Open-source analysis of military flights linked to Ukraine can offer a useful way to separate narrative from evidence. By examining observable data from Flightradar24 and comparing patterns across Eastern Europe in November 2024 and November 2025, it becomes possible to identify shifts in visible activity without overstating what the data can prove. The purpose is not to make sweeping claims about strategy or intent, but to show how the intelligence cycle can be applied to open-source information in a disciplined and evidence-based way.
The most striking findings point to major year-on-year changes in observable military air activity. Germany recorded a 600 per cent increase in visible military flights in 2025 compared with 2024, while Romania showed an even larger rise of 660 per cent. At the same time, the United States registered a 55 per cent decline in observable air presence. These figures do not automatically reveal the reasons behind the change, but they do indicate that the visible pattern of military aviation connected to the wider Ukraine theatre shifted significantly over the period in question.
The geographical distribution of that activity changed as well. Romania, particularly around Bucharest and Constanța, rose from 6 per cent to 37 per cent of the total observable activity, while Poland increased from 30 per cent to 41 per cent. By contrast, the types of mission observed showed only limited variation. Transport flights rose slightly from 61 per cent to 64 per cent, while reconnaissance missions fell from 25 per cent to 18 per cent. The aircraft most frequently seen remained broadly the same, including the C-27 Spartan, Antonov An-30, C-130 Hercules, C-17 Globemaster and Airbus A400M.
The wider lesson is that OSINT is not simply a matter of searching online, but of collecting, assessing and interpreting open data with care. Observable flight information can help analysts detect operational changes, identify patterns and test public claims against measurable evidence. At the same time, methodological caution remains essential. ADS-B, MLAT and Mode S data can all be affected by coverage gaps and other biases, which means the analysis should be understood as a study of observable patterns rather than definitive causal attribution. Done properly, this kind of work can produce a more grounded and credible picture of developments in the region.
