C‑DRONE
Waves rolling onto a sandy beach seen from above

C-DRONE GUIDE · 5 SEPTEMBER 2026

Bathing water vulnerability profile: what a drone survey brings a French municipality (method, limits, price)

In 2025 France listed 3,378 bathing sites — 2,089 coastal, 1,289 inland — and had close to 34,000 samples analysed by accredited laboratories commissioned by the regional health agencies. Behind those analyses sits a less visible duty on the party responsible for the bathing water, usually the municipality: drawing up, reviewing and updating a profile that identifies and assesses every source of pollution liable to reach the bathing area. That profile is not a water analysis: it is a desk-and-field study whose review interval depends on the classification obtained — every two years if the water is rated "poor". This is exactly where a drone survey fits: not to measure bacteriology, which it cannot do, but to see and date what the profile requires you to inventory. Here is how, within what limits, and at what price.

Published on 5 September 2026, reviewed on 11 September 2026 — regulations in force as of September 2026.

What the law actually requires: a profile, not a water analysis

The duty comes from directive 2006/7/EC of 15 February 2006 on the management of bathing water quality, which repealed the 1976 directive. France transposed it into articles L. 1332-3 and D. 1332-20 onwards of the public health code. Article L. 1332-3 designates the responsible party: whoever declared the bathing area under article L. 1332-1 or, failing a declarant, "the municipality or the group of local authorities with jurisdiction over the territory in which the bathing water is situated". That party must "draw up, review and update the bathing water profile, which includes in particular an inventory and an assessment of possible sources of pollution".

The content is set by article D. 1332-20, which mirrors annex III of the directive. The profile must contain:

The profile is sent to the mayor, who forwards it to the regional health agency (article D. 1332-21). It contains no analysis: those belong to the health agency's monitoring programme, which looks for Escherichia coli and intestinal enterococci, the only two microbiological parameters in annex I of the directive. At the levels usually found, these germs are not necessarily dangerous in themselves, but their presence signals the likely presence of pathogens. They underpin the four-category classification — poor, sufficient, good, excellent — established at the end of each season from the data of that season and the three preceding ones (article D. 1332-27). For coastal water rated "excellent", the directive sets values of 250 cfu/100 mL for E. coli and 100 cfu/100 mL for intestinal enterococci, assessed at the 95th percentile; in inland waters, 500 and 200 cfu/100 mL respectively.

The operational conclusion is clear: a drone does not count E. coli, it has no business in the official monitoring programme, and no aerial image will ever change a classification. What the profile asks you to inventory, locate and assess, on the other hand, is largely visible from the air.

The nearby catchment: what a flight sees and a ground visit misses

The heart of the profile is the inventory of sources. On a bathing site the list always revolves around the same objects: stormwater outfalls and combined-sewer overflows, discharges from treatment plants or on-site sanitation systems, streams and ditches emerging onto the beach or bank, grazed parcels and livestock watering points along watercourses, motorhome parking areas and campsites with their emptying stations, harbours and mooring zones, bird colonies, car parks and roads draining towards the water.

The difficulty is not knowing that list but locating it exhaustively and keeping it current. A ground visit follows walkable paths; an outfall buried in a wooded bank, a ditch re-dug the previous season, a herd moved onto a newly fenced upstream parcel, a motorhome area that appeared on private land — these easily escape an observer on foot. A drone flight produces, in a morning, a dated, georeferenced orthophoto of the site and its upstream surroundings, in which every identified source can be pinned, numbered and appended to the profile with its coordinates.

This attention to the catchment is not documentary overkill. A study by D. Kay, J. Crowther, C. M. Stapleton, M. D. Wyer and co-authors, published in 2008 in Water Research from 205 watercourse sampling points spread across mainland UK, measures gaps of four orders of magnitude between catchments depending on land use, and shows that faecal indicator organism concentrations exceed base-flow values by more than one order of magnitude during high-flow episodes — with export coefficients rising by roughly two orders of magnitude (see the study on Google Scholar). In other words: what matters for bathing plays out upstream and in wet weather, and a source whose position nobody wrote into the profile is a source no management measure covers.

