← Tutorials

Tutorial Series · Part 12 of 27

GNSS Error Sources: Atmosphere, Blockage and Multipath

Explain real-world GNSS degradation including atmosphere, blockage, multipath and interference.

By

  • gnss
  • multipath
  • accuracy
Series Context

Navigation from First Principles

This article is part of an ordered course. Use the previous and next links to stay in sequence.

Progress Part 12 of 27
Previous Satellite Geometry and DOP Next Carrier Phase, RTK and Why Centimetres Are Possible

GNSS is often presented as clean geometry, but real measurements travel through the atmosphere, reflect from buildings and can be blocked or jammed. GPS.gov lists common degradation causes including signal blockage, indoor or underground use, reflected signals, interference, jamming and atmospheric conditions.1 Navipedia’s measurement model includes clock, troposphere, ionosphere, multipath and noise terms.2

Why clean geometry is not enough

A phone can work well in an open field and then jump across the road in an urban canyon. The receiver may still see satellites, but some signals may be reflected, weakened or non-line-of-sight.

A simplified error model

A simplified pseudorange error model is:

Pi=ρi+cδtr+Ti+Ii+Mi+ϵiP_i = \rho_i + c\delta t_r + T_i + I_i + M_i + \epsilon_i

where MiM_i is multipath or non-line-of-sight error. In an urban canyon, MiM_i can be biased and non-Gaussian. That makes it much more damaging than simple zero-mean noise.

For a teaching demo, use:

Mi={0,line of sightΔi,reflected or NLOSM_i = \begin{cases} 0, & \text{line of sight} \\ \Delta_i, & \text{reflected or NLOS} \end{cases}

where Δi\Delta_i is usually positive for an extra path length.

Interactive demo

The embedded simulator makes measurement quality visible. Direct rays, blocked satellites, and reflected satellites are rendered differently, then the resulting pseudorange set drives a simplified 2D solution.

Interactive Demo

Urban Canyon GNSS Simulator

This is a teaching simplification of a hard GNSS environment. Tall obstructions block satellites, reflected paths bias the ranges, and rejecting the worst signals can improve the solution.

Controls

Start with “Buildings”, then compare the solution with and without reflected satellites included.

Readouts

Usable satellites0
Multipath affected0
Horizontal error0.0 m
Solution qualitystable

What to watch

  • Open sky produces a small error with a stable solution.
  • Tall buildings remove some satellites entirely.
  • Reflected signals bias the estimate away from truth instead of just adding symmetric noise.
  • Excluding the worst reflected measurements can improve the result.

Try it

Start with open sky and low noise. Then enable buildings and multipath. Watch the solution jump even if several satellites remain visible. Then manually exclude the worst reflected satellite and see if the estimate improves.

Limits of this demo

A real GNSS receiver uses signal tracking, carrier-to-noise ratios, Doppler consistency, residual checks, multiple constellations, multiple frequencies, and sometimes 3D building models. This demo is intentionally simplified so the reader sees the main idea: not all satellite measurements are equally trustworthy.

Footnotes

  1. GPS.gov, “GPS Accuracy”. https://www.gps.gov/gps-accuracy - Official GPS accuracy FAQ covering satellite geometry, signal blockage, atmospheric conditions, receiver design, multipath, jamming and smartphone accuracy. Accessed 2026-06-11. ↩

  2. ESA Navipedia, “GNSS Measurements Modelling”. https://gssc.esa.int/navipedia/index.php/GNSS_Measurements_Modelling - Measurement equations including geometric range, clock terms, troposphere, ionosphere, multipath and noise. Accessed 2026-06-11. ↩