Under the Canopy

The short version

• Across the 14-county region, tree canopy slipped from about 44.1% of land in 2015 to 43.5% in 2023, a loss of 0.6 percentage points, or roughly 1.4% of the region’s tree cover in eight years.
• The regional average hides a sharp tail. Nearly half of tracts (45%) lost canopy, and about 1 in 8 (12%) lost more than three percentage points. A handful lost more than ten.
• Losses concentrate in fast-growing exurban tracts, especially in the South Carolina border counties (Lancaster, York) and the southern edge of Mecklenburg. Where new subdivisions went up, trees came down.
• The relationship with wealth runs opposite to the national pattern. Nationally, lower-income neighborhoods have less canopy (McDonald et al. 2021). Here, in the region’s urban clusters, higher-income neighborhoods lost more canopy over this period, not less, because the loss tracks new construction, and new construction concentrates in affluent, growing areas.

1. A region still green, but thinning

At 44% tree cover, the Charlotte region is leafy, well above most U.S. metros. But the trend line bends the wrong way. Between 2015 and 2023 the region’s canopy trended steadily downward, settling about half a point lower (Figure 1).

Half a point sounds trivial. It is not, for two reasons. First, a regional average pools rapidly developing areas with vast tracts of protected or rural forest that change little, so it dilutes real local losses. Second, the losses are lumpy. Figure 2 shows the distribution of change across the region’s 752 tracts: most changed little, but a long left tail of tracts shed canopy heavily.

This matches, at a regional scale, what the City of Charlotte’s own high-resolution assessments have been warning for years: the city’s canopy fell from 49% to 45% between 2012 and 2018, a net loss of some 250,000 trees, and continued sliding to about 47% by 2023 (TreesCharlotte and University of Vermont Spatial Analysis Laboratory 2019; TreesCharlotte 2023; UNC Charlotte Urban Institute 2023). Our regional data extends that story outward from the city line to the exurban counties where the region’s growth is now happening.

2. Where the losses land

Canopy change is not evenly spread. Figure 3 maps it tract by tract. The brown (canopy loss) clusters along the region’s growth frontier: the southern arc of Mecklenburg into Union, and across the state line into the booming Lancaster and York county suburbs.

By county, the pattern is unambiguous (Figure 4). The heaviest losses are in the South Carolina border counties, Lancaster lost nearly 4 points of canopy and York more than a point, exactly the places absorbing Charlotte’s southward spillover. The slower-growing western counties (Cleveland, Gaston, Catawba) gained canopy as existing trees matured.

Zoom to the tract level and the losses have names. The ten hardest-hit tracts (Table 2) read like a map of the region’s construction boom: Providence / Blakeney in south Charlotte (78% of its homes built since 2010), Walnut Creek / Bent Creek in Lancaster (85% new), The Sanctuary on Lake Wylie, and Steele Creek, the same southwest Charlotte area the city’s studies flag as critically low and heavily developed (TreesCharlotte and University of Vermont Spatial Analysis Laboratory 2019).

3. Follow the bulldozers

The single strongest correlate of canopy loss in the data is not income, or race, or geography. It is development. Where the share of recently built homes is high, canopy change is negative (Figure 5); the correlation is r = −0.34. The relationship is even tighter against physical development: canopy change correlates −0.50 with the growth of impervious surface (roads, roofs, parking), the clearest signature of land being cleared and paved.

The gradient is stark. Tracts in the top quarter for recent construction lost 1.3 points of canopy on average; tracts in the bottom quarter gained 0.9 points. In slower-growing, built-out neighborhoods, existing trees kept growing; several dense Mecklenburg tracts gained six or seven points as their canopy matured. The region is running a race between two processes: mature trees filling in on one side, and forest cleared for subdivisions on the other. Since 2015, clearing has been winning.

4. The income twist

Here is where the Charlotte region complicates the national story. The most cited fact in urban forestry is that trees are unequally distributed by wealth: across 5,723 U.S. communities, lower-income blocks average 15% less tree cover and run about 1.5°C hotter in summer than higher-income blocks (McDonald et al. 2021). Programs like American Forests’ Tree Equity Score are built on closing exactly that gap (American Forests 2021).

That baseline gap exists here too, but it is modest. Restricting to the region’s urban clusters (tracts above 500 people per square mile, which excludes the rural fringe), higher-income neighborhoods held about 41% canopy in 2015 versus 37% in lower-income ones, a gap of a few points, not the chasm seen in older, denser metros.

