availability, not effort. Scores are relative within each city, so the tiers and the numbers are not comparable across cities, and reliability decreases from Tier 1 to Tier 4.

Why depaving matters

Pavement drives up summertime temperatures and concentrates stormwater runoff during Florida's rainy season. It also displaces the tree canopy that would otherwise cool streets and soak up rain. Some of that pavement is essential: roads and sidewalks carry the movement a city depends on. Plenty of it is surplus, though, like parking aprons, oversized driveways, and the unused back lot behind a strip mall. The work begins by finding where that surplus sits and which neighborhoods carry the greatest environmental burden.

How we identify pavement

We start with the U.S. Department of Agriculture's NAIP aerial imagery: 1-meter, 4-band (red, green, blue, near-infrared) photography flown over Florida every couple of years. A machine-learning classifier, a random forest, learns what pavement looks like by pulling its own training samples. Points along known OpenStreetMap road centerlines are labeled as pavement, building footprints from the Microsoft Global Buildings dataset as buildings, pixels with strong vegetation-index signals as plants, and known water bodies as water.

The classifier then labels every pixel in the NAIP imagery across the city. We refine the raw output in two passes. First we burn OpenStreetMap road centerlines back in at realistic lane widths, repairing places where tree canopy hid the road from the camera. Then we subtract airport footprints, so the runways and taxiways that will never be depaved don't inflate the numbers. Everything outside the Fort Lauderdale municipal boundary is clipped away.

To separate optional from essential pavement, we compare each piece against a "core" mask built from road centerlines and sidewalks. For state and federal roads we use FDOT Roadway Characteristics Inventory data with surveyed surface widths; for local streets we use OpenStreetMap centerlines with class-based width estimates. Sidewalks belong in the core mask. Parking lots, driveways, and service roads do not, so they become depave candidates. Anything inside the core mask is core pavement, and anything outside it is non-core.

Pavement inside airports and parks is excluded entirely, since depaving runways or park paths falls outside the scope of this analysis.

How we identify priority areas

For each of the city's 47 census tracts we compute four need scores, each normalized to a 0–1 scale:

Heat: mean surface temperature from the 2021 NOAA/CAPA heat-mapping campaign.
Stormwater flood risk: a proxy derived from lidar terrain and SSURGO soil permeability. Where does water pond, and where does the soil drain too slowly to take it away?
Canopy deficit: one minus the fraction of the tract covered by tree canopy (University of Miami lidar-derived canopy layer).
Pavement burden: non-core pavement acres divided by tract area.

We average the four scores into a single composite and flag the top quartile (tracts at or above the 75th percentile) as priority tracts. Overlapping the priority tracts with the federal Climate and Economic Justice Screening Tool (CEJST) disadvantaged-community designation highlights the equity hotspots, the places where environmental need and historical disinvestment coincide.

Headline findings (approximate, latest pipeline run)

Approximately 4,941 acres of pavement citywide, of which about 1,970 acres (40%) are non-core and therefore candidates for depaving.
Approximately 344 acres of pavement inside airports and parks were excluded from the analysis.
12 of 47 tracts (about 26%) flagged as priority on the composite needs score.
19 tracts are CEJST-designated disadvantaged communities, and 8 of those also fall in the priority set, where equity and need overlap.

Key caveats

Pre-screening only. This tool scopes where to look first. Choosing specific parcels is a separate step, and site-level decisions require ground-truthing, utility checks, ownership review, and community input.
FEMA flood zones are not used. FEMA maps regulate coastal and riverine flooding from hurricanes and storm surge, a hazard that depaving cannot meaningfully reduce. Our pluvial (rainfall) proxy targets the flooding depaving actually addresses.
Airport and park pavement is excluded. Runways, taxiways, and apron pavement at FXE and other airfields are removed before analysis, along with pavement inside 279 city and community parks.
The stormwater layer is a topographic proxy. It combines surface depressions and flow paths from lidar with soil permeability to rank where rainfall is most likely to pond, relative to the rest of the city. It does not model rainfall depth, drainage pipes, tides, or groundwater, and it has not yet been validated against observed flooding. Read it as a screen for where to look first, not a flood forecast.
Classifier imperfections. Concrete rooftops can look like pavement, and heavily shaded driveways can look like vegetation. The map is directionally accurate at the tract scale. At the parcel scale it is a starting hypothesis.

For the full technical methodology, including data sources, algorithms, hyperparameters, and known limitations, see the detailed methodology.