Guide · 6 min read
DU vs CU: understanding irrigation uniformity
Two statistics dominate irrigation audit reports: low-quarter distribution uniformity, written DU or DUlq, and Christiansen's coefficient of uniformity, written CU. Both are computed from the same catch-can readings and both are reported as percentages, but they answer different questions and they are not interchangeable.
DU asks: how badly does the driest quarter of the zone do compared with the average? CU asks: how far, on average, does any reading sit from the mean? A zone can carry a respectable CU while its driest corner is starving, which is exactly why scheduling practice leans on DU.
How DU is computed
Sort the catch readings, take the lowest quarter of them (the count is the whole number part of n divided by 4, with a minimum of one reading), average that low quarter, and divide by the average of all readings. Multiply by 100 for a percentage.
Worked example: readings of 0.10, 0.20, 0.30, and 0.40 inches average 0.25. The lowest quarter of four readings is one reading, 0.10. DU is 0.10 divided by 0.25, which is 40 percent. The driest quarter of this zone receives well under half the average application.
How CU is computed
Christiansen's CU is 100 times one minus the sum of absolute deviations from the mean divided by the sum of all readings. Every reading contributes, so one very dry can moves CU far less than it moves DU.
Same example: deviations from the 0.25 mean are 0.15, 0.05, 0.05, and 0.15, summing to 0.40. The readings sum to 1.00. CU is 100 times (1 minus 0.40 divided by 1.00), which is 60 percent. Note the spread: the same test scores 40 DU and 60 CU.
Which one to schedule from
Schedule from DU. The point of a runtime is to keep the driest meaningful fraction of the zone healthy, and DU is built from exactly that fraction. Dividing the plant water requirement by the precipitation rate derated by DU gives a runtime that waters the low quarter adequately instead of the average square foot.
CU remains useful as a cross-check and for comparing tests over time, and much older agricultural literature reports CU alone. When both are available, a large gap between them signals a distribution with a concentrated dry area rather than general raggedness.
What counts as a good DU
Achievable uniformity depends on hardware. Well-maintained rotors and rotary nozzles distribute more evenly than fixed sprays, so the bands commonly cited in Irrigation Association auditor training sit higher for them. Cut points vary by program, rebate scheme, and jurisdiction; treat any single threshold as guidance rather than a specification.
The reference tables on this site list the bands Irrigaudit uses for grading, together with that caveat.
Frequently asked questions
What is a good DU for spray heads versus rotors?
Guidance commonly cited in Irrigation Association auditor training treats roughly 65 to 75 percent as achievable for well-maintained fixed sprays and roughly 70 to 80 percent for rotors and rotary nozzles, with higher figures being excellent. Exact targets vary by program, so check any rebate or ordinance requirement you are auditing against.
Why is my CU higher than my DU?
CU averages deviations across every reading while DU compares only the driest quarter against the mean, so a zone with a concentrated dry area scores much lower on DU than on CU. CU is almost always the higher of the two on the same data.
Should I schedule runtimes from DU or CU?
From DU. A runtime computed from the precipitation rate derated by DU waters the driest quarter adequately. A runtime based on CU or the raw average leaves the low quarter chronically short.
Can DU exceed 100 percent?
No. The low-quarter average can never exceed the overall average, so DU tops out at 100 percent, which would mean perfectly even application across every can.