Glossary
CV (Coefficient of Variation) in Spreading
A single percentage that scores how evenly a spreader lays product across its swath — the standard deviation of the catch weights divided by their average. Lower is more even.
Definition
Coefficient of variation is the statistic that turns a pan test into a number you can act on. Lay a row of collection pans across the flight line, fly the pass, weigh what lands in each pan, then divide the standard deviation of those weights by their mean and express it as a percentage. A CV near zero means every pan caught about the same amount and the pattern is flat; a high CV means some pans are heavy and others nearly empty, so the product is landing in bands rather than a sheet. The number is always tied to the conditions it was measured under — a specific product, rate, flight height, speed, and disc setting — so a CV is a result for that setup, not a permanent property of the machine. Spreading references generally treat a CV under roughly 15% as acceptable for granular fertilizer and under about 10% as good, with values climbing past the low twenties being where uneven application starts showing up in the field as visible striping. The pattern shape matters alongside the number: a symmetric curve that tapers at the edges can be overlapped into an even sheet, while a lopsided or twin-peaked pattern will stripe no matter how the passes are spaced.
In an Alberta Context
On an Alberta granular job the CV is what justifies the pass spacing. Blends behave differently from a single straight product because particles of different size and density throw different distances, and a spreader that tests well on urea can test poorly on a coarse blend or a light forage seed. That is why the pan test is repeated when the product, the rate, or the flight height changes rather than carried over from last season. UAV AG sets the effective swath from the tested pattern, which is normally narrower than the raw throw width, so the overlap between passes fills the taper at the edges. Granular fertilizer and seed spreading is the everyday legal drone service in Canada today, and CV is the evenness measure that applies to it directly.
Why It Matters
A poor CV does not usually announce itself on the day of application — it shows up weeks later as green and pale strips running the length of the field, because the strips that got a light rate ran short of nutrient while the heavy strips got product that was paid for and wasted. Two passes can cover the same acres at the same average rate and produce very different crops if one had an even pattern and the other did not. Measuring CV converts "it looked fine from the truck" into evidence, and it is the number that decides how far apart the passes should be flown.
Frequently Asked Questions
+How is CV calculated in a spreading pattern test?
Collect the product in a row of pans laid across the flight line, weigh each pan, then divide the standard deviation of those weights by their average and multiply by 100. The result is a percentage: the spread of the catch weights relative to the typical catch weight.
+What is a good CV for granular spreading?
Spreading references generally treat a CV under roughly 15% as acceptable for granular fertilizer and under about 10% as good. Once the number climbs past the low twenties, uneven application tends to show up in the field as visible striping. Seed and high-value products are usually held to the tighter end.
+Does a low CV mean the pass spacing is correct?
Not on its own. CV scores the evenness of the pattern that was tested at a given pass spacing, so it is the pairing of the two that matters. The pattern shape decides how much overlap is needed — a curve that tapers at the edges overlaps into an even sheet, while a lopsided or twin-peaked pattern will stripe regardless of spacing.
+How often does the CV need to be re-tested?
Whenever the setup changes. A different product, a different target rate, a different flight height, or a change in disc or gate setting all move the deposition curve, so the tested CV no longer describes what the machine is doing. A blend can also test very differently from the straight product it is based on.