Densest Water Network Ever Built in Quebec Estuary Puts SWOT Satellite to the Test

Deep in the St. Lawrence Estuary, where tides surge up to seven meters and saltwater collides with one of the world’s greatest freshwater flows, an international team of scientists has quietly assembled the densest water-level and wave monitoring network ever deployed in the region. Their goal is as ambitious as the terrain is treacherous: to hold the new Surface Water and Ocean Topography satellite, known as SWOT, accountable to the centimeter. The result, published in Earth System Science Data, is a publicly available benchmark dataset collected from March 2023 onward, designed to verify the accuracy of a satellite that promises to transform how humanity measures the water on its planet.

SWOT, launched in December 2022 by NASA and the French space agency CNES, is the first satellite built to capture two-dimensional maps of water surface elevation at unprecedented resolution. Its heart is the Ka-band Radar Interferometer, or KaRIn, a dual-antenna instrument that sweeps wide swaths of the Earth’s surface rather than the narrow ground track of traditional altimeters. Before scientists can trust SWOT’s measurements of rivers, lakes, estuaries and oceans, however, the satellite must be validated against ground truth measured with even greater precision. Only a handful of estuaries worldwide fell beneath SWOT’s fast-sampling calibration orbit, and among them the St. Lawrence Estuary and Saguenay Fjord stood out as the most hydrodynamically complex, making them an ideal natural laboratory for the mission’s most demanding test.

The region’s physics are extraordinary. The St. Lawrence drains the Great Lakes basin, which holds roughly a quarter of the world’s liquid freshwater, with an average discharge of 12,200 cubic meters per second at Quebec City that swells past 16,800 downstream of the Saguenay. Because the estuary converges and narrows, propagating tides are amplified dramatically, reaching ranges of about seven meters in the estuarine transition zone where the saltwater intrusion limit shifts with every tidal cycle. Storm surges riding on high tides can push total water levels into extreme flooding territory, a risk expected to worsen as sea levels rise. The Saguenay Fjord, one of the southernmost fjords in the Northern Hemisphere and uniquely intracontinental, behaves differently still: its tides act like a standing wave rather than a traveling one, producing nearly synchronous high and low tides along its entire 95-kilometer length and remarkably flat instantaneous water surfaces for most of the tidal cycle.

To capture this complexity, the team, led by Pascal Matte of Environment and Climate Change Canada and involving partners from the Canadian Hydrographic Service, Fisheries and Oceans Canada, Université Laval, Université du Québec à Rimouski, the University of Sherbrooke and NASA’s Jet Propulsion Laboratory, expanded an existing network of just nine permanent tide gauges to 31 water level stations. Twenty-two new stations were installed on both shores of the estuary, on islands and along the fjord, strategically distributed across SWOT’s two 50-kilometer-wide observation swaths and the 20-kilometer gap around the satellite’s nadir track. The instrumentation combined 18 barometrically compensated pressure transducers with 13 low-cost GNSS-Interferometric Reflectometry, or GNSS-IR, sensors, a technique that derives water level from the interference pattern between direct and water-reflected satellite navigation signals. Three directional wave buoys, including the large Canadian AZMP IML-4 platform moored in 335 meters of water and two compact Sofar Ocean Spotter buoys, added measurements of wave height, period and direction.

The technical rigor behind the dataset is what elevates it from a pile of sensor logs to a genuine benchmark. Every water level measurement was referenced to a common vertical datum, the Canadian Geodetic Vertical Datum of 1928, using real-time kinematic GNSS corrections capable of centimeter-level accuracy. Because pressure sensors infer water depth from hydrostatic pressure, the team corrected for water density using collocated salinity and temperature readings, a seemingly minor detail that proved decisive: in the brackish estuarine transition zone, neglecting density variations can introduce errors exceeding one decimeter near high tide, when saline water floods in and the water column is deepest. The researchers even documented a complete conversion workflow from the Canadian geodetic reference system to SWOT’s own vertical reference, accounting for coordinate transformations between NAD83 and ITRF2014 and for the permanent solid Earth tide deformation, so that anyone can compare the in situ data directly with satellite products.

