Why does 60 GHz lose about 15 dB per kilometre?

Because oxygen absorbs there. Oxygen has 37 absorption lines between 50 and 69 GHz, and at sea-level pressure they merge into one broad band. At 1013.25 hPa, 15 °C and 7.5 g/m³ the calculator gives 14.50 dB/km from oxygen and 14.66 dB/km in total, so a 1 km hop loses about 15 dB to the air alone. For short links that is useful. The signal fades before it can disturb the next link, which is why 57–71 GHz is popular for dense short-range links.

Why is there a bump at 22 GHz, and does it hurt?

It is a water-vapour line, and a weak one. At 7.5 g/m³ it adds about 0.18 dB/km at 22.235 GHz, which only matters over long terrestrial paths or through the whole atmosphere on an Earth-space link. Weather radiometers use it to measure humidity. The strong water line sits at 183.31 GHz, where the same air absorbs 28 dB/km.

Does altitude matter?

Yes, through the air you enter. Use the pressure, temperature and humidity at the height of the path. In the ITU-R reference atmosphere at 2 km (795 hPa, 2 °C, 2.76 g/m³) the total drops from 14.66 to 13.32 dB/km at 60 GHz, from 0.40 to 0.14 dB/km at 94 GHz and from 28.3 to 13.9 dB/km at 183 GHz. The oxygen band barely moves because thinner air makes its lines both weaker and narrower, so their peaks stay tall.

Why are slant paths to satellites out of scope?

Because P.676-13 has no short formula for them. Its exact method integrates through 922 layers of a full height profile, and its approximate method needs a 700-row coefficient file for oxygen. Edition 12 still had a closed form, edition 13 removed it. This calculator covers horizontal and slightly inclined paths near the ground, where A = γ · d holds. For an Earth-space link, use an implementation of the full Recommendation such as ITU-Rpy.

Why does the ITU table say 14.62 dB/km at 60 GHz?

Because P.676 counts dry-air pressure. The ITU validation spreadsheet uses 1013.25 hPa of dry air plus 9.97 hPa of water vapour, 1023.22 hPa in total. A barometer shows the total, so this calculator treats 1013.25 hPa as total pressure and hands P.676 a dry pressure of 1003.28 hPa. Oxygen then drops from 14.62 to 14.50 dB/km. Enter 1023.22 hPa and you get the spreadsheet’s 14.62 and 14.78 dB/km.

← All tools

Atmospheric gas attenuation calculator (ITU-R P.676)

Atmospheric gas attenuation is the loss that oxygen and water vapour add to a radio path, in dB per kilometre. This calculator applies ITU-R P.676-13 line by line from 1 to 1000 GHz for your pressure, temperature and humidity, then multiplies by the length of a terrestrial path. Slant paths to satellites are out of scope.

Atmosphere

The global reference profile of P.835-7 at sea level, called the mean annual global reference atmosphere in earlier editions. Picking it fills in pressure, temperature and humidity. Editing any of them switches back to custom.

Frequency

P.676 Annex 1 is valid from 1 to 1000 GHz.

Barometric pressure (total)

The total air pressure along the path, as a barometer shows it. On high ground use the station pressure, not the sea-level value (QNH) from a weather report. The calculator subtracts the water-vapour part itself.

Air temperature

Relative humidity can only be converted between −40 and +50 °C (P.453).

Humidity given as

Water-vapour density

g/m³

7.5 g/m³ is the ITU standard value at sea level. Dry air is 0, humid tropical air reaches 20–25 g/m³.

Path length

A horizontal or slightly inclined path near the ground, through the same air all along.

Specific attenuation, total

14.66

oxygen and water vapour together

Path attenuation A

14.66 dB

γ × 1 km, gases only

Oxygen and dry air

14.50

at dry-air pressure p = 1003.28 hPa

Water vapour

0.1536

at water-vapour pressure e = 9.973 hPa

Pressure handed to P.676: dry air p = P − e = 1013.25 hPa − 9.973 hPa = 1003.28 hPa. Water vapour: ρ = 7.5 g/m³, relative humidity 58.2 %.

