Free vs. Commercial Satellite Imagery: What You Actually Get for the Price Difference

The Price Gap Is Enormous

Free satellite imagery — primarily from ESA's Copernicus programme (Sentinel-1, Sentinel-2) and USGS/NASA (Landsat) — costs exactly zero. You can access decades of global coverage without paying a cent.

Commercial very-high-resolution (VHR) imagery, by contrast, is sold by the square kilometre, with a minimum order area, and tasking (requesting a new acquisition rather than pulling one from the archive) costs substantially more than archive. Published rates differ by provider, resolution, licence terms, and volume commitment, and they change — check the vendor's current price list rather than any figure quoted in an article.

The gap between "free" and "priced per square kilometre" is what actually shapes project design. The question is whether it buys you something you need.

What Free Imagery Provides

Sentinel-2 (Optical)

• Resolution: 10m (visible + NIR), 20m (red edge + SWIR), 60m (atmospheric)
• Revisit: 5 days (at the equator, more frequent at higher latitudes)
• Bands: 13 spectral bands from visible through SWIR
• Coverage: Global land surfaces
• Archive: 2015-present

At 10m resolution, one pixel covers a 10×10 meter area. You can distinguish:

• Agricultural field boundaries
• Forest clearings larger than ~0.01 hectare
• Urban areas vs. rural areas
• Water bodies and coastlines
• Major roads and highways

You cannot distinguish:

• Individual buildings (unless they're very large)
• Vehicle types
• Small paths or trails
• Crop rows within a field

Sentinel-1 (SAR)

• Resolution: ~20m (IW mode), ~5m (SM mode)
• Revisit: 6 days
• Bands: C-band (5.405 GHz), dual-pol VV+VH
• Coverage: Global (systematic over land and coastal areas)
• Archive: 2014-present

Landsat (Optical)

• Resolution: 30m (multispectral), 15m (panchromatic)
• Revisit: 16 days (8 days with two satellites)
• Bands: 11 bands
• Coverage: Global
• Archive: 1972-present (the longest continuous Earth observation record)

The Landsat archive's length is its unique strength. No commercial provider offers 50+ years of consistent global coverage.

What Commercial Imagery Provides

Sub-Meter Optical

• Resolution: 0.3-0.5m (panchromatic), 1-2m (multispectral)
• Revisit: Daily or near-daily (from constellation operators)
• Bands: Typically 4-8 bands
• Archive: Varies by provider

At 0.3m resolution, one pixel covers a 30×30 cm area. You can identify:

• Individual vehicles and their approximate type
• Building footprints and roof materials
• Tennis courts, swimming pools, and playground equipment
• Individual trees in an urban setting
• Small boats at a marina

Commercial SAR

• Resolution: 0.25-3m
• Revisit: Daily or near-daily (constellation operators)
• Modes: Various polarizations, spotlight/stripmap/scanSAR

Commercial SAR typically offers higher resolution than Sentinel-1, enabling detection of smaller objects and finer structural details.

When Free Is Enough

For a surprisingly large number of professional applications, free imagery is not just adequate — it's actually preferred:

Agriculture (Field-Level)

Monitoring crop health across a farm or region with NDVI works well at 10m resolution. Each field typically contains hundreds to thousands of Sentinel-2 pixels, providing statistically robust index values. The 5-day revisit captures the full growing season at a frequency that matches relevant biological timescales.

Deforestation Monitoring

Detecting forest loss at the scale that matters for national reporting and carbon accounting works at 10-30m. A cleared area large enough to be ecologically significant (>0.1 hectare) is resolvable. Sentinel-1 SAR adds cloud-free monitoring capability in tropical regions.

Flood Mapping

Sentinel-1 SAR at 20m resolution reliably delineates flood extent for emergency response. The difference between flooded and non-flooded areas produces a strong backscatter contrast that doesn't require sub-meter resolution to detect.

Urban Expansion Tracking

Distinguishing urban from non-urban land cover at 10m resolution is straightforward. Multi-year time series reveal growth patterns clearly.

Water Quality Monitoring

Algal blooms, sediment plumes, and water clarity indicators work at 10-60m scales. The phenomena themselves span hundreds of meters to kilometers.

When You Need Commercial

Infrastructure Monitoring

Assessing the condition of individual bridges, dams, or pipelines requires resolving structural details that are invisible at 10m. A bridge might be 10-20m wide — that's 1-2 pixels in Sentinel-2, which tells you the bridge exists but nothing about its condition.

