Soil Steps

15 October 2026 · Carbon accounting

4 km of hedge, 170 tonnes of carbon, and a code that will not pay for any of it

What a drone survey of one farm's hedgerows actually tells you, and what the Hedgerow Carbon Code would do with it.

I spent last week working up a carbon assessment for a farm I have been advising. It is a mixed holding, a shade under 39 hectares across nine parcels, with 4.19 km of hedgerow threaded between the fields. The brief was simple enough: how much carbon is here, and is any of it worth money?

The answer to the first question took a drone, a soil lab and a fair amount of trigonometry. The answer to the second is shorter. None of it.

4.19 km of hedgerow across 41 boundaries
~170 t C held in that hedge network
93% of farm carbon is in the soil, not the hedges

What the farm actually holds

Across soil, hedges and a small patch of woodland, the farm holds an estimated 3,970 tonnes of carbon, or roughly 14,560 tonnes of CO₂ equivalent. The honest range is 3,600 to 4,360 tonnes.

The split matters more than the total. Soil to 30 cm holds about 3,710 t C, roughly 96 tonnes per hectare, worked up from six lab results with organic carbon running between 2.3 and 2.6%. The hedgerows hold about 170 t C in woody biomass, above and below ground. Two thirds of a hectare of woodland holds about 95 t C.

So the hedges, which are the visible, charismatic, much-photographed part of the farm's carbon story, are about 4% of it. The soil nobody can see is 93%. That ratio is worth sitting with, because almost all the policy energy and almost all the marketing goes the other way round.

Measuring a hedge from the air, honestly

Here is the part I found genuinely interesting. The hedge dimensions came from a 15 second 4K clip shot from a small drone hovering at the corner of one field, looking along a boundary. Not a photogrammetric 3D model, not LiDAR. One hovering clip and some geometry.

The method runs in four steps. The drone's own metadata gives gimbal pitch and lens field of view, which together place the true horizon line a known number of pixels above the centre of the frame. Anything on that line sits at exactly camera height. Two people happened to be standing in shot, and assuming they are average height puts the camera at 5.6 m above the ground. From there, a hedge whose foot sits an angle below the horizon and whose top sits an angle above it resolves to a distance and a height by simple trigonometry.

That gave a hedge of about 4.5 m tall, with emergent field maples pushing 6 to 7 m and one mature ash standard at around 20 m. On the Adams and Hedgelink ten stage management cycle it scores H7: overgrown and straggly on top, but still satisfyingly dense right down to the base. Condition fair. Hawthorn, hazel, field maple, blackthorn and elder, with a bramble skirt along the bottom.

And here is where I have to be straight about the limits. Height came out at ±23%. Width could not be resolved at all from that camera angle, because nothing in the clip looks down on the hedge from above, so it is a visual estimate with ±35 to 40% on it. Push both through to a carbon figure and the hedge biomass carries roughly ±50% uncertainty. I ran a 200,000 iteration Monte Carlo over every input, varying assumed person height, pixel picks, gimbal pitch, lens field of view and ground level, to get those bounds rather than guess at them.

There was also a discrepancy worth recording. The pilot remembered flying at about 13 m. No field in the flight data reads 13 m, the stored height above take off reads 3.3 m, and the people in frame imply 5.6 m. At 13 m the hedge would compute to 10 m tall and the ash to 45 m, both plainly wrong. The full flight log would settle it. I have left the number I can defend in the report and flagged the one I cannot.

The carbon itself comes from Axe et al. (2017), who measured hedges left untrimmed for three years and found 14.0 t C/km above ground and 11.5 t C/km in roots. Scaling those by hedge cross section, following Black et al. (2023), who showed biomass rises more or less linearly with hedge volume, gives about 40 t C/km for a hedge this size. Applying that to all 4.19 km assumes every hedge on the farm looks like the one I measured, which they certainly do not. The farm figure is an upper bound.

Now the part that stings

Having established that this farm holds roughly 170 tonnes of carbon in its hedges, the obvious question from any farmer is whether that can be sold.

It cannot. The Hedgerow Carbon Code, developed by the Allerton Project and the GWCT on a Natural Environment Investment Readiness Fund grant, credits only new carbon: planting new hedge, gapping up broken sections, or letting an existing hedge grow taller and wider than it is today. Existing stock is the baseline. The 170 tonnes already standing in those hedges is, in carbon market terms, simply the starting line.

The same logic governs the Woodland Carbon Code, which credits new planting only. The two thirds of a hectare of established wood on this farm is worth nothing under it either.

This is not a flaw. Additionality is the whole point of a credible code, and a scheme that paid farmers for hedges that were already there would be paying for nothing. But it does produce a result that takes some explaining on a farm walk: the better your hedges already are, the less a carbon code has to offer you.

There is a sharper version of this on the farm I surveyed. That H7 hedge is overgrown. Standard advice would be to cut it back towards H6 and tidy it up. Under the Code, trimming it back would count for nothing, while letting it grow out and gapping up the weak sections beside the ash would generate creditable carbon. The agronomy and the carbon accounting point in opposite directions.

Where this goes next

Two things need saying about the Code's current state. First, development is on hold. The Allerton Trust's own project page says the work is paused while further development opportunities and next steps are considered. After an £81,500 grant in 2021, pilots on three arable farms and trials on the Kellogg's Origins farms, the Code is not issuing credits and has no confirmed launch date.

Second, and more interesting to me, is the measurement question. The Code's method asks for hedge length from mapping software and hedge width and height measured in the field, by hand, with a tape. That is accurate and it does not scale. Measuring 41 hedges properly on one 39 hectare farm is a day's work. Doing it across a county is not happening.

Which is exactly why remote measurement matters, and exactly why I wanted to know what a cheap consumer drone could really deliver. The answer, on this evidence, is: the right shape of answer, at nothing like the right precision. ±50% on hedge carbon is fine for a farm advisory conversation about where the carbon sits. It is nowhere near good enough to underwrite a tradable credit.

Closing that gap is not conceptually hard, and that is the encouraging part. An overhead pass at a known height would resolve width properly. A side-on pass with a two metre pole in frame for scale would fix the height reference that the people in shot were standing in for. Both together would bring dimensions to roughly ±10%, which starts to look like something a verifier could work with. The Code's own developers have already trialled LiDAR scanning for verification, so the direction of travel is established.

My hunch is that the future version of a hedgerow code looks less like a surveyor with a tape and more like an annual flight, with the code's role shifting from prescribing how you measure to setting the precision a measurement has to hit. That would let the method improve without the standard having to be rewritten every time the hardware gets better.

For now though, I am writing to a farmer to explain that his 4 km of hedge holds 170 tonnes of carbon, that it is genuinely good for his farm, and that nobody is going to pay him for a single tonne of it.

A note on the numbers

This farm is anonymised. The figures, method and uncertainties are taken directly from the assessment, but the holding, its location and its parcels are not identified. The carbon figures are stocks, not credits, and the hedge total is an upper bound that assumes every hedge resembles the one surveyed. Bulk density was assumed rather than measured, at 1.3 g/cm³, which is the single largest uncertainty in the soil figure and therefore in the farm total.

References

Back to all posts