September 25, 2026

What a soil water content sensor reads

A soil water content sensor reports the share of your soil's volume that is water, written as a percentage. Most garden probes read this with a capacitive method. They sense how wet soil changes an electrical field, and water changes it far more than dry soil or air. A reading of 40% means water fills roughly 40% of the soil volume where the probe tip sits.

What the sensor measures

The number is volumetric water content. It counts water as a fraction of total soil volume, alongside the mineral grains and the air pockets. A capacitive probe estimates it by measuring the soil's dielectric response, which climbs steeply as pores fill with water. The probe then converts that response to a percentage on a fixed curve.

That curve is generic. It does not know your exact soil, so the percentage is an estimate, close enough to track drying day to day. For a home bed that resolution does the job. You watch the number fall and rise, and you learn where your plants start to struggle.

Volumetric water content and soil tension

Volumetric water content and soil tension measure two different things. Water content tells you how much water is present. Tension tells you how hard roots must pull to draw it out, measured in centibars (cb) by a watermark-style sensor. A sandy soil can hold a low water content yet release it easily, while a clay soil can hold more water and still grip it tightly.

Both readings point at the same decision from opposite directions. As a bed dries, water content drops and tension rises. Gardeners conflate the two because both get called "soil moisture", yet a 25% capacitive reading and a 60 cb tension reading come from separate instruments. If tension is the reading you trust, the soil water potential guide covers when to water by centibars.

What our beds read

In our own garden four beds carried capacitive probes from mid-July to mid-September. Raised bed 3 ranged between 32% and 62% over those weeks. The ground bed hit that same span, 32% to 62%, on other dates. Raised bed 1 kept a tighter 39% to 54%. Raised bed 2 landed between 36% and 57%.

None of them dropped below 32% in that window, even through a drier early September. That tells us the beds held water well after a wet July of 10.89 inches of rain. It also tells us one fixed percentage would flag these beds differently, since their soil and drainage differ bed to bed.

The reading that flags a bed as dry

A bed counts as dry once its two latest readings both sit under the threshold you set. The default is 25% for a capacitive probe, and you can set a different threshold per bed. For a watermark-style tension sensor, the trigger is both recent readings above 60 cb. Requiring two readings in a row keeps a single odd sample from firing a warning.

When that condition holds, the bed sends a low-moisture warning and a push. The push names the bed that needs water so you can skip the rest. The warning reports soil state only. You still turn the water on yourself.

The watermark path suits gardeners who already think in tension, and the capacitive path suits anyone who wants a plain percentage. Both feed the same dry-bed logic. If you are weighing a handheld meter against a fixed bed probe, that comparison covers what each one can and cannot track.

A good starting threshold depends on your soil and your plants. Our beds sat well above the 25% default all summer, so we would raise the threshold to match what our plants show, rather than wait for a number they may never hit. Set a per-bed baseline from a week of readings, then adjust it when a plant wilts before the alert.

To turn a raw percentage into an action, log a week of readings per bed and mark the level where growth slows. Our per-bed threshold guide walks through setting that baseline, and the measurement guide covers reading depth and timing.