Specific attenuation against frequency for the air you entered, both axes logarithmic. The solid green curve is the total, the grey dashed curve the oxygen and dry-air part, the light green dotted curve water vapour. Dotted vertical lines mark the absorption peaks at 22.235 GHz (water vapour), 60 GHz (oxygen band), 118.75 GHz (oxygen) and 183.31 GHz (water vapour). The dot is your frequency. The axis runs to 350 GHz, or to 1000 GHz once your frequency lies above 350 GHz. With zero humidity there is no water-vapour curve.

Clear air only. Rain, cloud and fog add their own loss on top. The result assumes the same pressure, temperature and humidity along the whole path, which holds for a horizontal path near the ground. Slant paths to satellites or aircraft are out of scope, because P.676-13 computes them by integrating through a height profile or with a 700-row coefficient file. Implements Recommendation ITU-R P.676-13 (08/2022), Annex 1, equations (1)–(10).

How it works

01
Describe the air

Enter the total barometric pressure, the temperature and the humidity where the path runs. Humidity can be water-vapour density or relative humidity. Relative humidity goes through the P.453 saturation formula first, so about 58 % at 15 °C and 1013.25 hPa becomes 7.5 g/m³.
02
Split off the water vapour

P.676 takes the dry-air pressure and the water-vapour pressure separately. The vapour pressure is e = ρ·T / 216.7, which gives 9.97 hPa for 7.5 g/m³ at 15 °C. The dry part is what remains, p = P − e = 1003.28 hPa.
03
Add up the absorption lines

Each of 44 oxygen lines and 35 water-vapour lines has a strength and a width that depend on p, e and T. The calculator evaluates every line at your frequency, adds them up and adds a smooth dry-air continuum. Rule of thumb at sea level: under 0.015 dB/km below 10 GHz, about 15 dB/km at 60 GHz, 0.4 dB/km in the 94 GHz window and 28 dB/km at 183 GHz.
04
Multiply by the path length

On a horizontal path the air stays the same, so A = γ · d. At 60 GHz over 1 km that is 14.66 dB. At 10 GHz over 30 km it is only 0.42 dB.

Formulas

Specific attenuation (eq. 1)

γ = 0.1820 · f · (N″_Ox + N″_W)

γ — specific attenuation, dB/km: the oxygen part plus the water-vapour part
f — frequency, GHz
N″_Ox — line strength times line shape, summed over the 44 oxygen lines, plus the dry-air continuum (eq. 8). It gives the oxygen part.
N″_W — the same sum over the 35 water-vapour lines. It gives the water-vapour part.
F_i — line shape at f (eq. 5), set by a line width that grows with pressure

Oxygen line strength (eq. 3)

S_i = a₁·10⁻⁷·p·θ³·exp(a₂(1−θ))

p — dry-air pressure, hPa: the total pressure you enter minus e, subtracted by the calculator
θ — 300 / T, with T in kelvin
a₁, a₂ — line constants from Table 1 of P.676-13
b₁, b₂ — the water-vapour lines use the same form, with e in place of p, 0.1 in place of 10⁻⁷ and θ to the power 3.5

Vapour pressure and path attenuation (eqs. 4, 10)

e = ρ · T / 216.7, A = γ · d

e — water-vapour partial pressure, hPa
ρ — water-vapour density, g/m³ (from relative humidity via P.453-14)
A — attenuation over the path, dB
d — path length, km

Worked example

60 GHz over 1 km in the ITU-R reference atmosphere at sea level

P = 1013.25 hPa, T = 288.15 K (15 °C), ρ = 7.5 g/m³
e = 7.5 · 288.15 / 216.7 = 9.97 hPa, so p = 1013.25 − 9.97 = 1003.28 hPa
oxygen and dry air: γ_o = 14.502 dB/km
water vapour: γ_w = 0.154 dB/km
γ = 14.502 + 0.154 = 14.656 dB/km, shown as 14.66
A = 14.656 · 1 = 14.66 dB

