26 October 2026
Displacement height and why it shifts your wind shear profile

Displacement height, also called zero-plane displacement and usually written as d, is the height above actual ground where the wind profile behaves as if the surface sits. Over bare ground or short grass, d is close to zero and you can ignore it. Over a forest canopy, a shelterbelt, or tall crop, the air near the tree tops drags on the canopy the same way it drags on open ground near the surface, so the whole log-law or power-law profile shifts upward by roughly d. Measure wind speed at 60 m and treat it as 60 m above ground on a forested ridge, and you've quietly mis-stated your effective height by however tall the canopy is.
The usual rule of thumb puts d at somewhere between 0.6 and 0.8 of canopy height h, with denser stands running toward the high end. A loblolly pine stand at 22 m might carry a d around 15 m. That's not a rounding error on a hub height extrapolation. It changes your shear exponent, your roughness length, and ultimately the number you report for hub-height wind speed.
What the correction changes in the math
The standard power-law shear equation,
v2 = v1 × (z2/z1)^α
assumes z1 and z2 are measured from the surface the wind actually sees. Over a canopy, you replace z with (z - d) everywhere height appears. Get d wrong, or skip it, and you've fed the shear-exponent fit two anemometer heights that aren't really the heights you think they are. The fitted α comes out distorted, usually flatter than it should be, because the model is trying to explain a profile shift with a slope parameter instead. Roughness length z0 gets dragged along with it. If you're back-calculating z0 from a measured profile without correcting for d first, you'll get a roughness estimate that's too large, and that roughness estimate then feeds into every extrapolation downstream, including the one from your mast height up to hub height.
This matters most on exactly the terrain where you'd consider skipping it: a ridge crest with mature timber on the upslope approach, a saddle lined with scrub oak, a bluff where the canopy thins toward the top but isn't gone. Treat the whole site as "d = 0" because the met mast itself will clear the canopy, and you're extrapolating from a shear profile that was never measured against bare ground in the first place.
Where this shows up before the mast goes in
Canopy height correction is a desk-study problem before it's a field problem. You need a canopy height number for every candidate pad and every stretch of upslope fetch feeding wind into it, not just an average for the parcel. A ridge with 25 m hardwood on the northeast approach and a logged clearing on the southwest approach has two different effective ground planes depending on wind direction, and a single-point d correction applied to the whole site will be wrong for half your wind rose.
This is the kind of terrain detail that's cheaper to catch from stereo imagery than from a crew walking the fetch with a clinometer. Obstacle height, canopy extent, and the surface roughness transitions along each approach direction are exactly what a terrain screen run against candidate turbine positions is built to surface before anyone drives a mast into the ground. Screening a ridge before you commit a crew to it means you walk in already knowing where the canopy correction bites and where it doesn't.
A couple of things worth saying plainly
Displacement height isn't a fixed constant per land cover type. A clear-cut that was forest two years ago has a different d than the stand next to it that wasn't logged, even though both show up as "forest" on an older land-cover layer. And d matters more for taller, denser canopies and matters less as fetch distance from the obstacle increases, since the internal boundary layer the canopy generates takes time to decay.
If your site has any forested upslope fetch at all, get a canopy height figure for it before you fit a shear exponent, not after.
Run a terrain screen on your ridge before the next mast goes up.