C-DRONE GUIDE · 2 SEPTEMBER 2026

Counting anchored boats and anchoring pressure on Posidonia seagrass by drone: method, legal framework, price

A coastal municipality on the French Riviera preparing an application for a light-equipment mooring zone, a marine nature park that must demonstrate how anchoring pressure on its seagrass meadow is changing, a consultancy commissioned to establish the baseline before a mooring order: all of them hit the same methodological wall. AIS data only sees part of the fleet, counts made from a signal station or a patrol boat are expensive and remain one-off, and diver ground-truthing covers only a few dozen square metres per day. A drone replaces none of these tools, but it fills precisely the gap between them: a dated, georeferenced, verifiable picture of an entire bay, obtained in an hour, repeatable every week of the high season. This guide describes what a flight actually delivers, what it does not, the legal framework that makes the data useful, and the prices observed in 2026.

Published on 2 September 2026, reviewed on 7 September 2026 — regulations in force as of September 2026.

What a drone sees of an anchorage, and what AIS does not

Anchoring pressure on Mediterranean seabeds is now documented at large scale using ships' AIS data. A study by Deter, Lozupone, Inacio, Boissery and Holon published in 2017 in Marine Pollution Bulletin estimated that pressure over 1,800 km of French Mediterranean coastline between 2010 and 2015: around 30% of habitats lying between 0 and −80 m showed anchoring pressure, and Posidonia oceanica meadows stood out as the habitat most affected in cumulative duration (read the study). The article's own title states the limit of the method: it requires bias estimation, because AIS is carried by only a fraction of vessels.

That is precisely where a drone contributes. In a cove in August, the anchored fleet is mostly made up of 6 to 15 metre boats that carry no transponder: they are invisible to AIS, yet their anchors plough the same seabed. A high-altitude nadir flight over the bay captures every vessel present at the moment of the shot, whatever its size, each with a position in Lambert-93 or WGS84, a length measured to within a decimetre on the orthophoto, and a timestamp.

In practice, such a flight produces three categories of data that are hard to obtain otherwise: a headcount (how many vessels, at what time, on what day), a spatial distribution (where they cluster in the bay, at what depth, over what type of seabed) and a fleet structure (breakdown by length class, share of units over 24 metres). All three feed concrete decisions: sizing the number of moorings in a future mooring zone, justifying the perimeter of a ban, or measuring the effect of an order one year after it was signed.

The flight protocol: counting objects that move

An anchorage is not a construction site: vessels swing on their anchor, come and go, and a boat photographed at the start of a mosaic may have turned 90° by the end. This calls for a protocol different from a standard topographic survey.

For counting, the fastest possible coverage is preferred: flight at the maximum legal height, wide-angle sensor, overlap cut to the strict minimum, so as to freeze the bay in a few minutes rather than half an hour. Over a cove of less than a square kilometre, a single high-altitude frame is sometimes enough, leaving no double-counting ambiguity. Over a wider bay, the area is split into blocks flown back to back with the time of each block recorded: the few vessels in transit between two blocks are tracked manually during processing. The ground sampling distance chosen need not be centimetric: 4 to 8 cm per pixel is enough to identify a hull, measure it and tell a sailing yacht from a RIB.

For seabed mapping, everything changes: lower altitude, higher overlap, ground control points positioned with differential GPS on the riprap or the beach, and a mandatory flight within a narrow window of conditions. Repeatability comes from standardisation: same solar time, same strip orientation, same exposure settings from one campaign to the next. That discipline, more than the equipment, is what allows two campaigns a year apart to be overlaid and a patch to be declared as having grown.

One often-overlooked point: data protection. At low height, vessel names, registration numbers and people on board become legible. Counting needs none of that information. Minimisation must be built into the flight plan from the outset — sufficient height, blurring on delivery, bounded retention period — along the lines detailed in our guide to GDPR applied to drone imagery.