FAQ

Why does 60 GHz lose about 15 dB per kilometre?: Because oxygen absorbs there. Oxygen has 37 absorption lines between 50 and 69 GHz, and at sea-level pressure they merge into one broad band. At 1013.25 hPa, 15 °C and 7.5 g/m³ the calculator gives 14.50 dB/km from oxygen and 14.66 dB/km in total, so a 1 km hop loses about 15 dB to the air alone. For short links that is useful. The signal fades before it can disturb the next link, which is why 57–71 GHz is popular for dense short-range links.
Why is there a bump at 22 GHz, and does it hurt?: It is a water-vapour line, and a weak one. At 7.5 g/m³ it adds about 0.18 dB/km at 22.235 GHz, which only matters over long terrestrial paths or through the whole atmosphere on an Earth-space link. Weather radiometers use it to measure humidity. The strong water line sits at 183.31 GHz, where the same air absorbs 28 dB/km.
Does altitude matter?: Yes, through the air you enter. Use the pressure, temperature and humidity at the height of the path. In the ITU-R reference atmosphere at 2 km (795 hPa, 2 °C, 2.76 g/m³) the total drops from 14.66 to 13.32 dB/km at 60 GHz, from 0.40 to 0.14 dB/km at 94 GHz and from 28.3 to 13.9 dB/km at 183 GHz. The oxygen band barely moves because thinner air makes its lines both weaker and narrower, so their peaks stay tall.
Why are slant paths to satellites out of scope?: Because P.676-13 has no short formula for them. Its exact method integrates through 922 layers of a full height profile, and its approximate method needs a 700-row coefficient file for oxygen. Edition 12 still had a closed form, edition 13 removed it. This calculator covers horizontal and slightly inclined paths near the ground, where A = γ · d holds. For an Earth-space link, use an implementation of the full Recommendation such as ITU-Rpy.
Why does the ITU table say 14.62 dB/km at 60 GHz?: Because P.676 counts dry-air pressure. The ITU validation spreadsheet uses 1013.25 hPa of dry air plus 9.97 hPa of water vapour, 1023.22 hPa in total. A barometer shows the total, so this calculator treats 1013.25 hPa as total pressure and hands P.676 a dry pressure of 1003.28 hPa. Oxygen then drops from 14.62 to 14.50 dB/km. Enter 1023.22 hPa and you get the spreadsheet’s 14.62 and 14.78 dB/km.

Gas loss is the same in every direction. Terrain is not. Waveshed draws where hills and buildings block the path, so you can see whether that short 60 GHz hop has line of sight at all.

Check line of sight on the map →

Related guides

LOS vs RF analysis — gas loss is one term of the RF budget, line of sight is the other question
Plan radio coverage — where to add a fixed loss such as this one when you plan a link

Sources & further reading

ITU-R Recommendation P.676-13 — Attenuation by atmospheric gases and related effects ↗ — edition 08/2022, Annex 1, equations (1)–(10) and line tables 1 and 2, implemented here
ITU-R Recommendation P.453-14 — The radio refractive index: its formula and refractivity data ↗ — equations (8)–(10): saturation vapour pressure over water and the conversion from relative humidity
ITU-R Recommendation P.835-7 — Reference atmospheres ↗ — Annex 1, the ITU-R reference atmosphere at sea level: 1013.25 hPa, 288.15 K, 7.5 g/m³ (the preset)
ITU-R SG3 validation examples — CG-3M3J-13-ValEx Rev 8.3.0 (spreadsheet) ↗ — sheet “P.676-13 SpAtt”: the reference values this calculator reproduces to one part in a million
ITU-Rpy — A python implementation of the ITU-R P. Recommendations ↗ — open-source cross-check of the line tables, and a way to compute slant paths

Atmospheric gas attenuation calculator (ITU-R P.676)

How it works

Describe the air

Split off the water vapour

Add up the absorption lines

Multiply by the path length

Formulas

Worked example

FAQ

Related guides

Sources & further reading

Related tools

Rain attenuation calculator (ITU-R P.838)

Cloud and fog attenuation calculator (ITU-R P.840)

Free-space path loss calculator

Link budget calculator