Vegetation encroachment on transmission-line corridors from canopy height models

Spaceborne lidar and multispectral time series can quantify how fast canopy is closing on transmission-line clearance zones, giving asset managers an evidence base before a fault, not after.

Sensors

GEDI (ISS spaceborne lidar): Full-waveform lidar at 25 m footprint diameter, footprints spaced 60 m along-track and 600 m across-track. Provides canopy top height, relative height percentiles (RH95, RH100) and ground elevation. Non-imaging: delivers corridor-level statistics, not span-by-span profiles. Coverage limited to latitudes 51.6° N/S by the ISS orbital inclination.
Sentinel-2 MSI: 10 m resolution in visible and near-infrared bands (B4, B8) used to compute NDVI; 20 m for red-edge bands (B5, B6, B7) sensitive to canopy vigour. Five-day revisit at mid-latitudes from the two-satellite constellation. Cloud cover is the primary operational constraint; persistent cloud can break time series for weeks in maritime climates.
Airborne discrete-return lidar: Point densities of 4 to 50 points per square metre are routine from manned or UAS platforms, resolving individual tree crowns and sub-metre height accuracy. Provides the authoritative span-by-span clearance measurement that GEDI cannot. Survey cost and mobilisation time limit repeat frequency; annual or biennial campaigns are typical for high-risk sections.
Planet SuperDove: 3 m resolution, eight spectral bands including red-edge, near-daily revisit globally. Useful for detecting newly cleared or rapidly re-vegetating sections between airborne survey epochs. Resolution is insufficient to measure canopy height directly but supports change-detection masking and encroachment extent mapping.

What the clearance standards actually require

Transmission-line clearance rules are not advisory. In Europe, ENTSO-E's operational handbook and national transpositions of the EU Electricity Regulation set minimum air-clearance distances between conductors and vegetation that vary with voltage class, conductor sag under thermal load, and wind deflection. In North America, NERC Reliability Standard FAC-003 mandates that utilities inspect rights-of-way and maintain clearances sufficient to prevent vegetation contact under the most adverse credible conductor position. The standard explicitly requires utilities to have a documented inspection and remediation programme.

The geometry matters. At 400 kV, minimum clearances to stationary vegetation are typically in the range of 3 to 5 metres, but the conductor sags under load and swings laterally in wind, so the effective danger zone is larger and three-dimensional. A tree that clears the conductor at rest in January may not clear it on a peak-load August afternoon when the conductor has sagged an additional metre or more. This is why static height measurements alone are insufficient; the analysis must account for conductor position under design conditions.

What GEDI gives you and where it stops

NASA's Global Ecosystem Dynamics Investigation lidar, mounted on the International Space Station, fires eight laser beams in a cross-track pattern. Each beam leaves a 25 m diameter footprint, and successive footprints along a beam are spaced 60 m apart. Across-track spacing between beam groups is approximately 600 m. For a transmission corridor that may be 50 to 80 m wide, this means GEDI samples the corridor intermittently rather than continuously. You can derive a statistically meaningful canopy height distribution for a multi-kilometre section, but you cannot pinpoint which specific span has a tree at 28 m versus 22 m.

What GEDI does well is change detection at the corridor level across epochs. Height-model differencing between two GEDI acquisition periods, normalised to the same ground returns, reveals whether the canopy envelope within the corridor has grown, been cleared, or remained stable. Combined with Sentinel-2 NDVI anomaly detection, which flags sections where greenness is increasing faster than the corridor-wide trend, the two datasets together can prioritise which sections warrant an airborne survey. That prioritisation is the practical value: airborne lidar campaigns are expensive, and GEDI-guided targeting reduces the lane-kilometres that need flying.

Measuring growth rate, not just height

A single canopy height model is a snapshot. What an asset manager needs is a rate: how many centimetres per year is the canopy closing on the clearance zone? Sentinel-2 provides a dense enough time series, typically 20 to 30 cloud-free observations per year in temperate regions, to fit a seasonal NDVI model and extract the inter-annual trend. Sections where NDVI is rising year-on-year after a clearance event indicate regrowth that will re-enter the danger zone on a predictable schedule.

Combining NDVI growth rate with the GEDI-derived height at the last survey epoch allows a simple projection: if the canopy is currently at 18 m, the clearance threshold is 22 m, and the regrowth rate is 0.8 m per year, the section is approximately five years from a potential violation. That projection carries real uncertainty because growth rates vary with species, rainfall, and management history. But even an order-of-magnitude estimate is more useful than no estimate, and it gives vegetation managers a defensible basis for scheduling remediation rather than reacting to complaints or post-incident audits.

One honest limit: Sentinel-2 NDVI measures canopy vigour and extent, not height. A dense 6 m shrub layer and a sparse 20 m tree canopy can produce similar NDVI values. Height attribution always requires lidar, whether spaceborne or airborne.

Airborne lidar as the compliance instrument

For regulatory reporting under FAC-003 or equivalent national standards, airborne discrete-return lidar remains the instrument of record. Point densities above 4 points per square metre allow individual tree crowns to be modelled, heights to be measured to sub-metre accuracy, and clearance distances to be computed against a surveyed conductor position. The workflow is well established: classify ground returns, generate a digital terrain model, subtract to produce a canopy height model, then compare canopy height against the three-dimensional danger-zone envelope derived from the conductor catenary and design sag tables.

