Deforestation alerts catch the obvious: bare ground where forest stood. Degradation is the silent precursor and the harder problem. Selective logging, understory burning, canopy thinning from drought stress, and fragmentation from new tracks each reduce biomass and biodiversity without clearing a single hectare. Conventional optical satellites miss most of this because the canopy closes over the wound within weeks; radar and shortwave-infrared are required to see beneath and through it. Nations relying on third-party alert services receive only the coarse signal and systematically under-report their actual carbon emissions to the UNFCCC.
A sovereign constellation combining L-band or C-band SAR with shortwave-infrared (SWIR) optical instruments resolves the problem at national scale. SAR coherence change detection identifies canopy structural disruption down to single-tree removal; SWIR distinguishes live green canopy from stressed or burned material that looks healthy in visible bands. Flown together at 12–16 day repeat cycles, the two payloads produce a spatially explicit degradation severity index updated monthly, at 10–25 m ground resolution across the entire national forest estate.
The operational outcome is threefold. Forest agencies can enforce concession boundaries against selective-logging violations before the damage compounds. Environment ministries submit verified Tier-2 emissions inventories rather than activity-data proxies, unlocking REDD+ carbon credits and bilateral climate finance. And national carbon registries gain an independent, court-admissible evidence base that is not subject to commercial licence restrictions or foreign government access controls.
Frequently asked
What is the practical difference between deforestation and forest degradation for satellite monitoring purposes?
Deforestation is a discrete, high-contrast event — complete canopy removal detectable even with moderate-resolution imagery (30 m Landsat). Degradation is cumulative and low-contrast: selective logging, fuelwood cutting, and repeated surface fire reduce biomass incrementally without triggering the spectral break that deforestation detectors look for. Mapping it reliably requires sub-10 m resolution, dense time-series compositing, and ideally multi-sensor fusion (optical plus SAR plus lidar where available).
Why should a sovereign nation operate its own forest-degradation satellite constellation rather than buying analytics from Planet or ICEYE?
Commercial providers can terminate contracts, raise prices, or be subject to foreign export-control restrictions. A sovereign constellation means the raw data never leaves national custody, the processing pipeline can be audited, and REDD+ and EUDR submissions carry the legal weight of nationally certified statistics. It also allows the nation to task satellites on its own schedule rather than competing with commercial customers for archive priority. Over a 15-year horizon, owned infrastructure is generally cheaper than sustained data-purchase contracts for any nation monitoring more than roughly 50 million ha.
What orbit and sensor type are best for degradation mapping?
A LEO constellation at 450–550 km altitude using multispectral imagers (10 m, eight-plus bands including SWIR) combined with C- or X-band SAR provides the cloud-penetration and spectral depth needed. A constellation of 8–12 microsatellites in sun-synchronous orbits can achieve 3–5 day revisit over any tropical target. SAR-only constellations (like ICEYE's 21-satellite fleet) can push to sub-24 h revisit but at a higher per-satellite cost and with more demanding data-fusion pipelines.
How does forest degradation mapping feed into REDD+ reporting?
UNFCCC REDD+ requires nations to report emissions and removals from all five activities, including forest degradation (activity 'D'). Satellite-derived Activity Data (AD) combined with Emission Factors (EF) from national forest inventories constitutes the standard Tier 2/Tier 3 reporting approach under the 2006 IPCC Guidelines. Nations with sovereign satellite capacity can update their forest reference levels annually rather than every five years, improving the statistical quality of their submissions and their eligibility for results-based payments under the Green Climate Fund.
Does the EU Deforestation Regulation (EUDR) require satellite evidence specifically?
EUDR Regulation 2023/1115 does not mandate a specific technology, but it requires operators to demonstrate due diligence — geolocation, date of harvest, and 'no deforestation or degradation after 31 December 2020'. Satellite-derived land-cover maps with documented chain of custody are the most scalable and auditable evidence base available. Nations that can provide certified, government-issued satellite evidence of forest status have a competitive advantage in accessing EU commodity markets.
Can AI / machine-learning models replace manual expert interpretation for degradation mapping?
Deep-learning models (U-Net variants, vision transformers) now achieve 85–92% overall accuracy on degradation classification in benchmark tropical datasets, but they remain prone to systematic error when applied outside their training domain — a forest type or sensor configuration not in the training set. Nations should treat AI outputs as a first-pass screening layer that still requires structured validation sampling. Sovereign control of the model weights and training data is also critical; relying on a foreign vendor's black-box model introduces the same dependency risks as relying on foreign raw imagery.
What is the minimum constellation size for operationally useful degradation monitoring over a large tropical nation?
For a country with around 100 million ha of forest (comparable to the DRC or Indonesia), a constellation of 6 optical microsatellites plus 4 SAR nanosatellites can achieve 5-day cloud-free composites over 95% of the territory in most seasons. This is sufficient for near-real-time degradation alerts at 0.5 ha sensitivity. Dropping below 4 optical satellites pushes average revisit past 10 days, which misses fast-moving selective-logging fronts that operators clear in 3–7 days.
How is forest degradation mapping different from — and complementary to — forest biomass estimation?
Degradation mapping tracks change: it tells you where and when biomass is being lost and at what rate. Biomass estimation provides the baseline stock against which that loss is measured. You need both: degradation maps without a biomass baseline cannot be converted to carbon accounting figures, while a biomass map without change detection is a static snapshot that ages out of relevance within 2–3 years in active forest frontiers. The two capabilities share sensor infrastructure but require different algorithm pipelines.