High-voltage transmission line corridor construction monitoring
Metallic transmission towers produce unmistakable double-bounce signatures in SAR imagery, allowing tower-by-tower erection verification. Sentinel-2 NDVI time-series independently confirms vegetation clearance along the right-of-way. Together they give project owners and lenders an objective construction record without site visits.
Sensors
- Sentinel-1 (C-band SAR): 5 x 20 m resolution in IW mode, 6-day revisit at mid-latitudes with both satellites active. C-band (5.405 GHz) produces strong double-bounce returns from vertical metallic structures such as lattice towers. Free and globally archived from 2014.
- TerraSAR-X (X-band SAR): Spotlight mode delivers 1–2 m resolution; StripMap 3 m. X-band is more sensitive to fine metallic geometry than C-band, improving discrimination of tower type and orientation. Tasked commercially; revisit on request, typically 2–4 days at most latitudes.
- Sentinel-2 MSI (optical multispectral): 10 m resolution in visible and near-infrared bands; 5-day revisit with both satellites. NDVI derived from Band 4 (red) and Band 8 (NIR) quantifies vegetation suppression along the cleared corridor. Cloud cover is the principal operational constraint.
- Planet SuperDove: 3 m resolution, daily revisit globally. Eight spectral bands including red-edge support high-frequency NDVI monitoring and optical confirmation of cleared ground between SAR acquisitions. Commercial licence required; archive from 2021 at 3 m.
What a steel lattice tower looks like to a radar
A high-voltage transmission tower is, from a SAR perspective, an almost ideal target. The vertical metallic legs and cross-members form dihedral corner reflectors with the surrounding ground. Incident radar energy bounces from the ground to the structure and directly back to the sensor, producing a double-bounce return that is substantially brighter than the surrounding terrain. In Sentinel-1 IW mode, a completed 40-metre lattice tower typically produces a backscatter anomaly detectable at 5 x 20 m pixel spacing. In TerraSAR-X Spotlight, individual leg geometry becomes resolvable.
The practical consequence is that tower installation can be tracked acquisition by acquisition. A bare earthwork pad appears first as a low-backscatter disturbed-soil patch. Once the steel structure rises above roughly two metres, the double-bounce signature emerges. By comparing sequential SAR amplitude images, an analyst can assign an approximate installation date to each tower with confidence, provided acquisitions are frequent enough. Sentinel-1's six-day repeat (or twelve-day with a single satellite) sets the temporal resolution of that record.
Conductor stringing: what satellites cannot directly see
This is worth stating plainly. Overhead conductors on a 400 kV line are typically 25–40 mm in diameter. That is far below the resolution floor of any freely available SAR or optical sensor, and below the practical detection limit of commercial constellations in standard imaging modes. Conductors do not appear as discrete features in satellite imagery.
Inference is the only option. Once towers along a span are verified as erected and the project schedule indicates stringing has begun, analysts can look for secondary signals: access track activity near stringing equipment, the appearance of temporary tensioning platforms, or the cessation of heavy vehicle movement along the corridor. None of these is a direct confirmation. Project owners relying on satellite data for milestone payments tied to conductor installation should treat satellite evidence as corroborating rather than definitive for that specific activity. A single site inspection or contractor photograph closes the gap.
Reading the right-of-way from NDVI suppression
Vegetation clearance along a transmission corridor is the other major trackable activity. Before any tower is erected, a right-of-way typically 30–80 metres wide must be cleared of trees and tall shrubs. Sentinel-2 NDVI time-series captures this well. A forested pixel carries NDVI values in the range 0.6–0.9; freshly cleared ground drops to 0.1–0.3. That contrast is large and persistent, detectable even with moderate cloud contamination if a multi-month time-series is composited.
The approach has a useful asymmetry: clearance detection is more reliable than tower detection in heavily forested terrain, because the spectral change is dramatic and spatially broad. SAR double-bounce works best in open or semi-open terrain where surrounding clutter is low. In dense forest, the canopy suppresses the double-bounce return from a newly erected tower until surrounding trees are cleared. Combining both sensors resolves the ambiguity: NDVI suppression confirms clearing has occurred, which in turn validates that any subsequent SAR anomaly is a tower rather than a forest artefact.
Regrowth is also detectable. If a cleared corridor begins to recover vegetation, NDVI rises back toward pre-clearance values within one to three growing seasons depending on climate. For long-duration projects or operational lines, this provides an ongoing encroachment signal without any additional sensor cost.
Building a tower-by-tower installation record
A practical monitoring workflow starts with the approved route alignment, digitised as a centreline. Tower positions from engineering drawings or early-phase optical imagery are registered as point features. Each SAR acquisition is then processed to extract backscatter values at those points, compared against a pre-construction baseline derived from the same sensor and geometry. A threshold exceedance flags probable tower presence; the date of first exceedance becomes the estimated installation date.
False positives arise from standing water (which can suppress backscatter and create confusing transitions), agricultural equipment parked near the corridor, and side-lobe artefacts from nearby infrastructure. A minimum of two consecutive acquisitions showing elevated backscatter at the same point substantially reduces false positives. TerraSAR-X, where budget allows, provides a useful cross-check on ambiguous Sentinel-1 detections because its higher resolution and X-band frequency respond differently to non-tower clutter.