Reading the meadow from the sky: what the method allows, and what it does not

Over a healthy Mediterranean seabed, a Posidonia meadow appears from the air as a dark, almost brown blanket, standing out sharply against pale sand. Repeated anchoring digs patches of sand into the matte, and a dragging anchor leaves a linear scar. These features are visible from the air, and that is the whole point of the method: a high-resolution orthophoto of a cove makes those patches measurable in area, countable, and above all comparable from one year to the next. A study by Ventura, Bonifazi, Gravina, Belluscio and Ardizzone published in 2018 in Remote Sensing demonstrated automated mapping and classification of ecologically sensitive marine habitats, Posidonia meadows among them, from drone imagery processed with object-based image analysis (see the paper).

The stakes are not cosmetic. A study by Abadie, Lejeune, Pergent and Gobert published in 2016 in Marine Pollution Bulletin described the mechanism by which the mechanical wound left by an anchor then evolves into chemical degradation of the matte, which explains how these anthropogenic patches are generated and why they persist in the meadow (read the study). In other words: a scar does not close over within a season, which is what makes multi-year monitoring relevant — and makes a cumulative degraded-area indicator credible.

The limits must be stated before the mission, not discovered afterwards. Aerial photogrammetry only sees through water under narrow conditions:

• Shallow water only. The exercise is reliable in very shallow water. Beyond that, the contrast between meadow and sand collapses and reading becomes interpretation, not measurement. Usable depth depends on the day's turbidity: it is observed on site, it is not promised in a quote.
• Clear water and flat sea. A light chop is enough to break the reading of the seabed and to defeat photogrammetric matching below the surface. Residual swell or a late-morning thermal breeze cancels the window.
• High sun, but controlled reflection. Light is needed to penetrate the water, hence a high sun, but a sun at its zenith throws a blinding specular hotspot into the middle of the frames. Work therefore happens in a mid-morning or mid-afternoon slot, with a polarising filter oriented to cut surface reflection.
• No depth without refraction correction. Through the air-water interface, every point of the seabed appears shallower than it is. A raw 3D model systematically underestimates depths: a refraction correction must be applied, a method formalised by J. T. Dietrich in 2017 in Earth Surface Processes and Landforms (see the publication). Without that correction, no depth value derived from a flight is usable.

In practice, a serious contractor therefore sells area mapping (outlines, surfaces, positions of patches and scars), not bathymetry. If precise depth is needed, it belongs to another tool: an on-board echo sounder, or a dedicated bathymetric survey. And ground-truthing by diver or towed camera remains essential to anchor the interpretation on a few control plots.

The legal framework that makes this data useful (regulations in force in September 2026)

An orthophoto is only valuable to a manager if it hooks onto a legal text. Three legal blocks structure the subject in the Mediterranean.

Species protection. Posidonia oceanica is listed in the order of 19 July 1988 on the list of protected marine plant species, under the vernacular names "pelote de mer" and "chiendent marin". That listing prohibits the destruction, transport, sale, purchase and use of specimens. The penalty falls under article L. 415-3 of the Environmental Code, which provides for up to three years' imprisonment and a €150,000 fine, doubled where the acts are committed in the core of a national park or in a nature reserve. This is the foundation that gives legal weight to a demonstration of meadow degradation.

Mooring enforcement. In the Mediterranean, prefectural order no. 123/2019 of 3 June 2019 from the maritime prefect of the Mediterranean sets the general framework for the mooring and stopping of vessels in French internal and territorial Mediterranean waters. It is implemented through a series of departmental orders — the first were signed for the Alpes-Maritimes in October 2020 — which define zones where mooring is prohibited for vessels of 24 metres and over, with a threshold locally lowered to 20 metres in certain bays. These local orders change regularly, including on maximum stay duration: the exact perimeter applicable to a given bay must be checked against the order in force published by the maritime prefecture of the Mediterranean, never copied from a secondary source.

Mooring organised by the local authority. This is the regime of the zone de mouillages et d'équipements légers (ZMEL), a light-equipment mooring zone. Article L. 2124-5 of the General Code of Public Property allows temporary occupation permits to be issued for the layout, organisation and management of mooring zones, provided the works and equipment are not liable to cause irreversible alteration of the site; these permits are granted as a priority to municipalities or groupings of municipalities, or after their opinion if they waive that priority. The detailed rules are in articles R. 2124-39 to R. 2124-55 of the same code: an agreement lasting a maximum of fifteen years, equipment exclusively intended for mooring or launching, mobile and removable installations, and demolition and site restoration at expiry at the holder's expense.

