Satellite Imagery Explained: How We See Earth from Space

Every day, hundreds of Earth observation satellites photograph our planet in wavelengths from ultraviolet to microwave, generating petabytes of data that feed applications from weather forecasting to military intelligence, from precision agriculture to insurance underwriting. Satellite imagery has become so embedded in daily life — through Google Maps, weather apps, and news coverage of natural disasters — that we rarely consider the extraordinary technology required to capture, transmit, and process images from hundreds of kilometers above Earth's surface.

Here's how it actually works.

Types of Satellite Imagery

Optical (Visible and Near-Infrared)

The most intuitive type — cameras that capture reflected sunlight, producing images similar to what your eyes see. Modern optical satellites use pushbroom sensors — linear arrays of thousands of detectors that scan the Earth as the satellite moves in orbit, building an image line by line. Leading providers include:

• Maxar (WorldView Legion): Sub-30cm resolution — detailed enough to identify vehicle types and assess structural damage to individual buildings
• Planet Labs (SuperDove): 3-meter resolution but with daily global coverage from 200+ satellites, enabling change detection at unprecedented temporal frequency
• Airbus (Pléiades Neo): 30cm resolution with stereo imaging capability for 3D terrain modeling
• BlackSky: Sub-meter resolution with rapid revisit from a constellation of 55kg microsatellites

Limitation: Optical imagery requires sunlight and clear skies. Cloud cover, nighttime, smoke, and haze all degrade or prevent image capture.

Multispectral and Hyperspectral

Multispectral sensors capture images in multiple specific wavelength bands — typically 4-12 bands spanning visible through shortwave infrared. Each band reveals different information: red and near-infrared bands detect vegetation health (the basis of the NDVI index used in precision agriculture), shortwave infrared reveals moisture content, and thermal infrared measures surface temperature.

Hyperspectral sensors capture hundreds of narrow, contiguous spectral bands, creating a detailed spectral "fingerprint" for every pixel. This enables identification of specific minerals, crop species, water quality parameters, and chemical compounds — capabilities impossible with fewer bands. Satellites like PRISMA (Italian Space Agency) and EnMAP (German Aerospace Center) provide hyperspectral data.

Synthetic Aperture Radar (SAR)

SAR satellites transmit microwave pulses toward Earth and measure the reflected signal, effectively creating their own illumination. This gives SAR its killer advantage: it works day or night, through clouds, smoke, and rain. SAR imagery looks very different from optical — it appears grainy and gray-scale, with bright returns from buildings and ships and dark returns from smooth water surfaces.

Leading SAR providers include:

• Capella Space: Sub-50cm resolution X-band SAR — the highest-resolution commercial SAR available
• ICEYE: Sub-meter SAR with a constellation of microsatellites enabling rapid revisit
• Umbra: Open-data SAR provider with 16cm staring spotlight resolution

SAR is particularly valuable for maritime surveillance (detecting ships that have turned off their AIS transponders), disaster response (mapping flood extent through cloud cover), and defense intelligence (monitoring military facilities regardless of weather).

Thermal Infrared

Thermal sensors detect heat radiation emitted by Earth's surface, measuring surface temperature independent of sunlight. Applications include urban heat island mapping, wildfire detection and monitoring, volcanic activity surveillance, and industrial facility monitoring. Landsat and Sentinel-3 carry thermal bands; dedicated thermal missions like Satellogic's HyperSat are expanding commercial thermal capabilities.

Understanding Resolution

Satellite imagery resolution has four dimensions:

• Spatial resolution: The size of the smallest feature distinguishable in an image. A 30cm resolution means each pixel covers a 30cm x 30cm area on the ground. Military-grade systems achieve 10-15cm; commercial leaders reach 25-30cm
• Temporal resolution: How frequently a satellite revisits the same location. A single satellite may revisit every 3-5 days; a constellation like Planet's achieves daily global coverage
• Spectral resolution: The number and width of wavelength bands captured. Panchromatic (1 band), multispectral (4-12 bands), hyperspectral (100+ bands)
• Radiometric resolution: The sensitivity of the sensor to small differences in reflected energy — typically 8-bit (256 levels) to 16-bit (65,536 levels). Higher radiometric resolution enables detection of subtle variations in dark shadows or bright snow

There is an inherent trade-off between spatial and temporal resolution. You can have extremely detailed images of specific locations (Maxar's approach) or moderate-resolution images of everywhere every day (Planet's approach). The industry trend is toward constellations that break this trade-off — achieving both high resolution and high revisit through sheer numbers of satellites.

From Raw Data to Actionable Intelligence

A raw satellite image straight from the sensor is surprisingly useless for most applications. The processing pipeline includes:

1. Radiometric correction: Removing sensor artifacts, calibrating pixel values to physical units of radiance
2. Geometric correction: Orthorectification — removing distortions caused by satellite viewing angle, terrain relief, and Earth's curvature to produce a map-accurate image
3. Atmospheric correction: Removing the effects of atmospheric scattering and absorption to recover true surface reflectance values
4. Pan-sharpening: Combining high-resolution panchromatic data with lower-resolution multispectral data to produce a high-resolution color image
5. AI/ML analysis: Automated detection, classification, and change analysis using deep learning models trained on labeled satellite imagery. Modern platforms can automatically detect buildings, roads, vehicles, ships, aircraft, crop types, and land use changes

Key Applications

• Defense and intelligence: The original and still the largest market for satellite imagery. Activity monitoring, force disposition tracking, weapons proliferation detection, and battle damage assessment
• Agriculture: Crop health monitoring (NDVI), yield prediction, irrigation optimization, and insurance claims verification. Companies like Descartes Labs and Gro Intelligence build agricultural analytics on satellite imagery
• Insurance: Property damage assessment after natural disasters, risk modeling for underwriting, and fraud detection
• Finance: Satellite imagery-derived alternative data — counting cars in retail parking lots, measuring oil storage tank fill levels, tracking construction progress — provides alpha for quantitative investors
• Environmental monitoring: Deforestation tracking, carbon emissions estimation, methane leak detection, ocean color monitoring, and ice sheet measurement
• Urban planning: Population estimation, infrastructure monitoring, building footprint extraction, and transportation analysis

Track Earth Observation Satellites on SpaceNexus

SpaceNexus provides tracking and profiles for Earth observation satellites across all sensor types through our Satellite Tracker. Monitor constellation deployments, compare sensor capabilities, and follow the evolution of the remote sensing market.

Explore Satellite Tracking on SpaceNexus