GPS jamming in the Russo-Ukrainian war has opened a growing gap in satellite atmosphere data

Why it matters

Radio occultation is a key source of long-term atmospheric temperature records and weather forecast input. If the gap grows over time, it can bias trends in a way that simple corrections do not fix. The authors say climate data records based on this technique need a critical review.

What they did

The authors counted radio occultation profiles from the Metop-B satellite over time and compared recent months with 2013 to 2017. They then took the years before 2018 and randomly removed profiles to match today’s reduced coverage. The difference between full and thinned data estimates the error from missing observations. They used Monte Carlo simulations to apply this to the 2018 to 2026 record and estimate the effect on temperature trends.

Key findings

• Over eastern Ukraine and western Russia, observation density fell by more than 90% from early 2018 to 2025. A central zone of around 15°×15° is now effectively unobserved.
• Coverage over the wider European region fell by about 40%. The area of reduced data reaches from the Iberian Peninsula to northeastern South Asia. Southern Hemisphere mid-latitudes show no significant change.
• In the thinned-data test, monthly mean temperature errors reached up to 4 K in the free atmosphere near the gap. Zonal-mean errors reached up to 0.35 K at 50°N to 60°N.
• The missing data cause a warm bias that grows over time and varies by year, so a simple bias correction is not enough. It leads to trend errors that are only a factor of 2 to 3 smaller than the actual observed trends.
• The authors estimate that zonal-mean trends from 2018 onward are likely underestimated at most altitudes, by up to 60% around 25 km.

Limitations

• The error estimates assume that errors found by thinning pre-2018 data match the real errors after 2018. They also assume the loss grew linearly from 2018 to 2025.
• The study is a preprint and has not been peer reviewed. The trend analysis focuses on the 50°N to 55°N band.
• Results depend on how zonal means are computed. A simple average of all profiles in a latitude band gives much larger errors.

Glossary

• GNSS Radio Occultation: A method that uses satellite navigation signals passing through the atmosphere to a satellite in low orbit, to derive temperature and pressure profiles.
• GPS jamming: Deliberate interference that drowns out navigation satellite signals.
• Zonal mean: An average of a quantity around a full latitude band.
• RMSE: Root-mean-square error, a measure of the typical size of errors.