Drone photogrammetry vs LiDAR for golf courses.

Photogrammetry builds the model from overlapping photographs, so it records the surface the camera can see, in full color. LiDAR measures distance with laser pulses, so it can reach the ground through gaps in vegetation, but it records no picture of its own.

For the parts of a golf course that are open, mown turf (greens, tees, fairways, bunkers), photogrammetry gives a detailed, photoreal model and terrain you can measure. Where the ground is hidden under trees or tall native grass, LiDAR sees what photogrammetry cannot.

In photogrammetry the drone flies a grid and takes hundreds or thousands of overlapping photos. Software finds the same points in many photos and solves for their position in 3D. The results are a textured 3D mesh, a stitched aerial image, and an elevation model of the visible surface.

A LiDAR sensor fires laser pulses and times their return. One pulse can return more than once: from a leaf, a branch, and then the ground. Sorting those returns separates vegetation from bare earth. The result is a point cloud, which is precise but colorless unless a camera is flown alongside.

Most of what a club wants to measure is closely mown and open to the sky. On a green cut to a few millimeters, the top of the grass and the ground are the same surface for practical purposes, so a photogrammetry model is a fair record of the contours. The same holds for tees and fairways.

The gap opens as the grass gets longer and the cover gets thicker. In deep rough a photo-based model sits on top of the grass. Under a tree line it shows the crowns. A drainage swale that runs through woods, or a creek bank under overhanging branches, is where LiDAR earns its cost.

Water is a blind spot for both. Ponds and streams move and reflect, so neither method gives a dependable surface, and neither sees the bottom.

Either method can produce very fine relative detail: the shape of a green, the fall across a fairway. How well the whole model sits in real-world coordinates is a separate matter. That depends on surveyed ground control, on how the flight positions were corrected, and on whether anyone checked the result against independent points.

So the useful question for any provider, whichever sensor they fly, is what the data is fit for. A working model for maintenance, planning and play is one thing. Work that will be engineered or built from, such as drainage design, construction staking or anything touching a boundary, should rest on a licensed survey.

Start from what you need the data to do.

Course Correct flies drone photogrammetry. That choice follows from the product: a golf course digital twin has to look like the course for golfers and measure like the course for staff, and one photo flight delivers both.

It also sets the limits we state plainly. The twin is working detail for maintenance, planning and play. It is indicative, not survey-grade, and under tree canopy it shows the trees, not the ground beneath them.