A gobbler sleeps in a single tree. Sitting eighty yards from that tree instead of eighty yards from another tree two hundred yards along the same bench can change whether he lands in front of you or flies off the back into the next hollow. Many hunters use aerial imagery today to decide where to sit. Ordinary aerial imagery shows the spring canopy as a pattern rather than separate trees. The hunter chooses a ridge while the bird chooses a trunk. This paper looks at ground sample distance, the amount of ground covered by one pixel, and what changes as that distance shrinks from half a meter to three inches.
KEY FINDINGS
- The beginning of spring turkey season overlaps with imagery taken before the leaves fill in. During that early stretch of the season, the hardwood canopy remains open. This is the only window each year when an aerial photo reveals the branches a bird roosts on instead of the layer of leaves above them.
- Consumer map services have partly addressed the difference. Some of the Midwest's turkey range has leaf-off basemaps with pixels measuring 30 to 50 cm. Our county orthoimagery records the bare canopy with pixels measuring 3 to 6 inches, or roughly 7.5 cm to 15 cm. That is roughly two times finer per side at the low end and six and a half at the high end, depending on the tile.
- The calculations show a large difference. A mature tree grown in the open, with a crown roughly forty feet wide, measures about 24 pixels from edge to edge at 50 cm. That reveals the tree's presence, but often cannot distinguish its boundary from the neighboring crown. With three-inch pixels, that very crown measures about 160 pixels in width.
- Smaller pixels make neighboring crowns easier to distinguish. Once the detail crosses a threshold, you can identify an individual tree within a cluster. If ground sample distance is coarse enough, the crowns of two neighboring trees become a single outline. Zooming cannot restore their separate forms.
IWhat a basemap shows you in April
Helps answer the hundred-acre question; offers no help with the fifty-yard question
In early April, any map app showing a hardwood drainage displays a gray-brown ribbed pattern of leafless branches. The pattern looks unbroken. The creek remains visible, along with field boundaries, the two-track road and the overall course of the woods. Those details really do help you decide where to begin across a hundred acres.
That view does little to answer the fifty-yard question: which tree does the bird use? Turkeys generally choose tall roost trees with broad, open crowns, level limbs, and clear space below for flying in and out, often close to a creek and an open field. These traits belong to an individual tree. A basemap that shows the stand clearly still leaves the tree unresolved. A hunter relying on that map makes an educated guess about a ridge, then walks it at four in the morning and hopes to be close.
IIThe resolution a crown needs
What ground sample distance means
Ground sample distance gives an honest description of this difference. Exact wording matters here because imagery sales language seldom provides it. With 50 cm imagery, each pixel represents about a foot plus half a foot of ground on each side. A crown forty feet wide therefore occupies roughly two dozen pixels across. You can see that the tree exists, but cannot reliably distinguish its boundary from an adjoining tree's crown. Two mature trees with touching canopies appear as a single outline.
Swipe table to see every column
| Ground sample distance | Pixels across a 40 ft crown | What resolves |
|---|---|---|
| 50 cm | 24 | A tree is present |
| 30 cm | 41 | Crown outline, edges still merging |
| 15 cm | 81 | Crown separation begins |
| 3 in (7.6 cm) | 160 | Limb structure, gaps, shadow shape |
With three-inch resolution, the pixel count across that same crown is a hundred and sixty or thereabouts. Its interior branches show up, and you can clearly distinguish the space separating neighboring crowns. A tree rising above surrounding trees throws a shadow you can judge for length and shape against the slope. The move between these resolutions is not a gradual improvement in how pretty the image looks. Shrinking pixels improves separation between objects. Beyond a threshold, an individual tree becomes identifiable within a cluster.
With three-inch pixels, you can pick out a particular roost tree. With half-meter pixels, a blurred patch of timber could contain one tree or three.
IIISeeing it, not just calculating it
Comparing imagery fairly
On paper, the pixel calculations resolve the question. Seeing the difference brings the point home. We can demonstrate it fairly using our imagery instead of presenting a competitor's poorest view. Select one creek-bottom drainage. Use its image at the original resolution of three to six inches, then reduce that same image's resolution to half a meter. Change pixel size alone. The capture date stays fixed, as do the sun's angle, leaf conditions and rendering settings. Trees distinguishable as separate objects in the original are expected to become texture in the reduced-resolution version.
IVWhat this does not tell you
Tree structure shortens the search but cannot pinpoint a bird
We do not publish roost pins and never will. This imagery can flag possible roost trees, but it cannot tell you whether birds roost there. Along a drainage, it picks out trees with the physical traits turkeys favor, giving you a much shorter list than “the woods on this bench.” Whether a bird roosted in one of those trees last night may depend on hunting pressure, the location of hens, recent weather, and other conditions a photograph cannot capture. The layer can guide you toward trees worth checking on foot. You still have to locate the bird.
There are two additional limits to state clearly. Counties have different capture dates. By mid-May, a late-season hunt can take place long after the leaves have filled in, concealing branches from scouts on the ground while the earlier leaf-off photo continues to show leafless treetops. A standing dead tree can closely resemble a living tree in leaf-off photos. What looks like a perfectly defined crown in the orthoimagery may prove to be a snag once you stand beneath it. The same honest approach applies to deer bedding: terrain features and physical structure guide the search, while cover and behavior determine the animal's location.
NOTES AND SOURCES
- Our county orthoimagery of the bare canopy uses a ground sample distance of 3 to 6 in. Coverage includes Indiana, with 67 counties, and Ohio, with 44 counties. Photos were taken in spring; capture dates differ across counties, and the canopy changes as the season progresses.
- The comparison cites a basemap resolution category (30 to 50 cm) based on publicly available specifications for consumer imagery, without assigning that category to any single named provider.
- For a crown measuring 40 ft, pixel counts come from dividing its width by ground sample distance. A resolution ratio of 2 to 6× comes from calculating with the given distances, 0.30 to 0.50 m compared with 0.076 to 0.15 m. It is not separately measured.
- The structure that marks a possible roost tree is an open crown, height above the trees immediately around it and space for flight access. This describes the tree's physical form and never claims that a bird is there.
Fully Drawn Research is an independent data analysis desk, not affiliated with any state wildlife agency, transportation agency, mapping provider or outfitter. Fully Drawn Research is an independent data and analysis desk. This paper is educational: it explains ground sample distance as an imagery property and what that property means for identifying roost-candidate tree structure during the bare-canopy window. FDR treats roost selection as cover- and behavior-driven and publishes structure as a search-narrowing layer, never as a prediction of bird location and never as a pin.
Keep reading
FDR Scout puts research like this on the map of the ground you hunt. Open Scout →