A useful deliverable combines three layers: the dated orthophoto, a layer of annotated source points (type, outfall, presumed operator, distance to the bathing area), and a trace of the surface flows visible towards the water. This output follows the logic already described in our guide to the municipal orthophoto by drone for a local authority, with a footprint tightened onto the catchment immediately around the site.

Outfall thermography and multispectral imagery: what they show, what they do not

Two sensors widen the reading beyond a standard photo, each with a narrow domain of validity that deserves to be stated honestly.

The thermal camera exploits a simple principle: water discharged by an outfall or a spring is rarely at the same temperature as the receiving water. Where that gap exists, it draws a plume visible in thermal infrared and reveals an active discharge that no visual trace announces — an outfall running in dry weather when it should not, an unrecorded inflow beneath riprap, a structure overflowing after a storm. The limit is equally simple: when temperatures equalise, the contrast vanishes and the discharge becomes invisible. A thermal flight only makes sense within a chosen time and weather window, and a thermal plume is not a contamination plume: it indicates a flow, not its bacterial load.

Multispectral imagery, for its part, maps optical water parameters. A study by A. Román, A. Tovar-Sánchez, A. Gauci, A. Deidun and co-authors, published in 2023 in Remote Sensing, flew a five-band multispectral sensor on a drone over two sites on the Maltese coast and retrieved chlorophyll-a and total suspended solids at centimetre scale, where the Sentinel-3 OLCI satellite imagery available for the same area works at 300 m resolution; the authors also note that standard photogrammetric chains fail over a moving water surface and that a dedicated workflow is required (see the study on Google Scholar). For a bathing water profile, this documents localised turbidity, a suspended-solids plume at a stream mouth after rain, or the algal proliferation potential the profile requires you to assess — a topic covered in detail in our guide to cyanobacteria and algae monitoring on a water body by drone.

What neither does: measure E. coli or intestinal enterococci. No airborne sensor counts bacteria, and no optical index substitutes for that. The meta-analysis by T. J. Wade, N. Pai, J. N. S. Eisenberg and J. M. Colford, published in 2003 in Environmental Health Perspectives from 27 studies retained out of 976 screened, is a reminder of why those parameters remain central: a one-log10 increase in enterococci concentration is associated there with a relative risk of gastrointestinal illness of 1.34 in marine water (95% confidence interval: 1.00-1.75), while in fresh water E. coli proves the more consistent predictor, with a relative risk of 2.12 per log10 — but a wide confidence interval (0.925-4.85) that includes no effect and which the authors flag as such (see the study on Google Scholar). The drone directs the sampling; it does not replace it. On the sampling act itself and its admissibility, see our dedicated guide to water sampling by drone.

Revising a profile after a downgrade, and supporting a preventive closure

This is where the topic becomes urgent for a municipality. Article D. 1332-22 of the public health code requires a review covering all elements of the profile at least:

Water rated "excellent" needs a review only if its classification is downgraded. The text further specifies that the frequency and extent of reviews must be adapted to the nature, frequency and severity of the pollution risks the water is exposed to: the schedule is a floor, not a ceiling. A profile must also be revised when works or infrastructure changes of significant scale, in the bathing zone or its immediate vicinity, are liable to affect water quality.

A downgrade is therefore mechanically the trigger for tightened fieldwork, on a short timetable, and often in a tense political context: the classification is public, displayed on site and published online. Of the 3,378 sites listed in 2025, 90.2% were rated "good" or "excellent" quality (74.4% excellent and 15.8% good), and 97 sites fell into the "no classification" category for lack of sufficient data. The remaining minority is the one that has to redo its profile most often.

In that setting, a drone campaign serves two precise functions. The first is comparative: if a flight was flown when the initial profile was drawn up, the repeat flight produces a dated comparison highlighting what changed in the catchment — a re-dug ditch, a parcel returned to grazing, an extended parking area, an outfall put back into service. That is exactly what the word "update" in article L. 1332-3 calls for, and it is very hard to produce other than with two superimposable orthophotos. The second is support for a preventive closure decision: article L. 1332-4 allows both the responsible party and the mayor to order a temporary preventive closure where bathers' health is endangered. That decision is often taken before any analysis result, on a bundle of indications — heavy rain, a combined-sewer overflow triggered, a visible plume. A short, logged and timestamped flight documents the state observed at the moment of the decision: it does not ground the closure legally, but it makes it objective in the file, which matters when the measure is contested by a beach concession-holder or a tourist office. The same logic of dated visual evidence is developed for a different regime in our guide to monitoring drinking water catchment protection zones.

