Satellite Change Detection Analysis: Commercial Imagery for Ukraine Battlefield Monitoring

The Russia-Ukraine war has been characterized by an unprecedented level of commercial satellite imagery access, fundamentally changing the information environment for analysts, journalists, military planners, and the public. Never before has a major conventional conflict been so thoroughly observed from orbit at high temporal and spatial resolution by freely available or commercially accessible commercial platforms. The applications of this imagery for change detection — identifying what has changed between two observation periods — have produced analytically significant insights into Russian fortification construction, frontline movements, equipment distributions, and infrastructure damage.distributions, and infrastructure damage.

The Commercial Satellite Landscape

Three categories of commercial satellite provider have been most significant for Ukraine-focused change detection. First, optical high-resolution providers: Maxar Technologies (operating WorldView-1, -2, -3, and -4 satellites at 30–50 cm resolution) and Airbus Defence & Space (Pleiades constellation at 50 cm resolution) provide the highest-detail public imagery, capable of resolving individual vehicles, fortification structures, and shell craters. Maxar's policy of sharing Ukraine-relevant imagery with media and open-source analysts has made it the dominant visible satellite provider in public discourse about the war.

Second, medium-resolution high-frequency optical providers: Planet Labs operates the largest commercial Earth observation constellation (200+ PlanetScope satellites at 3–5 meter resolution), providing near-daily coverage of any point on Earth. This temporal density — a revisit rate of 1–2 days globally — enables change detection over time periods where single-day high-resolution images cannot be obtained due to cloud cover, but the 3–5m resolution makes it difficult to reliably identify individual vehicles or small fortification features.

Third, synthetic aperture radar (SAR) providers: Capella Space, Iceye, and Umbra operate X-band and other SAR constellations that image through cloud cover and at night, making them complementary to optical providers that cannot image through cloud or darkness. SAR imagery at 25 cm resolution (Capella's SpotLight mode) can resolve vehicle-sized objects with radar returns, providing cloud-independent monitoring capability particularly valuable for Ukraine's frequently overcast autumn and winter months.

Change Detection Algorithms and Methods

Change detection in satellite imagery is the computational identification of differences between registered co-located images from different timestamps. For battlefield monitoring, the principal categories of change detected are: ground surface disruption (new earthworks, trench construction, shell crater fields), structural changes (building damage, new fortification element construction), vegetation removal (cleared fields of fire, logistic area preparation), and moving platform tracking through multi-temporal comparison. Each change type has optimal detection approaches depending on the imagery type and resolution available.

For trench and earthwork detection — one of the most extensively applied change detection tasks in the Ukraine context — a common approach compares multitemporal medium-resolution optical or SAR backscatter data to identify linear soil disturbance signatures consistent with trench lines. Machine learning approaches, particularly convolutional neural networks trained on labeled trench imagery, have been developed by several commercial and academic groups to automate trench identification at scale, enabling systematic monitoring of hundreds of kilometers of frontline fortification development. The ACLED (Armed Conflict Location & Event Data) project and academic groups at University College London and elsewhere have published methodological analyses of this approach applied to Ukraine data.

Frontline Mapping Precision and Applications

The combination of high-resolution optical imagery, SAR, and OSINT geolocated ground-truth validation has enabled frontline mapping at a precision not previously achievable in real-time public analysis of a major conflict. Organizations including Maxar, the Institute for the Study of War, and commercial mapping services like Esri have published frontline delineations that independent military analysts have assessed at roughly ±0.5–2 km accuracy for the active contact line in accessible observation conditions — far more precise than the 5–20 km uncertainties typical of Cold War-era public battle assessments.

Limitations and Analytical Caveats

Commercial satellite imagery for frontline analysis has several important limitations that analysts must account for. Cloud cover over Ukraine is substantial for much of the year, frequently producing data gaps of 5–15 consecutive days in heavily overcast winter and autumn periods. High-resolution tasking is expensive and cannot cover the entire 1,000 km frontline continuously; most providers prioritize specific areas of active change or media/analyst interest, creating systematic coverage biases. SAR interpretation requires specialized expertise; without proper processing, SAR images can be misinterpreted, particularly in distinguishing between scattering artifacts and actual ground features.

Additionally, Russian active camouflage and concealment measures — including netting over equipment, movement at night, and use of decoy signatures — have been adapted in response to the known satellite observation schedule, creating cat-and-mouse dynamics between satellite collection and Russian concealment that degrade the reliability of imagery-based military capability assessments. The public commercial imagery record, while unprecedented in value, should always be treated as partial and potentially manipulated, not as a ground truth.

Frequently Asked Questions

Q: What resolution of satellite imagery is needed to identify individual vehicles?

A: Reliable individual vehicle identification typically requires optical resolution of approximately 50 cm or better (Maxar WorldView class) or SAR resolution of 25–50 cm. The 3–5 meter resolution of Planet Labs PlanetScope is insufficient for reliable individual vehicle identification but adequate for detecting vehicle concentrations and area-level activity changes.

Q: Why is SAR imagery useful for Ukraine battlefield monitoring?

A: Synthetic Aperture Radar (SAR) imaging uses radar pulses rather than optical light, allowing it to image through cloud cover and in darkness. For Ukraine — where cloud cover and winter conditions frequently deny optical coverage — SAR from Capella Space and ICEYE provides continuity of monitoring that optical satellites cannot provide alone.

Q: How accurate is satellite-based frontline mapping?

A: Commercial satellite-based frontline mapping by organizations like ISW and Maxar is assessed at roughly ±0.5–2 km accuracy for the active contact line under normal observation conditions, far superior to the 5–20 km uncertainty typical of Cold War-era public battle mapping.

Q: Can Russia hide military activity from commercial satellites?

A: Partially. Russia uses camouflage nets, nighttime movement, decoy signatures, and scheduling maneuvers to occur between known satellite overpasses. These measures degrade but do not eliminate detection — the density of the current commercial constellation, multi-type (optical + SAR) observation, and the challenge of hiding large-scale activity make sustained concealment increasingly difficult.

Q: What is the significance of Maxar's open data policy for Ukraine analysis?

A: Maxar's policy of sharing high-resolution Ukraine imagery with media and open-source analysts has fundamentally democratized access to detailed battlefield imagery. Previously, sub-meter resolution imagery was accessible only to government and large commercial buyers; Maxar's Ukraine sharing made it available to academic researchers, NGOs, and journalists, enabling an unprecedented level of independent analysis.

Sources

• Maxar Technologies, Ukraine Crisis Imagery program documentation and imagery releases (2022–2025)
• Planet Labs, Ukraine monitoring and analytical notes (2022–2025)
• Capella Space, SAR imagery technical specifications and Ukraine releases
• Strava, Kellogg & Brown Root (KBR), "SAR Change Detection Methodology" (technical)
• University College London, Open Mapping project for Ukraine (2023–2024)
• ACLED, conflict event mapping methodology documentation
• Bellingcat, "Using Satellite Imagery to Track Ukraine Conflict" (methodology guides)
• NASA SERVIR, change detection tutorial materials (applied to conflict monitoring)