A granular matrix sensor, of which the Watermark is the best-known example, is an inexpensive buried probe that tells a grower how hard a crop's roots have to work to pull water from the soil. Rather than measuring how much water the soil holds, it measures soil water tension, the suction the soil exerts on that water, by reading the electrical resistance of a porous granular material inside the sensor. This guide explains how that resistance relates to tension, weighs the granular matrix sensor's trade-offs against capacitance probes, and describes how its raw resistance signal is conditioned so a datalogger or telemetry RTU can turn it into a usable reading.
Granular Matrix Sensor in one line: A granular matrix sensor is a soil moisture sensor that measures soil water tension by reading the electrical resistance of a granular filler material that exchanges water with the surrounding soil. As the soil dries and pulls water out of the sensor, the granular material's resistance rises, and as the soil wets, water re-enters and resistance falls, so resistance maps to how tightly the soil holds water. It is a low-cost, robust alternative to capacitance probes, with the trade-offs of slower response and a need for temperature correction, and its resistance signal must be conditioned before a datalogger or RTU can read it.
Soil moisture can be described two ways, and a granular matrix sensor measures the one that matters most to a plant. Volumetric water content says how much water is in a given volume of soil, while soil water tension, or matric potential, says how strongly the soil holds onto that water and therefore how hard roots must pull to extract it. A plant does not experience content directly; it experiences tension, so a sensor that reads tension speaks the language of crop stress. A granular matrix sensor is a tension sensor: its readings, usually in units of pressure such as centibars or kilopascals, rise as the soil dries and the water becomes harder to remove.
Inside the sensor is a porous granular material, historically a gypsum block and in the Watermark a granular matrix around electrodes, held in intimate contact with the soil through a porous membrane. Water moves freely between this internal material and the surrounding soil until the two reach the same tension. When the soil is wet and holds water loosely, the internal material stays saturated; when the soil dries and pulls water out, water leaves the internal material until its water content matches the soil's suction. The sensor thus continuously equilibrates with the soil's tension, and the state of its internal material reflects that tension at any moment.
The reading itself comes from electrical resistance. Two electrodes embedded in the granular material measure the resistance of the water held between the grains, and that resistance depends on how much water is present: a wet, well-connected film of water conducts easily and resistance is low, while a dry material with little water conducts poorly and resistance is high. Because the internal water content tracks soil tension, the measured resistance tracks tension too, and a calibration relates resistance to tension in engineering units. Reading tension therefore reduces to reading an electrical resistance across the buried sensor.
The granular matrix sensor's great virtue is that it is cheap and rugged. Each sensor costs little, so a grower can install several at different depths and locations to build a picture of the whole root zone without a large outlay, and the sensors have no delicate electronics buried in the soil to fail. Because they read tension directly, they are also relatively insensitive to soil type in the way that matters for scheduling, since tension relates fairly consistently to plant stress across soils, whereas a content reading has to be interpreted against each soil's water-holding characteristics. For low-cost, tension-based scheduling spread across many points, granular matrix sensors are hard to beat on economics.
Capacitance probes take the opposite position on the trade-off. They measure volumetric water content electronically by sensing the soil's dielectric property, respond quickly to changes in moisture, and can read continuously at fine time resolution and often at multiple depths in a single probe. That speed and resolution make them well suited to watching moisture move through the profile in near real time. The cost is that they are more expensive, require careful installation for good soil contact, and generally need soil-specific calibration to convert their dielectric reading into a content the grower can trust. Where the granular matrix sensor is cheap, slow, and tension-based, the capacitance probe is costlier, fast, and content-based.
The most practical differences for a grower are response speed and temperature. A granular matrix sensor responds more slowly than a capacitance probe, because water must physically move in and out of its granular material to reach equilibrium, so it lags rapid changes and reads best as a measure of the settled tension rather than of fast transients. It is also sensitive to temperature, because the resistance of water changes with temperature, so a reading taken without accounting for soil temperature can be misleading. Neither issue is fatal, but both mean the sensor's raw output needs handling with care, which is where signal conditioning comes in.
A granular matrix sensor does not output a tidy number; it presents a resistance that has to be measured and converted, and the measurement itself must be done carefully. If a steady direct current were passed through the buried electrodes to read resistance, it would drive ions through the moist material and gradually polarize and degrade the sensor. To avoid this, the reading circuit excites the sensor with a brief alternating or reversing current so that no net current accumulates, measures the resulting resistance, and does so only momentarily rather than continuously. Getting this excitation right is essential to a sensor that lasts and reads consistently over seasons in the ground.
Temperature correction is the next necessary step. Because water's resistance falls as it warms, a given soil tension will read as a different raw resistance at a cold morning temperature than at a hot afternoon one, so an uncorrected sensor would appear to change tension when only the temperature changed. A companion soil temperature measurement lets the conditioning electronics or the datalogger apply a correction, adjusting the resistance-to-tension conversion for the current soil temperature so the reported tension reflects real soil dryness rather than the time of day. Without this correction, a granular matrix sensor's readings drift with the daily temperature cycle and mislead scheduling.
In the field these sensors connect to a datalogger or a telemetry remote terminal unit that performs the excitation, the resistance measurement, the temperature correction, and the conversion to tension, then reports the finished reading. From there the data enters the monitoring layer, and this is where cloud SCADA brings the readings together, applying to agriculture the same remote data collection platforms like Merobix provide across oil and gas and other industries. RTUs at points scattered across fields push their conditioned tension values, along with soil temperature and diagnostics, to a central dashboard where a grower sees the tension trend at each depth and location, gets alerts when a zone reaches a stress threshold, and can spot a sensor reading erratically. The humble buried resistor thus becomes part of a live, operation-wide view of soil moisture, its raw signal turned by conditioning and telemetry into scheduling information a grower can act on from anywhere.
It measures soil water tension, the suction the soil exerts on its water, rather than the volume of water the soil holds. It does this by reading the electrical resistance of a porous granular material inside the sensor that exchanges water with the surrounding soil until the two reach the same tension. As the soil dries and pulls water out of the material, its resistance rises, and a calibration converts that resistance into tension in units such as centibars or kilopascals.
A granular matrix sensor is inexpensive, rugged, and reads tension, which relates fairly directly to plant stress, but it responds slowly and needs temperature correction. A capacitance probe is more expensive and measures volumetric water content electronically, responding quickly and often at multiple depths but usually needing soil-specific calibration and careful installation. In short, granular matrix sensors trade speed and resolution for low cost and simplicity, while capacitance probes trade cost for fast, high-resolution content readings.
The sensor reads soil tension through the electrical resistance of water in its granular material, but the resistance of water changes with temperature. Without correction, the same soil dryness would read as a different resistance in the cool morning than in the warm afternoon, making the sensor appear to change when only temperature changed. A companion soil temperature reading lets the datalogger or conditioning electronics adjust the resistance-to-tension conversion so the reported tension reflects real dryness rather than the daily temperature cycle.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.