How Satellite Intelligence Works: Seeing Everything from Space
Types of Satellite Intelligence
Satellite-based intelligence falls into three main branches:
| Type | Abbreviation | What It Collects |
|---|---|---|
| Imagery Intelligence | IMINT | Optical and radar images |
| Signals Intelligence | SIGINT | Electromagnetic emissions, communications |
| Measurement & Signature Int. | MASINT | Radar, nuclear, chemical signatures |
This article focuses primarily on IMINT and SIGINT.
How Imagery Intelligence (IMINT) Works
A) Electro-Optical (EO) Satellites
Basic principle: high-resolution photography from space. Works in daylight conditions with sunlight.
Resolution: Commercial satellites (GeoEye-1, WorldView-3) offer approximately 30–50 cm resolution — enough to distinguish the make of a vehicle. Classified military satellites (KH-11 class) are estimated to achieve significantly higher resolution.
Limitation: Clouds, night, smoke block imagery. Solution: SAR satellites.
B) SAR (Synthetic Aperture Radar) Satellites
Radar-based imaging: transmits microwaves to the ground, receives reflections, and constructs an image. Works at night and through cloud cover.
How SAR works:
1. Satellite emits a radar signal (typically C, X, or L-band)
2. Signal strikes the ground and returns — delay and Doppler shift encode range and motion
3. Serial measurements collected as the satellite moves forward are synthesized to create a large virtual antenna aperture (the “S” in SAR)
4. Processing: high-resolution 2D or 3D map generated from radar returns
SAR satellite examples: Sentinel-1 (Europe), TerraSAR-X (Germany), COSMO-SkyMed (Italy), ICEYE (Finland — commercial)
SAR signatures:
- Metal surfaces (tanks, ships) produce strong backscatter — “bright” spots
- Earthwork and trench digging appears as texture changes on the ground
- Vehicle motion detection: InSAR and change detection techniques
C) Hyperspectral and Multispectral Imaging
A normal camera captures visible light. Multispectral sensors add near-infrared and other wavelengths. Hyperspectral sensors record hundreds of bands simultaneously.
This enables:
- Distinguishing camouflaged vehicles from vegetation
- Detecting chemical activity or nuclear infrastructure (heat signature, material spectrum)
- Underground facility or tunnel entrance/exit activity
How Signals Intelligence (SIGINT) Works
SIGINT satellites collect electromagnetic emissions rising from the Earth’s surface:
| SIGINT Type | What It Monitors |
|---|---|
| COMINT (Communications) | Radio, telephone, encrypted messages |
| ELINT (Electronic) | Radar emissions, IFF signals |
| FISINT (Foreign Instrumentation) | Weapons test telemetry, missile parameters |
Technical Details:
Geostationary (GEO) satellites — approximately 36,000 km altitude — can continuously monitor a fixed region; ideal for detecting communications spread over large areas.
Low Earth Orbit (LEO) satellites — 200–2,000 km — closer to Earth’s surface to capture even weak signals; suited for precision facility monitoring or capturing short-duration emissions.
Major SIGINT Programs:
- ECHELON / UKUSA: Global communications monitoring network operated jointly by the US, UK, Canada, Australia, and New Zealand (Five Eyes); includes satellite capacity
- NROL programs (USA): National Reconnaissance Office classified satellite series
Orbit Types and Intelligence Use
| Orbit | Altitude | Characteristic | Use |
|---|---|---|---|
| LEO | 160–2,000 km | High resolution, short pass time | IMINT, SAR, SIGINT |
| MEO | 2,000–35,786 km | Medium coverage | GPS, navigation |
| GEO | 35,786 km | Fixed region, 24/7 | Communications monitoring, early warning |
| HEO (Molniya) | Elliptical orbit | Polar and high-latitude coverage | Russia’s early warning systems |
Image Interpretation: AI and Analysts
Raw satellite imagery is not intelligence — it’s raw data. Value comes from interpretation.
Traditional approach:
Trained Photo Interpretation Analysts (PIAs) examine images manually:
- Shadow analysis to calculate building heights
- Vehicle type identification via wheels/tracks/size/shape
- Change detection: comparison against prior imagery
AI-assisted analysis (today):
- Object detection: Automatically identifies tanks, missile launchers, aircraft
- Change detection: Two-date imagery of a location is compared; tunnel digging, new construction, vehicle buildup is auto-flagged
- Classification: Vehicle make/model estimation (commercial AI models advancing rapidly here)
Real-World Intelligence Operations
Corona Program (1960–1972, USA)
The first satellite imagery intelligence program. Film canisters were ejected from orbit into the atmosphere and caught by aircraft mid-air. Best resolution ~1.8 meters. Mapped Soviet nuclear infrastructure throughout the Cold War.
KH-11 Kennan (1976+, USA)
Replaced film canisters with real-time digital image transmission. Still active (upgraded versions). In 2011 it emerged the KH-11 shares its structure with the Hubble telescope — the NSA had donated two Hubble-sized satellites to NASA.
Commercial Satellites in the Ukraine Crisis (2022)
Commercial companies like Maxar and Planet Labs shared imagery documenting the Russian buildup along Ukraine’s northern border with the public. Russian military disposition was documented in the commercial press before the White House confirmed it — a watershed moment for the democratization of IMINT.
North Korea Nuclear Facilities
The Yongbyon nuclear complex and Punggye-ri test site are regularly monitored by international analysts. Analysis of every activity — tunnel digging, cooling water flow, truck movements — via satellite imagery has provided early warning ahead of tests.
Major Satellite Intelligence Assets
| System | Country | Type | Feature |
|---|---|---|---|
| KH-11 class | USA | Optical | Classified; estimated sub-30 cm resolution |
| NROL-42 / NROL series | USA | SIGINT | Classified; GEO SIGINT |
| Pleiades Neo | France | Optical | 30 cm; commercial export |
| WorldView Legion | USA (Maxar) | Optical | 30 cm commercial |
| COSMO-SkyMed Gen 2 | Italy | SAR | All-weather |
| Yaogan series | China | Optical + SAR + SIGINT | Classified; rapid expansion |
| Kobalt series | Russia | Optical | Renewed; revisit rate questions |
| ICEYE | Finland | SAR | Commercial; sold data to Ukraine |
Frequently Asked Questions
Are satellite images real-time?
Commercial satellites typically pass over a given area once to several times per hour; images are therefore not live but a snapshot at a specific moment. Large LEO constellations (Planet Labs, 200+ satellites) provide daily revisit rates. GEO satellites offer a fixed stare but at very low resolution.
How deep underground can satellites see?
Optical satellites can see surface and open-area structures. Underground facilities cannot generally be directly imaged; however, entrance/exit traffic, waste heat signatures, and soil disturbance anomalies are analytically interpreted.
Is commercial satellite data reliable?
Yes, largely reliable for image quality. Interpretation errors arise in the human analysis phase. Commercial companies can make policy decisions about whom they sell data to (e.g., declining to sell to certain countries or for certain purposes).
Sources
- National Reconnaissance Office (NRO) — Declassified Program Histories
- Federation of American Scientists (FAS) — Space Reconnaissance Database
- CSIS — Aerospace Security Project, Space Threat Assessment 2024
- John Pike, GlobalSecurity.org — Satellite intelligence classifications
- Maxar Technologies — Imagery Products Overview
- IISS, The Military Balance 2024
- David Burbach, Naval War College — Commercial Satellite Intelligence
