worldsbestfacts
NASA's Deep Space Atomic Clock hardware, illustrating precision timekeeping

38 μs

GPS would fail within a day if it ignored Einstein's relativity.

5 min readLast updated

Photo: Public domainPublic domainfalsefalse This file is in the public domain in the United States because it was solely created by NASA. NASA copyright policy states that "NASA material is not protected by copyright unless noted". (See Template:PD-USGov, NASA copyright policy page or JPL Image Use Policy.) Warnings: Use of NASA logos, insignia and emblems is restricted per U.S. law 14 CFR 1221. The NASA website hosts a large number of images from the Soviet/Russian space agency, and other non-American space agencies. These are not necessarily in the public domain. Materials based on Hubble Space Telescope data may be copyrighted if they are not explicitly produced by the STScI.[1] See also {{PD-Hubble}} and {{Cc-Hubble}}. The SOHO (ESA & NASA) joint project implies that all materials created by its probe are copyrighted and require permission for commercial non-educational use. [2] Images featured on the Astronomy Picture of the Day (APOD) web site may be copyrighted. [3] The National Space Science Data Center (NSSDC) site has been known to host copyrighted content. Its photo gallery FAQ states that all of the images in the photo gallery are in the public domain "Unless otherwise noted." NASA's Jet Propulsion Laboratory · Public domain

The short answer

GPS must account for relativity because satellite clocks and ground clocks do not naturally keep the same rate. For a nominal GPS orbit, the combined relativistic rate difference is about 38 microseconds per day, large enough to undermine accurate navigation if left untreated.

Why it's true

A GPS receiver estimates its location from precisely timed radio signals. Multiplying signal travel time by the speed of light gives a distance-like quantity called a pseudorange. Because radio waves travel so quickly, even a tiny timing mismatch corresponds to a substantial ranging discrepancy. Accurate clocks and a shared time model are therefore essential.

Two relativistic effects pull in opposite directions. A satellite's orbital motion makes its clock run slower relative to a comparable Earth-centered reference description, an effect associated with special relativity. Its higher altitude places it in a weaker gravitational field, making its clock run faster relative to clocks near Earth's surface, as described by general relativity.

For nominal GPS conditions, the motion contribution is roughly minus 7 microseconds per day, while the gravitational contribution is roughly plus 45. The combined result is about plus 38 microseconds daily. These rounded figures describe an important clock-rate correction, not the complete catalogue of timing and signal-propagation effects handled by the system.

By the numbers

~+45 µs/day

Nominal gravitational clock-rate contribution

~-7 µs/day

Nominal orbital-motion clock-rate contribution

~+38 µs/day

Approximate combined rate relative to ground clocks

How we know

Einstein published special relativity in 1905 and general relativity in 1915, long before satellite navigation became an engineering project. Atomic clocks later made their timing predictions experimentally testable. NIST describes comparisons involving airplanes, satellites and clocks at different elevations as evidence that motion and gravitational conditions affect measured clock rates.

GPS made those effects operationally important. Rather than assuming every good clock must agree everywhere, its designers incorporated a consistent relativistic timing framework. The predictable average orbital clock-rate offset is handled in the satellite clock design and frequency setting, while additional corrections address changing conditions during actual operation.

Orbits are not perfectly circular, so a satellite's speed and gravitational potential vary during each orbit. Receivers apply a broadcast-orbit-based relativistic correction for this periodic effect. Earth's rotation also matters when relating signal transmission and reception in the rotating terrestrial frame, commonly discussed through the Sagnac correction.

The familiar "GPS would fail within a day" statement captures the importance but overstates the precision of a deadline. An untreated rate offset would rapidly spoil accurate ranging, yet actual position errors depend on satellite geometry, receiver clock estimation and which corrections were removed. There is no universal instant when every receiver suddenly stops working.

What the famous distance estimate means

Multiplying 38 microseconds by roughly 300,000 kilometers per second gives about 11 kilometers. This illustrates the light-travel distance corresponding to that accumulated timing difference. It should not be presented as a guaranteed daily displacement of the location marker on every phone, because receiver solutions estimate several unknowns simultaneously.

A receiver commonly uses signals from at least four satellites to solve for three position coordinates and its own clock offset. That clock estimation does not make relativity optional: the satellite clocks, orbit calculations and transmission timestamps must still be mutually consistent. Relativity is part of the model that makes those measurements fit together.

Worth remembering

  • Satellite motion and weaker gravity affect clock rates in opposite directions.
  • The nominal net rate difference is about 38 microseconds per day.
  • An 11-kilometer light-travel equivalent is not a universal daily position error.

In context

GPS is a timing system used to infer position, not a satellite camera looking down at a receiver. Its relativity corrections belong alongside orbital information and atmospheric modeling. Accurate navigation depends on combining these pieces rather than crediting one correction with solving every source of positioning error.

A closer look

Why several signals are needed

A receiver compares timing information from different satellites to estimate its location and the offset of its own clock. That clock offset is an unknown in the solution, not a requirement that every phone contain an atomic clock. Consistent satellite time makes the comparison meaningful; an incorrect shared clock model would undermine the geometry used to determine position.

Different relativistic effects need different treatment

Satellite motion and gravitational potential contribute to clock-rate differences, while Earth's rotation affects signal timing in the rotating terrestrial reference frame. An orbit that is not perfectly circular also requires attention to changing conditions along the path. The familiar daily net rate offset is consequently an entry point to the subject, not the complete specification of a working navigation system.

Related facts

The cesium second explains the frequency standard behind precision timekeeping. The second is defined by a cesium-133 atomic transition.

The light-based meter connects timing measurements with distance. The meter is defined using the speed of light in a vacuum.

The kilogram’s definition shows another use of fundamental constants in measurement. The kilogram is no longer defined by a metal cylinder.

Orbital views of the Great Wall illustrate a very different consequence of observing from space. The Great Wall is difficult to distinguish from orbit with the naked eye.

Key terms

Clock bias
The offset of a clock’s displayed time from the system’s reference time.
Gravitational time dilation
A difference in clock rates associated with different gravitational potentials.

People also ask

Does a phone need its own atomic clock?

No. The navigation solution includes the receiver’s clock offset, using signals from multiple satellites to constrain position and time.

Is the correction only special relativity?

No. Both motion-related and gravitational effects matter, and their contributions do not have the same sign in nominal GPS orbits.

Does relativity explain every GPS error?

No. Atmospheric delays, satellite orbit uncertainty, reflections and receiver conditions also affect positioning.

Why not reset satellite clocks occasionally?

Navigation needs a consistent ongoing timing model. Predictable rate differences and changing orbital effects must be accounted for between clock adjustments.

Is the quoted daily offset universal for satellites?

No. It refers to nominal GPS orbital conditions; satellites on other trajectories experience different combinations of motion and gravitational potential.

Sources

Send this to someone

More like this

See all

Join the list.

Every now and then we'll send you a fact you've never heard before.

No spam. No set schedule. Unsubscribe anytime.