Methodology & FAQ

Key variables

Scoring method

Element score

county_value / regional_baseline × 100. All nine elements are distress measures, where a higher rate is worse: no broadband subscription, cellular-only, satellite, poverty, unemployment, SNAP, age 65+, education, disability. The workbook also carries raw columns that are not index elements, among them wireline subscription and median income, and this formula does not apply to those.

The baseline

A weighted regional aggregate. The household rates (the three technology elements and SNAP) are weighted by households. Unemployment and age 65+ are population-weighted means of the county rates, using the civilian noninstitutionalized population. The weighting of the three elements whose source is unconfirmed is not known. The simple mean of the 18 county scores comes to 109.1, so reading 100 as “the county average” gets it wrong.

Components

Each is the mean of three elements; OVERALL is the unweighted mean of the three components.

Fixed wireless depends on a clear line of sight from a tower to a home, and ridgelines and dense forest canopy break that line of sight regardless of how much a provider invests. Running new wireline gets more expensive as slopes get steeper. The county maps show this terrain from lidar, so you can see the ground a tower signal or a trench has to cross.

Baseline

Baseline

The weighted aggregate over the 18 Alliance counties (the workbook’s DOGWOOD sheet). The simple mean of the county scores is a different, higher figure.

Polarity

Lower is better. A score above 100 means more distress than the regional baseline.

Ranks

Rank 1 is the most distressed county (the highest score) and rank 18 the least. Overall, technology, economy and skills are each ranked on their own.

Source

Sources

Caveats

• Provider-reported. The Broadband Data Collection records what a provider says it could serve at an address within ten business days of a request. It is a claim about capability. No speed test is involved, and it says nothing about who subscribes.
• Two releases in one map. Coverage is FCC BDC release 8, NC DIT layer updated 19 June 2026. The provider detail is release 6, NC DIT layer updated 10 July 2025. NC DIT publishes no release 7 or 8 provider-level layer, so the provider view uses an earlier FCC release (6) than the coverage figures (8) and the two will not reconcile. Every provider percentage is a share of its release-6 count.
• Fixed wireless is one class here. The source carries three fixed wireless technology codes: unlicensed, licensed, and licensed by rule. They share a colour because all three depend on a clear path from a tower, which is the thing the terrain is here to explain. The hover tooltip lists every provider by name. It does not say which of the three codes a provider reported. Where one provider reported two of them, which happens in seven of the eighteen counties, that provider appears twice in the same list, sometimes against the same percentage twice.
• A share of locations. Colour is a share of the locations in a hexagon. Its area plays no part, so a sparsely settled hexagon and a dense one are coloured on the same scale. Height is what separates them.
• Columns hide things behind them. That is what extrusion costs. A tall column occludes shorter ones behind it and the ground beneath it, and the footprints are drawn at 82% so the terrain still reads between them. Orbit to see what a fixed view hides.
• One light on everything. The columns are lit from the same direction the shaded relief was baked from: azimuth 277 (just north of west) and altitude 75.
• Scores at component level. The index scores are Technology, Economy and Skills, plus the overall score. Each county page also shows five rates from the same ACS 2020–2024 workbook: households with a wireline subscription, households with a cellular data plan, unemployment, SNAP households and population age 65+. The other elements are placeholders until their sources are confirmed.
• Highest single provider. The provider layer carries one row per provider per hexagon, and two providers can serve the same address. Adding their served-location counts would count that address twice. Each hexagon is coloured by the highest share any one provider of the selected technology reports, which is a lower bound on that technology’s real reach.
• Height is on a square root. A column twice as tall holds four times the locations. The key gives ticks at real counts for that reason. Colour is the only channel here that is linear in what it measures.
• Bare ground means two different things. A hexagon only stands up if the selected measure found something in it. Bare ground is either a hexagon with locations and none of them unserved, which is the good case, or a hexagon that holds no broadband-serviceable location at all and is never drawn.
• Each column sits on its own highest ground. A hexagon is about 930 m across. Each column is measured from the highest ground in its own footprint and driven 180 m into the hillside, which is why no base floats above the terrain.
• Vertical scale. The county maps are drawn at true vertical scale, with no exaggeration. The regional map doubles heights (2×) so the relief still reads across 18 counties.

Terrain: NC OneMap DEM03 · EPSG:6542 · sub-metre lidar (3.125 ftUS native) → 2 m master → 64 m web grid, clipped to the county line.
Scores: ARBSI__ARC_202024.xlsx, DOGWOOD sheet.
Broadband: FCC Broadband Data Collection via NC DIT (NCDIT-BIO, CGIA) · coverage from FCC_BDC_Hexbins_v8, release 8, NC DIT layer updated 19 June 2026 · provider detail from fcc_bdc_provider_hexbins_v6, release 6, NC DIT layer updated 10 July 2025 · H3 level-8 hexagons, clipped to county lines by NC DIT.
Cited as NC OneMap asks: “NC OneMap (2019), NCDIT, GDAC, CGIA”, their portal citation, which does not give the date of the lidar. Lidar by the NC Floodplain Mapping Program, NCDPS Division of Emergency Management.

FAQ