Distribution Uniformity

No irrigation system waters a field evenly. Distribution uniformity measures how unevenly, by the low-quarter method: lay out catch cans across the field, run the system, sort the catches from lowest to highest, average the driest quarter of them, and divide by the average of all of them. Cans averaging 27 mm overall with the driest quarter averaging 21.6 mm give $21.6/27 = 80\%$.

The low quarter is chosen rather than the minimum or the standard deviation because it is the number a scheduling decision actually turns on. A single freak dry can may be a knocked-over container; the driest quarter is a real part of the field, large enough to matter to the yield and stable enough to be measured again next year. It is also directly actionable: to bring the low quarter up to a target depth, the whole field must be run $1/DU$ times longer than the average alone would suggest. At 80% uniformity, satisfying the dry quarter with 25 mm means applying $25/0.80 = 31$ mm on average, and the extra 6 mm is deep percolation by design.

That relationship is why uniformity is a ceiling on efficiency rather than a separate concern. Application efficiency can be anything from zero up to roughly the distribution uniformity, and no amount of careful scheduling lifts it past that line — the water is already in the wrong places by the time the scheduler sees it. Improving uniformity is therefore the only route to improving efficiency on a system that is already being run sensibly. Well-designed sprinkler systems reach 75 to 85%, centre pivots with correctly sequenced nozzle packages higher, and drip 90% and above.

Poor uniformity usually has a findable cause: worn or mismatched nozzles, pressure variation along a lateral or up a slope, sprinkler spacing too wide for the operating pressure, or wind distorting a pattern that was uniform in still air. Pressure is the one people underestimate, because output varies with the square root of pressure — a lateral losing a third of its pressure end to end delivers about 18% less water at the far end even with identical nozzles, which is exactly the kind of gradient a catch-can test reveals and a walk down the row does not. Run the test in the wind conditions the system normally operates in, not on the calmest morning of the year, or the number will describe a system nobody owns.