EO·Atlas
Optical time-series change detection
Flags surface change by testing each new optical observation against a per-pixel statistical baseline built from the historical time-series, accumulating evidence across dates before confirming a change. Underlies near-real-time deforestation and disturbance alert systems.
Optical time-series change detection answers the question: has this surface location changed, and when? Rather than comparing two single-date images, the method builds a per-pixel statistical baseline from the historical time-series of optical reflectance, then tests each new observation against that baseline. Change is not flagged from a single anomalous scene but confirmed only when evidence accumulates across multiple dates, reducing false alarms from cloud contamination and sensor noise.
Two algorithm variants have been deployed in near-real-time deforestation and disturbance alert systems. The GLAD-L alert system (Hansen et al. 2016) applies bagged decision trees to Landsat red, near-infrared, and shortwave-infrared temporal metrics, confirming a disturbance when at least 2 of 4 consecutive observations within a 180-day window cross the change threshold.[1] The GLAD-S2 extension applies decision trees to Sentinel-2 bands 2, 3, 4, 8, 11, and 12 with a 4-tier confidence scheme (single, low, medium, high).[2] The RADD alert system (Reiche et al. 2021) takes a different approach: it uses Bayesian conditional probability of forest disturbance, fitting a per-pixel Gaussian mixture model as the temporal baseline and reporting low-confidence detections above 85% and high-confidence detections above 97.5% probability on Sentinel-1 C-band 10m imagery.[3]
The method applies to any surface monitored by a dense optical or SAR time-series and is not limited to forest. Known failure modes include persistent cloud cover that prevents baseline accumulation, seasonal phenological cycles misread as change, and rapid re-vegetation that obscures confirmed disturbances within the confirmation window.
- BlackShore Satellite Imagery Forensics
- Copernicus EMS On Demand Mapping
- Starling
Starling detects deforestation via monthly optical time-series change detection over Airbus constellation imagery (Pleiades Neo, SPOT, 10m-30cm range)
- Readar Change Detection
Readar change detection service applies optical time-series change detection over nationwide aerial imagery to flag property mutations for BAG/BGT registry maintenance.
- AI models & automation
Reef Support AI change detection tracks ecosystem change between survey epochs (coral bleaching events, seagrass loss, mangrove degradation) using time-series satellite and survey imagery.
- SpaceKnow China Satellite Manufacturing Index
China Satellite Manufacturing Index uses optical satellite imagery and proprietary AI algorithms to detect changes at 6000+ industrial facilities in China; activity patterns aggregated into a PMI-like index published 3x weekly; Bloomberg Terminal distributed; optical change detection is the primary sensing methodology
- Marmoris Coastal Ecosystem Insights
- Terraprisma ESG Monitoring and Compliance
- OpenAtlas TERRAIN
- OpenAtlas RECON
- Class Location Monitoring
- COSMIC-EYE Area Monitoring
- COSMIC-EYE Infrastructure Monitoring
- CBERS-4A Wide Swath Panchromatic and Multispectral Camera
via CBERS-4A
wpm-cbers-4a contributes CBERS-4A optical land-imaging observations suitable for time-series change workflows.
- CBERS-4A Multispectral CCD Camera
via CBERS-4A
mux-cbers-4a contributes CBERS-4A optical land-imaging observations suitable for time-series change workflows.
- CBERS-4A Wide Field Imager
via CBERS-4A
wfi-cbers-4a contributes CBERS-4A optical land-imaging observations suitable for time-series change workflows.
- Wide View CCD camera - 2
via Huan Jing-2A
HJ-2A WVC-2 wide-swath VIS/NIR imaging supports recurring land/disaster monitoring.
- Panchromatic and Multi-spectral CCD Camera
via Gaofen-1 (GF-1)
GF-1 high-resolution land observation/disaster-monitoring PMS imagery supports optical land-change monitoring.
- Wide Field Camera
via Gaofen-1 (GF-1)
WFV wide 800 km swath and 3-day coverage support recurring land-cover and disaster monitoring.
- MSS (KANOPUS)
via Kanopus-V
MSS repeat VIS/NIR land imaging supports optical land-cover change workflows.
- PSS
via Kanopus-V
PSS is a high-resolution panchromatic land imager on Kanopus-V spacecraft.
- MSU-200
via Kanopus-V
MSU-200 provides wide-swath panchromatic imaging for land and fire-boundary observation.
- CBERS-4 Multispectral Camera
via CBERS-4
MUXCAM contributes multispectral land observations for optical time-series change workflows.
- CBERS-4 Wide Field Imager 2
via CBERS-4
WFI-2 contributes wide-field optical time-series observations.
- CBERS-4 Panchromatic and Multispectral Imager
via CBERS-4
PANMUX contributes panchromatic and multispectral land observations.
- Jilin Imager
via Jilin-1
Jilin-1 provides high-resolution optical imagery for repeated land-observation and change-monitoring use cases.
- NewSat multispectral camera
via Satellogic NewSat constellation
NewSat multispectral camera is the optical payload used by the Satellogic NewSat constellation for repeat RGB/NIR imaging.
