C-DRONE GUIDE · 7 SEPTEMBER 2026

Coral reef and lagoon monitoring by drone: what the aerial view adds in France's overseas territories, and what it does not replace

With around 60,000 km² of reefs spread across twelve overseas territories, France is the world's fourth reef nation and holds 10% of the global reef area. Those reef flats and lagoons are also among the hardest habitats to monitor: transect diving gives fine ground truth but only over a few dozen square metres at a time, while satellites cover everything at a resolution far too coarse to read an anchor scar or a two-metre bleaching patch. Between the two, drones occupy a precise — and narrow — niche. They only see what the light lets them see: a few metres of clear water, at low tide, on a flat sea, with the sun neither too high nor too low. Here is what they genuinely bring to marine protected area managers, environment departments and marine consultancies, what they cost, and what they will never replace.

Published on 7 September 2026, reviewed on 7 September 2026 — regulations in force as of September 2026.

An overlooked reef nation, under measured pressure

The figure often comes as a surprise: according to Chiffres clés de la mer et du littoral, published by the French environment ministry's statistics service (2024 edition), France's overseas territories hold around 60,000 km² of coral reefs across twelve territories — New Caledonia, French Polynesia, Wallis and Futuna, Clipperton, the Scattered Islands, Guadeloupe, Martinique, Saint Martin, Saint Barthélemy, Mayotte and Réunion. That is 10% of the world's reef area, placing France fourth in the world. New Caledonia's lagoons, inscribed on the UNESCO World Heritage List on 8 July 2008 across six marine zones and 1,574,300 hectares, alone cover nearly 60% of the reefs under Caledonian jurisdiction. In Mayotte, the marine nature park created by decree on 18 January 2010 spans some 68,400 km² and a lagoon of around 1,300 km² with a double barrier reef — a configuration with fewer than ten examples worldwide.

The state of that heritage is documented: Ifrecor produces a health assessment every five years. The 2020 assessment found 62% of monitoring stations in a degraded state in the West Indies and the Indian Ocean, against 30% in the Pacific and the Scattered Islands; between 2015 and 2020, degradation was recorded at 33% of stations in the West Indies, Mayotte and Réunion. Globally, NOAA and ICRI declared the fourth global bleaching event on 15 April 2024; in April 2025, ICRI reported that since 1 January 2023, bleaching-level heat stress had affected 83.7% of the planet's reefs, across at least 83 countries and territories.

Local pressures stack on top of warming: anchoring and boating pressure on accessible reef flats, terrigenous inputs and turbidity after heavy rain — a major issue in Réunion and Mayotte, where gullies and eroded catchments discharge sediment plumes straight onto fringing reefs — macro-litter, and coastal development (rock armour, fill, port structures) that alters circulation inside the lagoon. Those pressures, visible from the surface, are precisely what defines the drone's working ground.

What a flight over a reef flat actually produces

The core deliverable is a high-resolution orthophoto of the reef flat and shallow lagoon, georeferenced, at a few centimetres per pixel. On that base, a trained operator delineates the main habitat classes visible from the surface: sand, seagrass, massive coral, branching coral, coral rubble, pavement. A team led by Brian O. Nieuwenhuis published in 2022, in the journal Remote Sensing, a demonstration of what that approach yields once automated: over three very shallow reef areas (under 5 m) surveyed with an RTK drone, processed with SfM photogrammetry and object-based image analysis, adding geomorphometric variables derived from the digital elevation model to the orthomosaic's spectral variables raised overall classification accuracy by up to 11 percentage points (see the study on Google Scholar). In other words: colour alone is not enough, and bottom roughness and relief carry independent, decisive information.

Beyond habitat mapping, four uses come up repeatedly in managers' specifications: multi-year monitoring through exact repetition of the same flight plan, turning two orthophotos into a usable change map; detecting a bleaching episode at reef-flat scale, where the white of bleached coral stands out sharply on the imagery and lets you quantify an extent rather than extrapolate from a handful of transects; mapping turbid plumes after heavy rain, documenting the extent and trajectory of a terrigenous input during the few hours it stays visible; and counting visitor pressure and anchored vessels, with the exact position of each moored boat and, over light bottoms, the trace of anchor scars — the same method detailed in our guide to counting anchored boats and anchoring pressure on seagrass beds, transposed from the Mediterranean to a tropical lagoon.

Then there is the most contractual use: the baseline survey and as-built check for a coastal development project subject to an environmental impact assessment. A dated orthophoto, with its control points and method note, is admissible evidence: it fixes the state of the seabed before works and lets you measure the extent actually affected afterwards. That is the same repeatable geometric-evidence logic as in our guide to coastal erosion monitoring by photogrammetry, applied this time beneath the water surface.

