Technology & Devices
Why We Measure Soil Moisture at Two Depths, Not One
A single probe near the surface gives an accurate reading of a place that does not matter. The second probe, at the base of the root zone, is what tells you whether you are over-irrigating.
Published July 20, 2026 6 min read
Most smart irrigation systems on offer install one moisture probe at 20 to 30 centimetres and take the valve decision from it. The number that probe reports is entirely accurate. The problem is not its precision - it is that it answers a different question from the one being asked.
Surface moisture moves fast: it evaporates within hours of an irrigation event, rises immediately with any light application, and responds to canopy shading and daytime heat. Drive the irrigation decision from that reading alone and you get a recognisable pattern - short, frequent events that keep the surface damp while the base of the root zone dries out progressively. Roots stay shallow and the crop becomes far more sensitive to any interruption.
The second probe, at 45 to 60 centimetres depending on rooting depth, answers the other question: did the water pass below the root zone? A rise at that depth two to four hours after the event ends is direct evidence of deep percolation - the loss nobody sees, because it happens underground and leaves no mark on the surface.
Together the two readings give a simple, usable operating rule: start irrigating when surface moisture falls below the crop's stress threshold, and stop when the deep probe begins to respond. That rule alone removes most over-irrigation, and it needs two probes rather than ten.
The lag between the two readings adds a third piece of information at no cost: how fast the wetting front travels describes the soil itself. In sand the deep probe responds within minutes; in heavy clay it may take hours. That is what sets event duration and the number of pulses per day, and it is a local property of each plot that cannot be copied from a catalogue.
One installation note that matters as much as the theory: both probes must sit in the same vertical column and at the same distance from an emitter, the cable must be buried carefully, and the position must be marked visibly above ground. The most common cause of bad data we encounter is not a failed sensor - it is a probe installed half a metre off the emitter line, then displaced during weeding.