- PMS-3 high-resolution optical sensor
via SuperView / GaoJing constellation
PMS-3 is the SuperView optical payload supporting repeat panchromatic and multispectral imaging.
- PMS-2 Panchromatic and Multi-spectral CCD Camera
via Gaofen-2
GF-2 high-resolution repeated optical coverage supports optical change-detection workflows.
- AEISS-A
via KOMPSAT-3A
KOMPSAT-3A AEISS-A 0.55m PAN and 2.2m MS imagery used for urban and land-cover change detection; commercial imagery product.
- SpaceView-24
via BlackSky constellation
USGS characterizes BlackSky Global satellites using the SpaceView-24 multispectral sensor for high-resolution repeat optical imaging; WMO lists land observation and disaster monitoring objectives.
- 2C Degrees Water-Risk Signals
- CARMA Climate Risk Platform
- Cerberus
- Carble Remote Sensing for Coffee and Cocoa Supply Chains
- Readar Aerial Imagery Data Extraction
Readar change detection service demonstrates optical time-series change detection over aerial imagery for property mutation signalling.
- Remondo Imagery Data-as-a-Servicepending review
Remondo DaaS persistent monitoring mode is designed for optical time-series change detection applications; capability pending first satellite launch (mid-2027).
- OpenAtlas RANGER
- CS-101 Kikas
CS-101 Kikas is a visible-spectrum RGB camera; optical change detection capability applies to spacecraft monitoring and SSA use cases
- CS-202 Tsillopistri
CS-202 Tsillopistri is a visible-spectrum 12 MP camera with 4K at 30fps; capable of optical time-series imaging
- CS-252 Suupistri
CS-252 Suupistri is a 12 MP 70fps camera qualified to ESA and NASA standards; capable of optical time-series imaging with edge-compute processing
- CS-292 Suupistri-EO
CS-292 Suupistri-EO 67 MP camera with GSD <1m at orbit; designed for EO; optical change detection is a primary use case
- PAIS Optical Payloadpending review
- Valk VHRpending review
- Coastal Zone Imager
CZI is a coastal optical imaging payload on operational HY-1D; detailed use-case evidence should be strengthened in full sensor research.
- AWiFS
AWiFS wide-swath repeat optical imaging supports change-detection workflows.
- LISS-III (Resourcesat)
LISS-III repeat optical land imaging supports change-detection workflows.
- LISS-IV
LISS-IV repeat high-resolution optical land imaging supports change-detection workflows.
- Pléiades HiRI
Tasked high-resolution optical observations can contribute to optical time-series change detection where repeat acquisitions are available.
- MSC (KOMPSAT-2)
KOMPSAT-2 archive imagery can support optical change analysis across repeated scenes, but current new tasking is not available.
- PAN (Cartosat-3)
Very-high-resolution panchromatic optical imaging supports optical time-series change detection where repeat acquisitions are available.
- MX (Cartosat-3)
Four-band VNIR optical imaging supports land-cover and infrastructure change detection when repeat acquisitions are available.
- Pléiades Neo Imager
Pléiades Neo offers 30 cm optical tasking and daily revisit for site monitoring and change detection.
- AEISS
KOMPSAT-3 AEISS provides repeat high-resolution optical imagery for land and disaster monitoring.
- KMSS High-resolution VIS/IR Radiometer
KMSS 60-120m 6-band VNIR imager with 960 km swath and 3-day repeat enables land surface change detection time-series; used operationally for Russian Federation territory monitoring
Detects land-cover or structural change by comparing backscatter intensity across dates; good for flood mapping, deforestation alerts, and disaster damage assessment.
Detects fine-scale surface disturbance by measuring the loss of interferometric phase coherence between two SAR acquisitions, sensitive to changes too subtle to alter backscatter amplitude. Used for damage assessment and disturbed-ground mapping.
Maps the extent and timing of fire-affected area by tracking persistent post-fire changes in surface reflectance across a time-series, often anchored by coincident active-fire detections. Produces systematic burned-area records.
- [1]Humid tropical forest disturbance alerts using Landsat datapeer reviewed2026-06-04(2016) Primary GLAD-L algorithm paper: bagged decision trees on Landsat red, NIR, SWIR temporal metrics; confirmation rule 2/4 observations within 180d
- [2]Forest disturbance alerts for the Congo Basin using Sentinel-1peer reviewed2026-06-04(2021) RADD: Bayesian conditional probability of forest disturbance; per-pixel Gaussian mixture baseline; low-confidence >85%, high-confidence >97.5%; Sentinel-1 C-band 10m
- [3]Integrating satellite-based forest disturbance alerts improves detection timeliness and confidencepeer reviewed2026-06-04(2024) Describes GLAD-L, GLAD-S2, RADD side-by-side; GLAD-S2 uses 4-tier confidence (single, low, medium, high) via decision trees on Sentinel-2 bands 2,3,4,8,11,12
Edited from public sources. Last reviewed date pending by SpectraWorks editorial. See the data dictionary for field definitions.