Research Methodology
How Strategic Threat Index conducts historical geospatial change analysis
Platform Overview
Strategic Threat Index is a non-operational platform designed for academic and policy research into historical geospatial infrastructure change. It does not provide real-time surveillance, operational tracking, or actionable intelligence. All data is historical, delayed, and coarse-grained by design.
Gather public imagery, public datasets, source URLs, and archived site notes.
Normalize dates, locations, imagery vintage, and source metadata for review.
Compare change indicators and infrastructure patterns across delayed imagery.
Assign confidence and require corroboration before significant findings are logged.
Present non-operational research outputs with limitations and source caveats.
Historical OSINT Analysis Approach
What is historical OSINT analysis?
Open Source Intelligence (OSINT) analysis, as applied in this platform, refers primarily to the systematic review of archival, publicly available geospatial data. This includes historical satellite imagery, open-source map data, and publicly available infrastructure records. The Conflict Timeline is the separate exception: it uses a server-side GDELT article feed for recent news developments and shows source URLs returned by that feed.
Data sources
The platform draws on archived imagery from publicly available satellite programs including historical Sentinel-2, Landsat, MODIS, and ESRI World Imagery/Wayback where available. No classified imagery sources are accessed. Source labels indicate the archival program or public imagery service from which data was derived.
Site selection
Research sites are selected based on their relevance to academic and policy research on infrastructure development patterns. Sites are fixed, bounded polygons representing specific geographic areas. Centroid coordinates are coarse approximations only; precise coordinates are intentionally withheld to prevent misuse.
Confidence Scoring Methodology
How confidence is calculated
Confidence scores reflect a qualitative analyst assessment of the reliability of a given change detection or observation. Three levels are used: High (strong visual evidence across multiple acquisitions), Medium (single-acquisition evidence or partial cloud cover), and Low (inferred or uncertain change, significant data gaps). Confidence is not a statistical probability.
Change score computation
The site-level change score (expressed as a percentage, 0–100%) is a composite metric derived from the frequency, severity, and confidence of recorded change events across the archival review period. A score above 70% indicates a high volume or severity of observed structural change in the archive. Scores are normalised against a 36-month baseline.
Automated vs. analyst review
Initial change records in this build are source-labelled archival entries seeded from public reporting and analyst review. In a production deployment, this layer would be refreshed by a semi-automated change detection algorithm with mandatory analyst review and source confirmation before any change event is logged.
Limitations and Uncertainty
Resolution constraints
All imagery used in this platform is coarse to medium resolution. High-resolution sub-metre imagery is not used. This means that structural identification (what a building or object is) is not reliably possible from the data available. Observations describe general spatial change, not specific asset identification.
Temporal gaps
Archival imagery availability is irregular and subject to cloud cover, satellite revisit cycles, and archive completeness. Gaps in the temporal record may result in missed events or misattributed change dates. Analysts note temporal gaps where known.
Interpretation bias
Human analyst review introduces inherent interpretation bias. Findings represent one analyst's assessment of ambiguous visual data. Independent review and corroboration from additional sources should always be sought before any research conclusion is drawn.
No ground truth
All findings are derived solely from aerial or satellite imagery. No field verification, human intelligence, or signals intelligence is incorporated. All observations should be treated as preliminary and unverified.
Ethical Safeguards
Non-operational design principle
Strategic Threat Index is designed from the ground up to be non-operational. No feature of this platform supports or enables real-time tracking, live military surveillance, strike planning, or operational targeting of any kind. The platform is intended solely for academic research and policy analysis.
Delay model
A mandatory minimum 30-day delay is applied to the geospatial imagery workflow and monitored-site assessments. The recent Conflict Timeline is article-level news monitoring, not operational tracking; its map coordinates are country or region centroids when exact coordinates are not supplied by the external source.
Coordinate coarsening
Precise coordinates are never displayed in the platform. All geographic references are bounded to site polygons and coarse centroid estimates. This prevents the platform from being used to produce precise targeting data.
Responsible use commitment
Users of this platform are expected to comply with all applicable laws, regulations, and ethical guidelines governing open-source research. The platform may not be used to facilitate harm to individuals, to conduct surveillance of persons, or for any military operational purpose. Research outputs derived from this platform should be treated as preliminary and subject to independent verification.