Square Miles to Square Kilometers for Coverage Planning

Licence Areas in Square Miles, Link Budgets in Metric

The build plan lands on your desk quoting a serving area of 412 square miles and a target of 1,200 subscribers per square mile. Then you open the planning tool: path loss in dB over distance in kilometres, cell footprints reported in km², and a vendor site-count model whose spreadsheet columns are all metric. Half the inputs are in one system and half in the other before a single antenna has been picked.

Converting the area early — before the site count, before the capacity check — keeps the whole model in one language and stops a factor of 2.59 hiding inside a business case.

Conversion factor: 1 mi² = 2.589988 km², so 412 mi² is 1 067 km². Per-area figures reverse: 1,200 subscribers per mi² is 1,200 × 0.386102 = 463 per km², since a square kilometre covers less ground.

Where the Mismatch Shows Up in a Coverage Plan

Filings measure in square miles

Licence footprints, county serving areas and broadband availability reporting are all framed in square miles, because the underlying census geography carries land area that way.

Propagation output is metric

Recommendation-based path loss models take distance in kilometres and frequency in MHz, so predicted footprints and clutter statistics arrive back in km² whatever you fed in.

Site counts follow area, not reach

Halving the usable radius quarters the footprint, so an error in the area unit multiplies straight through the tower count and the capital plan behind it.

Demand figures invert the ratio

Subscribers, homes passed and offered traffic per unit area shrink when restated per km², which is the opposite of what happens to the coverage polygon itself.

Getting a Build Plan onto One Set of Units

Both fields are live and both accept typing, so a planning session can move between the filing and the tool without anyone reloading anything.

Put the imperial area in

Enter the mi² value on the left — 412 for a serving area, 10.4 for a single macro footprint. The box starts at 1 and the metric side follows every keystroke, so there is no submit step. Dots and commas both read as the decimal mark and spaces are ignored.

Set the unit on each side

The dropdowns list all 13 area units under metric and imperial headings with a filter box at the top — type km or the symbol itself, and Esc shuts the list. Whichever unit you choose is promoted to the top for the next lookup.

Reverse it for the tool output

When the planning tool hands back a footprint in km² and the report template needs imperial, type into the metric box instead, or press the swap arrows so the pair runs the other way for the rest of the session.

Send it to the site-count sheet

Each box has its own copy control that puts the plain number on the clipboard, without the unit and without the space between thousands, so it lands in a model cell as a number. Ctrl + C over a selection behaves the same way.

Watch the terrain assumption: a converted area is exact, but the footprint it came from was not. A predicted radius on flat rural ground and the same radius in a river valley can differ by more than the unit conversion ever will.

Cell Radius, Footprint per Site and Sites per 100 mi²

Planners size a build from a hexagonal cell laid out around each site, whose area is 2.598 × the radius squared. Below, each radius is worked through to a footprint in both units and to the number of sites needed for a 100 mi² block of the licence area, rounded up because part of a site is not a thing you can build.

Cell radius	Footprint per site (mi²)	Footprint per site (km²)	Sites per 100 mi²
0.25 mi (dense small cell)	0.162	0.42	616
0.5 mi (urban macro)	0.65	1.68	154
1 mi (suburban macro)	2.60	6.73	39
1.5 mi	5.85	15.14	18
2 mi	10.39	26.9	10
3 mi (rural low-band)	23.4	60.6	5
5 mi	64.9	168.2	2
10 mi (fixed wireless, clear line of sight)	259.8	672.9	1

Read down the last column and the economics of spectrum choice appear on their own. Going from a 3-mile low-band cell to a quarter-mile small cell is a jump from 5 sites to 616 over the same block — more than 120 times the structures, backhaul drops and lease agreements, for a licence area that never changed size.

Sanity check both directions: multiply mi² by 2.589988 for the footprint, but multiply per-area demand by 0.386102. A capacity model fed 1,200 per km² when the source said per mi² overstates offered traffic by a factor of 2.59.

What Helps While Building a Site-Count Model

Typing works on either side

Filing figures go in on one side, tool output on the other, and the swap arrows fix the direction for a run of numbers so you are not re-reading the labels every time.

Clean values for the model

The copy button yields the bare figure, so a converted footprint pastes into a capital-plan cell as a number instead of arriving as text with a unit stuck to the end.

Every area unit in one picker

Thirteen units sit behind a searchable list on both fields, which covers the hectare and acre figures that turn up in tower-lease and easement paperwork alongside the coverage numbers.

Plan data stays local

The arithmetic is done by the page itself, so unannounced build footprints and competitive coverage figures are not being posted off somewhere while you work through them.

Coverage and Site-Count Questions

How many cell sites does a 400 mi² county need?

That county is 1 036 km². With a 2-mile usable radius each hexagonal cell covers 10.39 mi² (26.9 km²), giving roughly 39 sites for coverage alone. Push the radius to 3 miles on low-band and it drops to about 18; pull it back to 1 mile because the band is mid-band or the clutter is heavy and it climbs past 150. Capacity can then force more sites than coverage does, so the figure you get here is a floor, not the final count.

Why is a cell area 2.6 r² instead of πr²?

A circle of radius r has area 3.1416 r², but circles tile a plane only by overlapping. Planning layouts use the regular hexagon inscribed in that circle, whose area is (3√3)/2 = 2.598 times r² and which tessellates with no gap and no double counting. The missing 17 % is exactly the overlap you would otherwise count twice. Use the hexagon figure for site counts and reserve the circle for a single isolated site with nothing adjacent.

How do I restate subscribers per square mile as subscribers per square kilometer?

Multiply by 0.386102, or divide by 2.589988 — the same operation either way. So 1,000 per mi² is 386 per km², and 1,200 per mi² is 463 per km². Note that this is the reverse of what you do to the area itself, which is the single most common slip when a US market plan is handed to a metric capacity tool. A fast check: the number of subscribers in a market must not change when you switch units, only the area you divide it by.

How far does coverage per site fall moving from low band to millimeter wave?

Steeply, because area scales with the square of whatever reach you lose. A rural 700 MHz macro working to about 3 miles serves 23.4 mi², or 60.6 km². A millimetre-wave node held to roughly 0.15 miles by building loss and foliage serves about 0.058 mi², near 0.15 km² — on the order of 400 nodes to replace one macro footprint. This is why high band is deployed as capacity infill in dense pockets rather than as an area-coverage layer.

Why do availability maps report square miles when planning tools work in km²?

They inherit different ancestries. Availability and licence reporting is built on census geography — blocks, tracts and counties whose published land area is imperial and whose boundaries trace a survey grid laid out in sections of one square mile. Propagation modelling grew out of international recommendations that specify distance in kilometres, so every footprint a tool draws is metric by construction. Neither side is going to move, which is why converting once at the boundary between them, and labelling the column, beats converting repeatedly inside the model.

Square Miles to Square Kilometers