For a municipality, the articulation is direct: the aerial count documents the need (how many units, of what size, in what period), the meadow mapping justifies placing moorings off the matte, on existing sand patches, and later campaigns measure the effect of the scheme. The same data serves as baseline evidence in a ZMEL application and as a monitoring indicator in the management plan of a marine protected area.

Flying over a bay and a marine protected area

The flight itself falls under the ordinary coastal framework, which we do not rewrite here: our guide to flying a drone on a beach and along the coast covers permissions, municipal police orders and beach practices. When the bay lies within a marine Natura 2000 site, a nature reserve or the core of a national park, the manager's own rules stack on top of the aviation rules — that is the subject of our guide to flying in Natura 2000 sites and nature reserves. On the French Riviera, airspace itself is a sizing factor: the controlled zones of Nice and Cannes-Mandelieu cover a good part of the coastline and require prior coordination whose lead time must be built into the campaign schedule.

Three precautions specific to flying over the sea are worth adding. A loss at sea is final: an incident over water means losing the aircraft and any memory cards not yet downloaded, which justifies floats, a wider battery margin than over land, and downloading data after every sortie. The take-off point is an operational choice: a jetty, a mole or a beach gives better visual line of sight than a spot set back behind the pines. Marine co-activity is real: shuttles, jet skis, parasailing and liaison helicopters share the same bay in August, and the flight must be announced to the water-body managers and to the rescue posts.

What is delivered to a municipality, a mooring-zone manager or a marine park

The deliverable determines the usefulness of the mission. On this type of campaign, a complete package contains:

• A georeferenced orthophoto of the study area, dated and timestamped, in the coordinate system expected by the client's GIS department — formats and their uses are described in our guide to drone photogrammetry deliverables.
• A vector layer of "vessel" points, one point per unit, with attributes: identifier, estimated length, length class, vessel type where identifiable, approximate depth, nature of the seabed under the anchor, date and time of the frame.
• A vector layer of "sand patch" polygons and "scar" polylines, with the area of each feature and the difference measured against the previous campaign.
• A method note documenting without ambiguity the flight conditions (time, height, GSD, sea state, observed turbidity, filter used), the processing applied and, above all, the areas where the seabed could not be read. That last item is what distinguishes an admissible report from a pretty picture.
• A quantified summary per survey date: total count, breakdown by length class, occupancy rate of the bay, cumulative degraded meadow area.

The principle remains that of any drone photogrammetric survey: the data is produced to be loaded into a GIS and replayed the following year, not to illustrate a presentation. A campaign whose flight conditions cannot be reproduced has no monitoring value.

Prices observed in 2026

Price depends mainly on the water surface to be covered, the number of passes over the season, and whether or not a seabed mapping component is included — considerably more demanding than counting alone, because it requires ground control points, a narrow weather window and possible postponements. The ranges below are orders of magnitude observed on the French market in 2026, given for guidance; they are not a fixed tariff and a quote remains necessary.

Two items make the difference between a low and a high quote. The number of weather postponements accepted: on a seabed mapping component the usable window is short, and a contractor committing to a given reading quality budgets for cancelled days. The level of GIS processing: delivering raw images costs little; delivering attributed, checked vector layers compared against the previous year often accounts for half the budget. For a local authority, the order usually goes through a public services contract; specifying standardised flight conditions and data ownership in the tender documents is what guarantees successive campaigns remain comparable.

Scoping your campaign

A serious estimate rests on four elements: the exact perimeter of the water body to cover, the period and frequency of passes required, whether a meadow mapping component is included, and the format expected by your GIS department. If the monitoring must also cover the emerged shoreline of the same bay, it can be combined with coastal erosion monitoring by photogrammetry during the same flights, which appreciably reduces the cost per campaign. Describe your bay and your deadlines in our quote request: you will receive a costed proposal, with realistic flight windows and postponement terms.

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