What does forestry mapping software do, and what does it leave out?

Forestry mapping software records where things are: stand boundaries, roads, streams, harvest areas, plots and protected zones. It overlays imagery, collects points and lines from phones and GPS units in the field, and prints the maps crews work from. Some of it can also store how accurately each line was captured, but only if someone sets that up, and the accuracy is rarely carried through to the reports built on the map. Boundaries digitised from imagery or inherited from old surveys usually carry no such record at all. A map looks equally certain everywhere, and so do the per-hectare figures computed from it.

What the usual answer says

The pages that answer this question list features: stand maps, layers, offline field collection, task tracking. Where they mention limits, they mean usability, such as rough terrain, weather and poor signal in the woods. The less visible limit is that a polygon drawn from a weak signal looks exactly like one drawn by a surveyor, unless the software was set up to say otherwise.

Accuracy is recorded only if someone asks for it

Esri's documentation for ArcGIS Field Maps shows how it works in one widely used product. "If you include GPS metadata fields in your feature layer, Field Maps can write GPS metadata to the respective fields when editing features," it says, and "GPS metadata is populated on point, line, and polygon layers". It also notes that with a device's internal GPS, "not all metadata fields are populated". The capability exists. Whether a business's stand layer uses it is a setting someone has to choose, and it says nothing about lines drawn on screen.

How accurate a boundary captured under canopy is

Two recent studies measured it. Thomas Purfürst, a forest operations researcher, tested ten smartphones and a geodetic receiver at 15 sites under forest canopy, published in Sensors in 2022. The best phone reached a DRMS of 4.56 metres. He also warned that "the accuracy varies greatly between smartphones, even between identical or quasi-identical tested smartphones".

Taeyoon Lee and colleagues tested three receivers at 26 control points at a forest GPS test site, published in PLOS ONE in 2023. The average horizontal error at each point ranged from 2.28 to 9.77 metres. They report that "The effect of the size of nearby trees on horizontal position error could not be generalized; however, the location of nearby trees on horizontal position error could". Error depends on the trees around each point.

What that does to per-hectare figures

Area is where boundary error turns into money. As a hypothetical, suppose every side of a square stand is drawn 5 metres too far out, a systematic error rather than the random scatter the studies measured. On a 20 hectare stand that adds about 4.5% to the area. On a 2 hectare stand it adds about 15%. Random errors partly cancel around a boundary. Consistent ones do not, such as a line walked along the inside of a fence. Every per-hectare figure divides by that area, and the report never shows it.

Set against other errors, this is often small. A 2025 study in Forests, which checked inventories against logs scaled at the mill, found cruise volume estimates between −16% and 6% of scaled volume. On a large, compact stand, boundary error is minor beside that. On small, irregular or high-value stands, it can matter as much.

What else it leaves out

Time and attribute age are data-model questions. A layer that isn't set up to keep history loses the old boundary when a stand is re-delineated, and figures from past years then divide by the wrong area. An attribute such as stocking needs an "as of" date, or nothing on the map says it is out of date. The loads, contracts and payments tied to each stand live elsewhere, and joining them is covered in our piece on why maps and numbers sit in different systems.

What to record with each line

For every boundary, store how it was captured (survey, receiver, phone, digitised from imagery), the date and an expected accuracy in metres. Then compare each small or high-value stand's mapped area with an independent one, such as a deed or an earlier survey, and flag those that differ by more than a set share.

Quarri's own map features are in design, not delivered, and this piece describes mapping software in general.

When it doesn't apply

Boundaries surveyed to a legal standard, or captured with correction services in open ground, are accurate enough that area error rarely matters. Large, compact stands are less affected than small, irregular ones. And for planning where crews go, a few metres of error changes nothing.

Quarri for forest management is built around how a forest operation runs, from the cruise to the settled account.

Quarri is an AI-native data platform for the timber supply chain. It connects buying, production, sales and inventory for forest management, sawmill, wood products and pulp, paper and packaging operators.