The output is a georeferenced table: tower ID, planned position, first-detection date, confidence rating. That table is directly usable by project managers tracking earned value, and by lenders verifying drawdown milestones. Archive depth matters here. Sentinel-1 data from 2014 onward means that for lines already under construction or recently completed, a retrospective record can be reconstructed without any new tasking cost.
Honest limits of the method
Resolution sets a hard floor. Sentinel-1 at 5 x 20 m can confirm a tower is present but cannot distinguish a 220 kV lattice from a 500 kV guyed structure. TerraSAR-X at 1–2 m begins to resolve structural type, but at significantly higher cost per acquisition.
Revisit frequency determines the precision of installation dating. A six-day revisit means installation dates carry an uncertainty of up to six days. For milestone verification tied to specific calendar dates, that uncertainty should be disclosed to lenders or off-takers.
Persistent cloud cover over tropical corridors can interrupt Sentinel-2 NDVI time-series for weeks. SAR is cloud-independent, which is one reason the two sensors are complementary rather than redundant. Even so, heavy tropical rainfall can increase SAR surface moisture returns and temporarily depress the contrast between cleared and vegetated ground.
Finally, the method assumes the approved route alignment is known and accurate. If a contractor deviates from the permitted corridor, the analysis will flag anomalies in the wrong locations unless the reference geometry is updated.
From raw detections to a project-finance deliverable
The analytic products that matter to a project owner or infrastructure lender are not backscatter maps. They are milestone tables, percentage-complete curves and exception alerts when construction falls behind schedule or strays outside the permitted corridor.
Satellize structures corridor-monitoring analytics around those outputs: a fortnightly GIS layer showing tower installation status colour-coded by confidence, a NDVI-change report covering the full right-of-way width, and an alert feed triggered when backscatter anomalies appear outside the approved centreline buffer. The Tonga crop-estimation programme established the organisation's approach to time-series analytics on open constellations; the same Sentinel-2 compositing and change-detection pipeline applies directly to corridor vegetation monitoring.
For projects where Sentinel-1 resolution is insufficient, commercial TerraSAR-X tasking can be added on client licence. The decision point is usually whether the project involves towers spaced at less than 200 metres in complex terrain, where adjacent double-bounce signatures begin to overlap in C-band imagery. If you are at that stage of project planning, the right conversation is about sensor selection before construction begins, not after the first towers are up.
Typical figures
| SAR spatial resolution (Sentinel-1 IW) | 5 m range x 20 m azimuth (ground range detected) |
| SAR spatial resolution (TerraSAR-X Spotlight) | 1–2 m; StripMap 3 m |
| Optical resolution (Sentinel-2 / Planet SuperDove) | 10 m (Sentinel-2 NIR/Red); 3 m (SuperDove) |
| Revisit frequency | Sentinel-1: 6 days (dual satellite); Sentinel-2: 5 days; SuperDove: daily; TerraSAR-X: 2–4 days on tasking |
| Minimum detectable structure (SAR double-bounce) | Metallic vertical structures approximately 2 m tall and above; conductor wire (25–40 mm diameter) not detectable |
| Spectral bands used | C-band 5.405 GHz (Sentinel-1); X-band 9.65 GHz (TerraSAR-X); Red (665 nm) and NIR (842 nm) for NDVI (Sentinel-2) |
| Installation-date uncertainty | Up to one revisit interval: ±6 days (Sentinel-1), ±1–2 days (SuperDove optical proxy) |
| Archive depth | Sentinel-1: from 2014; Sentinel-2: from 2015; Planet SuperDove: from 2021 at 3 m |
| Cloud penetration | SAR: all-weather; optical (Sentinel-2, SuperDove): blocked by cloud cover |
| Deliverable formats | GeoPackage / Shapefile tower-status layer; GeoTIFF NDVI-change raster; CSV milestone table; PDF periodic report |
Analytics Satellize can run
| Tower installation status map | SAR amplitude change detection against pre-construction baseline; double-bounce threshold exceedance at registered tower positions | Fortnightly GeoPackage layer with tower ID, first-detection date, confidence rating (high / probable / undetected) |
| Right-of-way vegetation clearance report | Sentinel-2 NDVI time-series compositing; pixel-wise comparison against pre-construction NDVI baseline within corridor buffer | Monthly GeoTIFF NDVI-change raster and tabular clearance percentage by corridor segment |
| Construction progress curve | Cumulative tower detections expressed as percentage of planned total; plotted against project schedule | Bi-weekly PDF chart and CSV data table suitable for lender reporting |
| Off-corridor encroachment alert | SAR backscatter anomaly detection outside approved centreline buffer (configurable width, typically 100 m either side) | Email or webhook alert within 48 hours of acquisition processing; georeferenced point feature attached |
| Vegetation regrowth monitoring (operational phase) | Ongoing Sentinel-2 NDVI monitoring post-construction to detect encroachment by regrowth into statutory clearance zone | Quarterly NDVI report flagging segments exceeding regrowth threshold |
| Retrospective construction timeline reconstruction | Batch processing of Sentinel-1 archive from 2014; sequential double-bounce detection to assign approximate erection dates to existing towers | One-off historical milestone table for completed or partially completed lines |
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.