Data & methodology
DepthScout maps 39,800+ lakes, rivers and coastal waters across 50 US states, 8 Canadian provinces and 10 European countries. Every depth number on this site traces to a named source, and every chart is labeled with how it was made. This page explains both.
Where the depth data comes from
- State fisheries agencies. GPS depth surveys and digitized lake maps from agencies including the Ohio DNR Division of Wildlife, Indiana DNR, Michigan DNR, Minnesota DNR, NYSDEC Lake Map Series, Texas Water Development Board, Vermont/New Hampshire/Maine fisheries departments, Montana FWP, ADF&G's Morphometric Atlas of Alaskan Lakes, and the Nova Scotia Lake Inventory.
- Federal survey programs. USGS bathymetric surveys (including reservoir surveys in Puerto Rico and South Carolina), US Army Corps of Engineers reservoir data, and NOAA hydrographic surveys (ENC soundings, referenced to low tide) for coastal bays and harbors.
- Canadian provincial programs. Québec's Géobase des bathymétries de lac (GBLQ, MELCCFP), British Columbia's FISS bathymetric and lake-survey databases, and Ontario lake survey data.
- OpenStreetMap. Real shorelines and island geometry for every water, © OpenStreetMap contributors.
The two labels, and what they promise
“Survey depth data” (18,419 waters): Contours interpolated from real measured bathymetry — government GPS soundings, digitized agency survey maps, or hydrographic surveys. The humps, channels and drop-offs you see were measured, not guessed — these waters draw solid contour lines.
“Approximated contours” (everything else): A computer-generated estimate built from the water's real shoreline and a verified maximum depth — not measured contours. Solid facts, estimated lines: that's why they're drawn dashed. The shape and location are right and the depth scale is anchored to a real, named-source figure — but individual contour lines are estimates, which is why these waters draw dashed contour lines on the map. We never claim a survey badge for modeled contours, and waters with no real depth figure at all simply don't ship.
A note on survey vintage
A survey badge means the contours were measured, not that they're recent or high-resolution. Minnesota is the clearest example: our MN depth charts are the state DNR's official lake surveys, but the DNR program stretches back to the 1930s and most of its maps are digitized historical surveys drawn at 5-foot contour intervals — the median survey year across our dated MN lakes is the 1960s, and only a handful were done with modern GPS/sonar. The same is true of Ontario's MNRF transect surveys, most of them run in the 1960s through 1980s. Wherever the agency publishes a date — every dated Minnesota and Ontario water — the survey year is printed on the lake's data badge itself, so you can judge for yourself before you trust a contour. Where you need dense, boat-ready detail, a commercial sonar chart (Navionics, Humminbird LakeMaster) or your own live sonar will show more than an older agency survey.
How fishing spots are scored
A rules engine scores each piece of detected structure — points, humps, weedlines, drop-offs, creek channels, timber — against the selected species' seasonal depth range and structure preferences, drawn from published fisheries biology. Each highlighted zone shows its score and a plain-English reason. Spots are recommendations grounded in structure and season, not guarantees of fish.
How we decide what season a water is in
Season drives everything downstream — target depths, feeding windows, bait advice — so we don't read it off the calendar. March 1 is spring on a Georgia pond and dead winter on a Minnesota lake still under a foot of ice. Instead, each water's season boundaries are shifted by its latitude, following the classic rule of thumb from Hopkins' Bioclimatic Law: biological spring moves about four days later for every degree of latitude north. The shift is asymmetric — a northern water's warm season is squeezed from both ends (spring arrives late and winter arrives early), it's capped at six weeks in either direction, and no season is allowed to vanish entirely.
This is a deliberately simple, first-order model, and it has known limits. It doesn't account for elevation, so a high mountain lake reads warmer-season than it really is — an alpine lake in the Arizona high country is the sharpest case. It doesn't model thermal mass: very large, deep waters warm and cool slower than their latitude suggests, so their transitions run late. And real seasons move year to year with the weather, which a fixed latitude rule can't see. Local knowledge beats the model at the margins; as anglers log catches with conditions attached, we'll use that data to calibrate these boundaries against what fish actually do.
Where the species lists come from
For most waters the species we list come from state and provincial fisheries records — survey rosters or observed per-waterbody catch data. Two cases carry an informed estimate instead, each labeled so it never reads as a confirmed survey. The depth chart on these lakes is still real surveyed bathymetry.
Québec publishes no open per-lake fish survey, so every Québec lake's species are estimated from the province's official fish-distribution range maps (MFFP / Données Québec), narrowed to what's ecologically plausible for each lake's depth and habitat — deep-coldwater fish only where there's a coldwater refuge, big-water fish only on big water. These lakes carry a “Species estimated from range maps” note.
Ontario does publish observed per-waterbody species (MNRF's Aquatic Resource Area data), but a few hundred lakes with real survey bathymetry have no fish record. Rather than leave those blank or guess from depth alone, we estimate each from the species MNRF actually recorded in the nearest surveyed lakes, kept only where they fit the lake's depth and habitat. For remote northern lakes with few neighbours we widen the search until there's enough real survey data to lean on — so the estimate is always grounded in observed lakes from the same region, where the fish community is similar. These estimated lakes carry a “Species estimated from nearby surveyed lakes” note.
A small share of U.S. lakes — pockets in Minnesota, New Hampshire, Indiana, Montana, Connecticut, North Dakota and Nebraska — likewise have real survey bathymetry but no published fish list. They get the same nearest-surveyed-lakes estimate, bounded by the same depth and habitat rules, and carry the same “Species estimated from nearby surveyed lakes” note.
Weakest of all — and labeled as such — are waters where the only fish information published for the region is a regional list, not a record of the water itself: most of our international waters (the UK, Finland, Sweden, Norway, and the unsurveyed part of British Columbia and Alberta) and a handful of U.S. and Canadian batches whose agency publishes bathymetry but no per-water fish list. For those we start from the region's common species and narrow to the water by depth, size and each species' range. They carry a “Species estimated for the region” note, and it makes no claim about the depth chart — the depth badge speaks for that separately.
When locals correct us
The best QA we have is a person who has fished a water for thirty years. Lake associations and anglers send us their own surveys, hazard marks and corrections, and we fold them in with credit: Clary Lake in Maine was rebuilt from a lakefront association's full sonar survey; Lake Talon in Ontario carries the conservation association's Transport Canada hazard buoys; and anyone with a depth-finder log can send data through our contribute depth data page. DepthScout reviews every file against the source geometry before anything ships.
Corrections and limits
Reservoir pools rise and fall, surveys age, and models are models — if you find something wrong, use the feedback button on any lake page and we'll trace it back to the source data. Nothing on DepthScout is for navigation. Depths can be wrong or out of date; always follow official charts and local regulations on the water.