The practical constraint is cost and scheduling. A single airborne lidar survey of a 500 km transmission corridor can take several days of flying and weeks of processing. Utilities typically prioritise high-voltage, high-consequence lines and accept longer inspection intervals on lower-voltage feeders. Spaceborne data does not replace this; it extends the temporal coverage between campaigns and flags the sections most likely to have changed.

Putting the layers together in an operational workflow

A workable operational architecture layers the sensors by cost and resolution. Sentinel-2 runs continuously as a background monitor, generating monthly NDVI composites for the full corridor network. Planet SuperDove provides higher-resolution change detection at sections flagged by the Sentinel-2 anomaly filter. GEDI acquisitions, which are not on-demand but are archived and updated as the ISS orbit precesses, supply canopy height context at corridor level. Airborne lidar is deployed to the sections that the satellite stack has ranked as highest risk.

Satellize applies this layered approach in analytics programmes built on open constellations, with commercial tasking added where revisit or resolution requirements exceed what Sentinel-2 alone provides. The Tonga crop-estimation programme demonstrated the same principle in a different domain: satellite-derived indices guide where ground truth is needed, rather than replacing it.

The output for a transmission operator is not a map of every tree. It is a ranked list of corridor sections by encroachment risk, updated on a defined cycle, with the evidence trail that satisfies a regulator asking how the inspection programme was prioritised. That is a different kind of deliverable from a pretty canopy height render, and it is the one that survives an audit.

Honest limits of the method

GEDI's 51.6° latitude ceiling excludes transmission networks in Scotland, Scandinavia, Canada, and Russia from spaceborne lidar coverage entirely. In those regions, airborne lidar or UAS surveys are the only height-measurement option. GEDI's non-imaging geometry also means that a corridor running diagonally across the ISS ground track may be sampled very sparsely on any given acquisition cycle.

Cloud is a persistent problem for optical sensors in maritime and tropical climates. A Sentinel-2 time series with fewer than ten cloud-free observations per year is too sparse to fit a reliable seasonal model, and gaps of two to three months are common in winter at high latitudes. Radar sensors can see through cloud but do not measure canopy height directly; SAR-based height inversion methods exist but carry larger uncertainties than lidar. No single sensor solves all of this. The honest answer is that the method works well in seasonally dry, mid-latitude environments and degrades in proportion to cloud cover and latitude.

Typical figures

GEDI footprint diameter	25 m; footprints spaced 60 m along-track, ~600 m across-track between beam groups
GEDI height accuracy	RH95 canopy height RMSE typically 2–5 m depending on terrain slope and canopy density (published GEDI validation studies)
GEDI latitude coverage	51.6° N to 51.6° S (ISS orbital inclination limit)
Sentinel-2 spatial resolution	10 m (B4, B8 NDVI bands); 20 m (red-edge bands B5–B7)
Sentinel-2 revisit	5 days at equator (2-satellite constellation); effective cloud-free revisit varies from ~10 days (dry tropics) to >30 days (maritime climates)
Airborne lidar point density	4–50 pts/m² typical; height accuracy sub-0.3 m RMSE on flat terrain
Planet SuperDove resolution	3 m; 8 spectral bands including red-edge; near-daily revisit
Minimum detectable canopy change (GEDI epoch differencing)	Approximately 2–3 m height change at corridor level; individual tree-level change not resolvable
Archive depth	GEDI: April 2019 to present; Sentinel-2: June 2015 to present; Landsat: 1972 to present (lower resolution context)
Delivery formats	GeoTIFF canopy height models, GeoPackage/Shapefile corridor risk rankings, CSV encroachment rate tables, PDF regulatory evidence reports

Analytics Satellize can run

Corridor-level canopy height model	GEDI RH95/RH100 waveform extraction and spatial aggregation per corridor section; epoch differencing for change detection	GeoTIFF height raster and section-level summary CSV, updated per GEDI acquisition cycle
NDVI growth-rate anomaly map	Sentinel-2 monthly composites, harmonic seasonal model fit, inter-annual trend extraction per 10 m pixel within corridor buffer	GeoTIFF anomaly layer and ranked section list, monthly refresh
Encroachment rate projection	Linear extrapolation of GEDI height change rate and Sentinel-2 NDVI trend to estimate years-to-threshold per corridor section	Tabular report with confidence intervals, suitable for maintenance scheduling input
Airborne survey prioritisation score	Multi-layer risk index combining GEDI height proximity to clearance threshold, NDVI growth rate, and line voltage class	Ranked corridor section list with supporting evidence layer, GIS-ready Shapefile or GeoPackage
Post-clearance regrowth monitoring	Planet SuperDove and Sentinel-2 change detection against pre-clearance baseline; species-agnostic greenness recovery tracking	Monthly change-detection report flagging sections exceeding defined regrowth thresholds
Regulatory evidence pack	Aggregation of sensor metadata, processing provenance, height measurements, and clearance standard references into a structured audit trail	PDF report with embedded GIS data, formatted to support FAC-003 or ENTSO-E inspection documentation requirements

Sources

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.