Open-cast mine expansion monitoring for cable route conflict detection
Open-cast mines expand their boundaries continuously, often encroaching on buried cables or microwave corridors surveyed years earlier. Multi-temporal optical change detection and InSAR subsidence mapping can project conflict zones before earthworks arrive.
Sensors
- Sentinel-2 MSI: 10 m resolution in visible and near-infrared bands, 5-day revisit at mid-latitudes with two satellites. Tracks bare-earth advance of the pit boundary through NDVI suppression and spectral change detection across a 290 km swath. Free archive from 2015.
- Planet PlanetScope: 3 m resolution, daily revisit globally. Resolves individual bench-face advances and spoil-heap growth that are sub-pixel in Sentinel-2. Useful for high-cadence monitoring during active blasting seasons, though archive access requires commercial licence.
- Sentinel-1 C-band InSAR: 6-day repeat pass (12-day single satellite), C-band at 5.4 GHz. Interferometric processing detects line-of-sight ground displacement at sub-centimetre precision over stable surfaces. Identifies subsidence halos around pit walls weeks before visible boundary change. Free archive from 2014.
- ALOS-2 PALSAR-2 L-band InSAR: L-band at 1.27 GHz penetrates dry overburden and maintains coherence over longer intervals and rougher terrain than C-band, making it better suited to detecting deep-seated deformation behind pit walls. 14-day repeat; archive from 2014. JAXA access required.
Why the installation survey date is the problem
A buried fibre or power cable is routed to avoid a mine's footprint as it exists at the time of design. Permits, however, are typically issued for a much larger area than the initial pit. Mines expand progressively over years or decades, and the rate of advance is rarely communicated to telecoms operators whose infrastructure crosses the permitted zone.
The result is a slow-motion conflict: the pit edge advances at rates that can range from tens to several hundred metres per year depending on commodity prices and extraction method, while the cable record sits unchanged in a GIS. By the time a contractor's survey flags the encroachment, rerouting options are constrained by active earthworks, blasting exclusion zones and access restrictions. Earlier warning changes the economics of rerouting entirely.
What a floating roof gives away: reading pit-edge advance from optical time series
Open-cast mines produce a distinctive spectral signature. Bare overburden, blasted rock and standing water in pit sumps all differ sharply from surrounding vegetation or soil in Sentinel-2's shortwave infrared bands (SWIR1 at 1610 nm, SWIR2 at 2190 nm). A multi-temporal stack of Sentinel-2 scenes, cloud-masked and atmospherically corrected using the Sen2Cor processor, allows per-pixel classification of disturbed ground across each acquisition date.
Comparing classified extents across quarterly or monthly epochs yields a pit-edge advance vector: direction and rate of boundary movement. At 10 m resolution, Sentinel-2 can resolve bench advances of roughly 20 m or more reliably. PlanetScope's 3 m pixels push that detection floor down to individual bench widths of 10 to 15 m, which matters when a cable route runs close to the current boundary. Neither sensor sees through cloud, so in high-rainfall mining regions (parts of central Africa, Indonesia, northern Queensland) optical revisit is effectively degraded and must be supplemented by SAR.
The subsidence halo: deformation arrives before the diggers do
Pit-wall failure and stress redistribution in the surrounding rock mass cause measurable ground deformation well outside the visible pit boundary. Persistent-scatterer InSAR (PS-InSAR) applied to a Sentinel-1 time series can detect line-of-sight displacement rates as low as a few millimetres per year over stable corner reflectors or coherent surfaces. Around active pit walls, deformation rates of 10 to 50 mm per year in the subsidence halo are not unusual in published literature, and these can extend 50 to 200 m beyond the visible edge depending on geology and pit depth.
For buried cable operators, this matters independently of the surface boundary. A cable duct subjected to differential settlement of even 20 to 30 mm along a short span can experience joint separation or conduit cracking. InSAR deformation maps therefore serve two functions: they identify where the ground is already moving outside the mapped pit extent, and they act as a leading indicator of where the visible boundary is likely to advance next.
C-band Sentinel-1 is the practical workhorse here given its free access and 6-day repeat, but it loses coherence quickly over disturbed spoil heaps and wet ground. ALOS-2 PALSAR-2's L-band maintains coherence better over rough terrain and is worth adding where C-band decorrelates.
Projecting future conflict zones
Advance-rate vectors derived from the optical time series can be extrapolated forward to intersect with cable route geometries held in a GIS. A simple linear projection gives a first-order conflict date: the year in which the pit boundary, at its current rate, reaches the cable corridor. That projection carries honest uncertainty. Advance rates are not constant: they accelerate when commodity prices rise and slow or reverse during care-and-maintenance periods. A responsible analysis presents a range of scenarios (current rate, 50% acceleration, 50% deceleration) rather than a single date.
