Wind in the lowest few hundred meters of the atmosphere is measured almost entirely from fixed masts. A mast returns continuous high-frequency data, but only from one height in one place, and it costs enough that networks of them are rare. Radiosondes drift away unrecovered, while lidar and sodar average each reading over a sample volume and smooth out the sharp fluctuations that make this layer significant. Small drones look like the obvious way to fill the gap, but the propellers keeping a drone in the air push flow straight past the sensor meant to be measuring the wind. The established fix is to move the anemometer well clear of the rotors, which forces a larger and heavier aircraft.

The goal is to find out whether a small consumer drone can measure wind and turbulence with the sensor mounted close in on a short rod, how short that rod can be before the propeller wash corrupts the readings, and what the resulting records say about the air above the site.

The author does this by building the instrument and testing it, rather than modeling it. A TriSonica Mini ultrasonic anemometer sits above a DJI Mavic 3E on a carbon-fiber rod, logging all three velocity components at 40 Hz. The rotor wash was characterized indoors first, hovering the drone where the true wind is near zero so that every reading is effectively an error, with a continuous run for each of ten rod lengths from 5 to 30 cm. The chosen setup was then flown through full outdoor wind profiles.

Main findings

• Indoors, propellers-on readings ran roughly 0.5 to 0.9 m/s against a 0.0765 m/s propellers-off background, so rotor wash, not sensor noise, dominates what a close-mounted anemometer reports.
• Outdoors, neither the morning nor the afternoon profile rose smoothly with height between 20 and 120 m, and the two flights differed by up to 2.04 m/s at the same altitude.
• The standard deviations differentiated the flight attempts more clearly than the means. Morning tests were positively skewed with tails heavier than Gaussian, while the afternoon holds were near-Gaussian or lighter, so the intermittent gusting belongs to the morning air, not the platform.

Additional Considerations

• A propellers-off baseline should be reported alongside any drone wind measurement, since, without one, there is no separating rotor wash from instrument noise.
• Each configuration needs repeated sessions, and each altitude sequence should be flown more than once in varied order, so that height dependence can be held apart from the air simply changing through the morning.
• Any campaign of this kind should run against a calibrated reference anemometer at known heights, the single change that would turn these results from indicative into quantitative.

In short, a short rod on a small consumer drone is enough to return usable turbulence statistics, and what now limits the measurement is not the airframe but the lack of a reference to check it against.