How to Map an Orchard With Just a Phone

Ask five growers where their trees are and you will get five answers: a hand-drawn sketch in a shed, pins on Google Maps, a boundary traced in a farm app, or nothing at all beyond memory. All of them work, right up to the day a new hire needs to find the young rambutan that got knocked over in a storm, or a buyer wants to know how many bearing trees a block holds. Mapping an orchard with a phone is free or close to it, and it is the foundation every other record sits on. Here are the four methods, what each one is good for, and a worked walkthrough of the one that gives you the most.

Method 1: the paper sketch (start here if you have nothing)

A drawing of the block with trees as circles and rows as lines costs nothing and takes an afternoon. Its value is not precision; it is completeness. Sketching forces you to walk every row and notice the tree you forgot existed. Its limits show fast: no coordinates, no way to hand it to anyone else without explanation, and it goes stale the first time a tree is removed. Treat the sketch as a draft to be replaced, not a system.

Method 2: pins on a satellite map

Open any satellite basemap (Google Maps, Google Earth, or a farm app with a map layer), zoom to your block, and drop a pin on each tree. Satellite imagery in the dry season often shows canopy shadows well enough to place trees without walking at all, and you can name each pin with its tree ID. It is genuinely useful for a first inventory.

• Cost: free, done from a chair, workable in one evening for a few dozen trees.
• Watch out for: young or small trees that are invisible from above, imagery that is years old, and offsets between the satellite image and the real ground (check a corner you know, like a gate or a well).
• The fix for both: walk the block once with the phone and correct each pin to the actual trunk.

Method 3: the GPS walk (boundaries and acreage)

Turn on a GPS tracker and walk or drive the boundary of each block. Modern phones locate you within a few metres in the open, which is plenty for block boundaries and acreage: a one-hectare block measured with a few metres of boundary error lands within a few percent of its true area. This method answers how big is the orchard, not where is each tree, so it is the base layer rather than the inventory itself. It matters more than most growers expect: it is easy to misjudge your own acreage, and planted area is what you buy inputs for, report for certifications, and base planting plans on.

One honest caveat applies to all three phone methods: accuracy degrades under canopy. In the open, a few metres is normal. Under dense, wet foliage the error can grow substantially, which matters at the tree level because fruit trees are often only 5 to 10 metres apart. There are field habits that keep errors small: stand at the trunk when you shoot, give the phone a few seconds to settle on each fix, and prefer fixes with a clear patch of sky overhead. At orchard spacing, metre-level ambition is unnecessary; tree-level ambition is not.

Method 4: the geotagged photo survey (trees, not just boundaries)

The step up from pins is a photo survey. Every photo a phone takes carries GPS coordinates in its EXIF metadata, so a walk through the rows photographing trees produces a set of location-tagged frames that can be clustered back into individual trees: photos taken within a few metres of the same trunk belong to the same tree. One afternoon of walking turns into a map where every marker is a real tree with its own photos attached, not a dot you drew. This is the method that produced a 72-tree durian orchard twin from 243 geotagged photos in a single survey.

The rule that makes the photo survey work: turn on location tagging before you start, and photograph from beside the trunk, not from ten metres away zoomed in.

Worked walkthrough: one hectare of mixed tropical trees

1. 1Prep (10 minutes): confirm the camera can access location, take one test photo, and check it shows coordinates in the photo details. Charge the phone.
2. 2Walk (allow an afternoon for a first block): go row by row in one direction so nothing is skipped. At each tree take two or three frames from different angles: trunk, canopy, and any fruit or flowers. For cauliflorous crops like durian, jackfruit, and cacao, shoot the trunk deliberately, because that is where the flowers and much of the fruit hang.
3. 3Count back at the desk: the photo set clusters into trees by location. Expect roughly the number of trees you walked past; a large mismatch means missed trees or doubles, and both are easy to spot on the map.
4. 4Tag the identities: give each mapped tree its ID and, if you use physical labels, print QR tags that scan straight to each tree's page.
5. 5Repeat next season: the second survey is where the method pays off. The same trees, photographed the same way, give you change over time: which flowered, which lagged, which died.

Choosing between the methods

• A handful of trees you know by name: the sketch is enough. Do not over-engineer a 12-tree yard.
• One block, 20 to 100 trees, no crew: satellite pins corrected by one walk, or a photo survey if you also want per-tree photos and counts.
• Multiple blocks, a crew, or buyers asking questions: the photo survey feeding a tree-level map. Pins alone cannot carry photos, health notes, or yield history.
• Large acreage where walking is impractical: a drone flight for the layout, then a phone survey for the per-tree detail. The two feed the same map.

Whichever method you pick, keep one habit from the sketch era: never reuse a position. When a tree dies, the map keeps the dead tree in the record with its end date, and the replacement becomes a new tree at a new point in time. An orchard map that forgets its dead trees cannot explain its gaps.

CanopyTwin is built around the photo-survey method: you photograph trees with your phone, GPS clusters the photos into individually tracked trees on a live satellite map, and each tree keeps its own page with photo history, AI fruit and flower counts, and an AI arborist diagnosis. It works on the web and on iOS, CSV import brings in an existing list, and every plan starts with a 30-day free trial. The demo orchard at canopytwin.com/map shows the end result before you map your own first row.