The surprise is in the change. Over 2015–2023, canopy was more stable in lower-income urban neighborhoods, not less (Figure 6). Lower-income tracts gained a bit on average (+0.7 points), while higher-income tracts lost ground (−0.3 points), and nearly half of higher-income tracts lost canopy versus a third of lower-income ones.

Why the reversal? Because in a fast-growing Sun Belt region, canopy loss is driven by new development, and new development lands disproportionately in affluent, tree-rich, greenfield areas, the Providence / Blakeney and Lake Wylie subdivisions of the world, that had abundant canopy to lose. The region’s lower-income urban cores, by contrast, are already built out; they have less canopy to begin with, but also less new clearing, and some are quietly re-greening as street trees mature. The City of Charlotte’s own assessments tell the same story from the opposite direction: some of the steepest recent losses were in affluent, leafy neighborhoods like Myers Park, Dilworth, and Eastover, where large-lot teardowns and rebuilds removed mature trees (TreesCharlotte and University of Vermont Spatial Analysis Laboratory 2019; UNC Charlotte Urban Institute 2023).

The takeaway is not that tree equity is a non-issue here. The level gap is real, and it compounds a heat gap. But the dynamic threat over this period is different from the disinvestment story: it is a growth story. Protecting the region’s canopy means governing how the region builds.

Other threads worth pulling

Canopy and heat move together. The tightest correlate of canopy loss was impervious-surface growth (r = −0.50). As trees give way to pavement and roofs, the same tracts trap more heat, the mechanism behind the canopy-temperature disparity documented nationally (McDonald et al. 2021). Canopy loss is, in effect, an early-warning signal for future urban-heat exposure.

Race is not the axis here; development is. Unlike the strong racial patterning of canopy in many older cities, canopy change in this region shows almost no correlation with a tract’s share of Black residents (r ≈ 0.05). The engine is where and how fast land is being built, not the demographics of who lives there.

The city-versus-region gap is a lesson in measurement. The City of Charlotte reports ~47% canopy; our regional NLCD figure is ~44%. They differ because they measure different things over different areas with different sensors (see below). The value of a regional view is that it captures the exurban frontier, Lancaster, York, Union, that a city-limits study cannot see, and that is precisely where the sharpest losses are now occurring.

What it means

The Charlotte region is not running out of trees. But it is trading a slow, compounding loss of mature canopy on its growth frontier for the houses and roads a booming region needs, and it is not replanting fast enough to keep even. The losses are concentrated, predictable, and tied to development, which means they are also, in principle, governable, through tree-protection ordinances, replanting requirements, and conservation of the large forest tracts that fall fastest. The region’s own tree-canopy goals will not be met by street-tree plantings alone while subdivisions clear woodland faster than saplings can grow (TreesCharlotte 2023).

The equity conversation deserves a regional update, too. The familiar, important call to plant trees in under-canopied low-income neighborhoods still holds. But in a growing Sun Belt metro, an equally urgent task is to keep the canopy that development is removing, much of it in the affluent, fast-building edges where, counter to the national script, the trees are coming down fastest.

Data and methods

Tree canopy. Percent tree-canopy cover per census tract from the USGS/USFS National Land Cover Database (NLCD) Tree Canopy Cover product (U.S. Geological Survey and U.S. Forest Service 2023), a satellite-derived, 30-meter modeled estimate, for 2015 through 2023, as compiled in the Carolinas Regional Explorer. Regional and county figures are area-weighted across tracts (each tract’s canopy percentage weighted by its land area, derived from population ÷ population density), because canopy is a share of land. Because NLCD is a modeled raster product, it is not directly comparable to the City of Charlotte’s parcel-level, aerial-imagery assessments (TreesCharlotte and University of Vermont Spatial Analysis Laboratory 2019); NLCD tends to read differently, especially in fine-grained urban settings, and it carries no survey margin of error, so small single-tract changes should be read with care. Both measurement approaches agree on the regional direction of travel.

Housing and income. Housing activity is measured by the share of a tract’s homes built since 2010 and by change in developed and impervious land (NLCD); income is median household income; both from the American Community Survey via the Explorer. “Urban clusters” are tracts above 500 people per square mile, which excludes the region’s most rural tracts (198 of 752) from the income analysis, as the question specified. Income groups are tertiles of median household income within those urban tracts.

Correlations are descriptive (Pearson’s r), not causal, and canopy change over an eight-year window should be read as an association with development, not proof of it. The full tract-level table and the scripts that produced every figure are in this story’s folder.

Explore the underlying indicators (tree canopy, homes built since 2010, income, and more) for any tract in the Carolinas Regional Explorer.

References