Light dictates everything: sun glint, tide, water column, and what the drone does not replace

This is the part honest providers state before quoting. A drone does not penetrate water: it exploits the light that crosses the water column, reflects off the bottom and comes back up. Three constraints follow. First depth: in clear lagoon water, bottom reading stays usable down to a few metres and degrades fast beyond that; Lyzenga's water-column correction algorithm, the historical reference in reef remote sensing, gives reliable results down to roughly 5 metres, and is not a checkbox in a software package but a genuine processing step to calibrate. Then sea state: a light chop is enough to shred the imagery and make the mosaic unusable. Finally specular reflection of the sun — sun glint — which burns out whole swathes of the orthomosaic.

That last point is covered in the literature. A team led by Aidy M. Muslim published in 2019, in the journal Remote Sensing, a comparison of the main sun glint correction methods applied to drone imagery (Lyzenga et al., Joyce, Hedley et al., Goodman et al.), on a coral site at Pulau Bidong in Peninsular Malaysia (see the study on Google Scholar). Their finding is directly operational: applying the Lyzenga et al. correction to the glint-affected regions only, rather than to the whole image, delivers the best habitat classification accuracy — branching coral, tabulate coral, patch coral, rubble, sand. Translated for a specification: deliverable quality depends as much on processing as on the flight itself.

In practice the usable window is short: low tide or close to slack water to minimise water thickness, a flat sea, a sun high enough to light the bottom but not exactly overhead in the camera axis, and clear water — so typically not in the days following heavy rain, unless the turbid plume is precisely what the flight is about. On a given site, that can mean a handful of usable slots per month. A serious quote budgets for weather postponements.

Be clear about what a drone does not do. It does not give live coral cover by species: only transect diving, with taxonomic identification and quadrat counts, produces that data — exactly what standardised monitoring protocols require. It does not measure bathymetry beyond a few metres: past the light penetration limit you need a multibeam echo sounder or bathymetric LiDAR, as explained in our guide to drone bathymetry and its limits. It cannot tell you whether a white coral is bleached, dead or overgrown by algae without ground truth. And it sees nothing below the drop-off, on outer slopes, or in turbid water. The drone is a tool for extent and repeatability, not biological diagnosis: it tells you where and how much, diving tells you what.

Who commissions this kind of mission, method and 2026 prices

First things first: overflight. A protected reef cannot be flown over freely, and the rule comes from the instrument that created the protected area, not from a general provision. A documented example: the Réunion national marine nature reserve, created by decree no. 2007-236 of 21 February 2007 — 3,500 hectares along 40 km of the west coast, covering around 80% of the island's reefs. Its article 19 prohibits powered aircraft from flying over the reserve below 300 metres, except with authorisation issued by the prefect. The manager adds that flight below 30 metres is never authorised, that strict protection zones remain off-limits, and that every request is jointly reviewed by the reserve's management body and the Réunion environment department, on production of the remote pilot's theoretical and practical training certificates. The general logic is the one set out in our guide to drones in Natura 2000 sites and French nature reserves: read the decree for the specific area, never transpose from one site to another.

On top of that comes a difference in aviation framework by territory, often missed by mainland providers. In the overseas departments (Guadeloupe, Martinique, French Guiana, Réunion, Mayotte), EU outermost regions, European law applies exactly as in mainland France — a topic covered in our guide to drone services in France's overseas departments. In French Polynesia, the European drone regulation also applies, supplemented by the airspace order of 3 December 2020, with operator registration on AlphaTango (source: SEAC-PF). In New Caledonia, by contrast, the European regulation does not apply: special-activity flights are conducted under the national standard scenarios S1, S2 and S3 of the order of 17 December 2015 as amended, under the authority of the local civil aviation directorate. Check this before drafting a multi-territory specification.

Who commissions the work: managers of marine nature reserves and marine nature parks, the French biodiversity office, overseas authorities and coastal municipalities, State services, marine consultancies working on impact assessments, operators contributing to Ifrecor monitoring, and coastal development promoters. Private individuals are not the audience here: over a protected reef, a recreational flight is most often simply prohibited.

Ranges observed in 2026, excluding VAT, and excluding the team's flights and accommodation when the provider travels from mainland France:

These amounts cover the flight, processing and delivery of the layers; they exclude dive-based ground truth, taxonomic identification and acoustic bathymetry, which fall to other providers and remain essential to any ecological diagnosis. Nor do they cover the overflight authorisation process, whose lead time needs generous anticipation: on a marine reserve, allow several weeks. For a marine protected area, a local authority or a consultancy, request a quote stating the territory, the target extent, the maximum expected depth and whether an admissible baseline survey is required.

Request a free quote

Also worth reading

• Race circuits and karting tracks: drone track surveys, and the rule that changes everything once there's a crowd
• Sargassum strandings in the French Caribbean: what a drone actually measures on the foreshore
• Sugarcane seen from a drone: vigour, ratoons and biomass in Réunion and the French Antilles