Microwave line-of-sight links face a different geometry. As a pit deepens and spoil heaps grow, the terrain model changes. A link that cleared the horizon at installation may be obstructed by a new spoil heap, or conversely may gain clearance as the pit floor drops. This is a terrain-model problem rather than a boundary-encroachment problem, and it sits on a sibling page in this library. The present analysis focuses on the physical encroachment of earthworks on buried infrastructure.
Honest limits of the method
Cloud cover is the most consistent operational constraint. In persistently cloudy mining regions, Sentinel-2 may yield only four to six usable scenes per year rather than the theoretical 72. PlanetScope improves this but does not eliminate it. SAR fills part of the gap but InSAR coherence degrades over actively disturbed ground, meaning the pit interior itself is often incoherent and only the surrounding stable terrain yields displacement measurements.
Advance-rate projections assume that the mine continues operating under its current permit. Permit extensions, commodity-price shocks and regulatory interventions all alter the trajectory in ways no satellite programme can anticipate. The analysis should be treated as a physical monitoring input to a planning process, not as a substitute for engagement with the mine operator or review of the current mining permit boundary.
Satellize runs this type of multi-temporal change detection on open Sentinel archives, with optional PlanetScope tasking added on client licence. The approach is methodologically similar to the crop-area estimation work delivered for the Kingdom of Tonga, adapted here for industrial boundary tracking rather than agricultural classification.
What to do with the output
The primary deliverable is a conflict-risk register: each cable segment within a defined buffer of the mine permit boundary is ranked by estimated years to encroachment and by current deformation exposure. Segments flagged as high-risk within a three-year horizon warrant physical survey and rerouting feasibility work. Segments in the five-to-ten-year window can be scheduled for periodic re-assessment.
For microwave operators, the deformation map identifies tower-base locations within the subsidence halo that may require foundation inspection, even where the surface boundary has not yet reached the site. A quarterly update cadence is usually sufficient for mines advancing at typical rates, with an alert trigger if the advance rate accelerates by more than 30% between epochs.
Typical figures
| Optical spatial resolution | 10 m (Sentinel-2 MSI); 3 m (PlanetScope) |
| SAR spatial resolution | 5 × 20 m to 20 × 20 m IW mode (Sentinel-1); 3 × 3 m to 10 × 10 m (ALOS-2 PALSAR-2 spotlight/stripmap) |
| Optical revisit | 5 days at mid-latitudes (Sentinel-2 two-satellite); daily (PlanetScope) |
| SAR repeat pass | 6 days (Sentinel-1 single pair); 14 days (ALOS-2) |
| InSAR displacement sensitivity | Sub-centimetre line-of-sight precision over coherent surfaces; practical detection threshold typically 5–10 mm/year for PS-InSAR time series |
| Minimum detectable pit-edge advance (optical) | ~20 m per epoch (Sentinel-2); ~10–15 m per epoch (PlanetScope) |
| Archive depth | Sentinel-2 from 2015; Sentinel-1 from 2014; ALOS-2 from 2014; PlanetScope from approximately 2016 (commercial) |
| Cloud limitation | Optical only; SAR is all-weather. High-rainfall regions may yield fewer than 6 usable optical scenes per year |
| Delivery format | GeoTIFF change layers, GeoPackage conflict-risk register, PDF quarterly report, optional WMS/WMTS feed |
| Update cadence | Quarterly standard; monthly available; alert-triggered on advance-rate threshold breach |
Analytics Satellize can run
| Pit-boundary advance map | Multi-temporal binary change detection on atmospherically corrected Sentinel-2 SWIR composite; per-epoch classified extents differenced to yield advance polygons | GeoTIFF change layer and advance-vector GeoPackage, updated quarterly |
| Advance-rate and trajectory model | Linear regression on pit-edge position time series; scenario-bracketed projection (current, +50%, -50% rate) intersected with cable route geometry | Conflict-date range table per cable segment, delivered as PDF report and CSV |
| InSAR subsidence halo map | PS-InSAR or SBAS processing of Sentinel-1 IW SLC stack; mean line-of-sight velocity map clipped to buffer around pit permit boundary | GeoTIFF displacement-rate raster and vector deformation-zone polygons |
| Cable-segment conflict-risk register | Spatial join of advance trajectory, deformation zones and cable route GIS; risk scoring by years-to-encroachment and current displacement exposure | GeoPackage attribute table ranked by risk tier, with recommended action per segment |
| Advance-rate acceleration alert | Epoch-over-epoch rate comparison against configurable threshold; triggered when current-period advance exceeds prior 12-month mean by defined percentage | Automated alert email with supporting map extract |
| L-band coherence supplement | ALOS-2 PALSAR-2 interferometric processing over areas where Sentinel-1 C-band loses coherence; deformation products merged into unified displacement map | Merged displacement GeoTIFF flagging areas of C-band decorrelation |
Who does the work
We can get this done for you. Satellize runs its own analyst desk and a strong science team. You do not buy a data feed and work out what it means; our people source the imagery, run the analysis described on this page, and hand you the answer with its confidence limits stated. Discuss this requirement.