Quality control was applied in layers. At permanent Canadian Hydrographic Service stations, water levels from three independent co-located sensors are compared against each other, against tide forecasts and against station-specific thresholds, with daily expert verification. For the newer instruments, the team flagged outliers using a scaled Median Absolute Deviation over moving 14-day windows, roughly the neap-spring tidal period, and discarded readings when temperature or salinity values betrayed a sensor left high and dry by a retreating tide. GNSS-IR data passed through a two-stage pipeline: a Lomb-Scargle Periodogram converts signal-to-noise ratios into irregular water level estimates called arcs, and a moving-window spline fit produces a smooth, regularly sampled time series, with any node carrying an error above one meter rejected. The result is a dataset released at three processing levels, from raw sensor files to fully corrected, quality-controlled products ready for scientific analysis.

The validation results are encouraging for the satellite and revealing for the estuary. Collocated pressure gauges and GNSS-IR sensors agreed with root-mean-square differences of about 7 centimeters at Saint-Laurent on Orleans Island and roughly 10 centimeters at Vieux-Québec, with near-zero mean offsets, confirming that inexpensive GNSS reflectometry can stand shoulder to shoulder with conventional gauges. In the Saguenay Fjord, twin instruments at L’Anse-Saint-Jean differed by barely 8 millimeters on average. The data also laid bare the fjord’s standing-wave character, with tidal range amplifying landward from 5.1 to 6.3 meters during spring tides while remaining almost perfectly in phase along its length, and exposed tidal asymmetries at Isle-Verte, where low waters on the shallow south shore lag behind the deeper north channel, a signature of the shallow-water tides that govern sediment and ecosystem dynamics.

Beyond validating SWOT’s elevation measurements, the network offers a rare testbed for the satellite’s ability to detect water surface slopes, a quantity SWOT must measure to about 1.7 centimeters per kilometer. The Saguenay’s standing-wave tides produce slopes so small they likely fall below the satellite’s detection threshold for most of the tidal cycle, while the propagating tides of the St. Lawrence generate gradients that should be detectable nearly continuously, including asynchronous water levels between the north and south channels around Orleans Island. The wave buoy records, meanwhile, showed that the estuary’s waves are overwhelmingly wind-driven, with wave provenance locked to the two dominant wind directions along the river axis, and that two buoys more than 60 kilometers apart recorded strikingly similar wave spectra, a consistency check that strengthens confidence in the entire observing system.

The dataset is publicly accessible on Zenodo and will be updated approximately annually as new observations are retrieved, processed and quality controlled, extending validation through SWOT’s 21-day repeat science orbit and beyond. The team plans to incorporate additional tide gauges, GNSS-IR sensors and wave buoys from the tidal river upstream to the Gulf of St. Lawrence, and future releases may include derived products for detecting ice-on and ice-off periods, characterizing river ice and estimating wave heights from GNSS signals. As climate change and rising seas intensify flood risks along the St. Lawrence, and as estuarine habitats face the so-called estuarine squeeze between saltwater intrusion and human barriers, this evolving benchmark promises to serve far beyond its original purpose: a foundation for hydrodynamic model calibration, for understanding tidal and wave processes in complex estuaries, and for a new era in which satellites and ground sensors watch the water together.

Subject of Research: A benchmark in situ dataset of water levels and waves for validating the SWOT satellite mission in the St. Lawrence Estuary and Saguenay Fjord, Quebec

Article Title: A benchmark dataset of water levels and waves for SWOT validation in the St. Lawrence Estuary and Saguenay Fjord, Quebec, Canada

Article References: Matte, P., Chartrand, X., Purnell, D., Simard, M., Christensen, A., Fortin-Legault, F., Morin, J., Côté-Nadeau, Y., Desrosiers, É., Ouellet, A., Chavanne, C., Anctil, F., Siles, G., Bedart, S., Innocenti, S., Dabboor, M., Peters, D. L., & Trudel, M. (2026). A benchmark dataset of water levels and waves for SWOT validation in the St. Lawrence Estuary and Saguenay Fjord, Quebec, Canada. Earth System Science Data, 18(10), 7269-7299. https://doi.org/10.5194/essd-18-7269-2026

Image Credits: AI Generated

DOI: 10.5194/essd-18-7269-2026

Keywords: SWOT, satellite altimetry, water levels, St. Lawrence Estuary, Saguenay Fjord, tides, GNSS-IR, wave buoys, hydrology, benchmark dataset, sea level, quality control