Damage Assessment (Building-Level)

After an earthquake or conflict, determining which specific buildings are damaged vs. intact requires sub-meter resolution. At 10m, you can see that a neighborhood was affected, but not which individual structures were destroyed.

Precise Change Detection

If you need to detect a new structure, a removed tree, or a filled excavation, you need resolution that matches the size of the change. A 5m-wide construction site is half a pixel in Sentinel-2.

Defense and Intelligence

Identifying equipment types, counting vehicles, and monitoring specific facilities are inherently high-resolution tasks.

The Hybrid Approach

The most cost-effective strategy for many organizations:

1. Use free imagery for wide-area screening — Sentinel-2 and Sentinel-1 cover the entire Earth every few days. Use this for ongoing monitoring, anomaly detection, and identifying areas of interest.

2. Task commercial imagery only where needed — When free data flags something that requires closer inspection, order a high-resolution scene of that specific location. This limits commercial imagery costs to the small fraction of your study area that actually requires it.

3. Use free data for time-series context — Even when you have a commercial image for a specific date, the free archive provides temporal context (what did this area look like last month? Last year?) at no additional cost.

Because commercial imagery is priced by area, restricting it to the small fraction of your study area that free screening has already flagged is what makes the cost difference — not negotiating a better rate.

Resolution, Cost, and Application: Consolidated Reference

Consolidating the capability trade-offs into one table for practical decision-making:

Tier	Sensor Examples	Resolution	Cost model	Key Use Cases	Clear Limitation
Free Optical	Sentinel-2, Landsat	10–30m	Free global archive; no tasking	Agriculture, land cover, forest monitoring	Cannot resolve individual buildings
Free SAR	Sentinel-1	10–20m	Free global archive; no tasking	Flood, subsidence, ship detection	Lower resolution than commercial SAR
Commercial mid-res	Planet SuperDove	3–5m	Area subscription	Construction monitoring, small-scale change	Limited spectral bands (4–8)
Commercial high-res optical	Maxar WorldView, Airbus Pléiades	0.3–0.5m	Per km², archive cheaper than tasking	Damage assessment, vehicle detection	Highest cost per km²; tasking constrained by revisit and weather
Commercial SAR	Capella Space, ICEYE	0.25–3m	Per km² or per scene	Maritime, rapid revisit through clouds	Complex processing; high cost
Nighttime/Thermal	VIIRS, Landsat TIRS, ECOSTRESS	375m–100m	Free	Urban energy, fire, ET mapping	Coarse resolution; ECOSTRESS ISS-limited

The actual cost of "free" data: Sentinel-2 and Landsat are free to download, but processing, storage, and analysis are not. A full-year, full-resolution time series over a region of continental scale runs to terabytes of raw data and non-trivial compute. Cloud platforms such as Google Earth Engine absorb much of that cost but introduce platform dependency. When comparing against a commercial offer that bundles processing, price your own storage and compute into the comparison rather than treating the data licence as the whole cost.

Where the hybrid approach delivers the best value: in monitoring programmes over large areas where only a small share of the area changes in any given period. Free imagery answers where and when something changed across the whole footprint; commercial imagery is then bought only for those locations and dates. The saving scales with how selective the screening step is — the smaller the flagged fraction, the larger the gap between blanket VHR coverage and a screened approach.

Resolution Is Not Everything

A common misconception is that higher resolution is always better. In practice:

• More pixels means more storage and processing — A 30cm scene is ~1,000× larger than a 10m scene of the same area
• Spectral coverage matters — Commercial imagery typically offers fewer spectral bands than Sentinel-2, limiting index calculations
• Temporal frequency matters — Daily commercial revisit sounds impressive, but off-nadir viewing angles at the edges of the swath introduce geometric distortions. Consistent nadir viewing from Sentinel is often analytically cleaner
• Archive depth matters — For trend analysis, Landsat's 50-year record is irreplaceable regardless of resolution

Start with Free

If you're beginning a new satellite monitoring project, start with free data. Many regional-scale remote sensing questions can be answered at 10–30 m resolution, although the resolution actually required depends on the target and the decision threshold. Only invest in commercial imagery once you've confirmed that your specific question genuinely requires it.

Off-Nadir Delta provides access to Sentinel-1 SAR, Sentinel-2 optical, and VIIRS nighttime lights data — all free and open. Explore these first, and you may find they're all you need.