Flight framework, GDPR on a busy beach, deliverables and 2026 prices

When to fly. The bathing season generally runs from 15 June to 15 September for coastal bathing and from 1 July to 31 August for fresh water; in the French overseas departments it most often covers the whole year. For profile work, the best window is out of season: vegetation is low, outfalls are exposed, the beach is empty — which incidentally settles the question of people being filmed. Two passes remain useful: one in dry weather, one within 24 to 48 hours of a marked rainfall event, since that is when flows become visible.

GDPR. Flying over a busy beach captures identifiable, often scantily clothed people: the processing is then sensitive and the "bathing water profile" purpose does not justify it. Three simple rules: prefer an out-of-season or early-morning flight, frame on structures and flows rather than on visitor numbers, and systematically blur in post-processing anyone left in frame. A catchment orthophoto needs no faces. The exact obligations — legal basis, retention period, impact assessment where required — are detailed in our guide to GDPR applied to professional drone work.

Airspace framework. A bathing site often stacks several constraints: proximity to a built-up area, a Natura 2000 site or nature reserve, municipal by-laws specific to the coast. The full picture is set out in our guide to drone rules on the beach and the French coast; it constrains the flight window as much as the weather does.

Deliverables. The set useful to a consultancy commissioned for the profile, or to a technical department drafting it in-house, includes: the dated georeferenced orthophoto of the site and its near catchment, the GIS layer of inventoried sources (annotated points, exportable as Shapefile or GeoPackage), the photo report per outfall with coordinates, and — where the campaign provides for it — the thermal mosaic of outfalls with the time window and weather conditions recorded. That last point is not cosmetic: without the capture conditions, a thermal image cannot be interpreted after the fact.

Orders of magnitude observed in France in 2026 (excl. VAT):

A municipality managing several beaches has every interest in grouping the campaign into a single weather window: mobilising the remote pilot is the dominant cost item, not flight time. A retention pond upstream of the site is a separate diagnostic, described in our guide to stormwater retention pond surveys by drone. To scope an intervention, request a quote stating the number of sites, the classification obtained last season and the profile's review deadline.

Frequently asked questions

Can a drone survey replace the official health monitoring analyses?

No, and that limit is structural, not technical. The classification of a bathing water rests on two microbiological parameters — Escherichia coli and intestinal enterococci — counted in an accredited laboratory from samples taken under the regional health agency's monitoring programme. No drone-borne sensor, thermal or multispectral, counts bacteria: thermography reveals a flow, multispectral imagery reveals optical parameters such as turbidity or chlorophyll-a, never a faecal load. The flight belongs to the documentary part of the duty — the inventory and location of sources required by article D. 1332-20 — and to directing sampling, not to regulatory measurement.

How often must a bathing water profile be reviewed?

The interval depends on the classification obtained, under article D. 1332-22 of the French public health code: at least every four years for water rated "good", every three years for "sufficient", every two years for "poor". Water rated "excellent" is only reviewed if its classification is downgraded. These are minima: the text requires the frequency and extent of reviews to be adapted to the nature, frequency and severity of pollution risks. A review is also required when significant works or infrastructure changes, in the bathing zone or its immediate vicinity, are liable to affect water quality.

Who must draw up the profile: the municipality or the beach operator?

Article L. 1332-3 of the French public health code designates the party responsible for the bathing water: whoever declared the bathing area under article L. 1332-1 or, failing a declarant, the municipality or the group of local authorities with jurisdiction over the territory where the bathing water is located. For a public beach open to all, that is in practice the municipality or the competent inter-municipal body; for a bathing area declared by a private operator — leisure park, campsite —, it is the declarant. The profile is sent to the mayor, who forwards it to the regional health agency (article D. 1332-21). Municipalities also list, each year, the bathing waters located on their territory.

Request a free quote

Also worth reading