August 31, 2026
Soil Water Potential vs. Percent Moisture: The Two Numbers That Tell You a Bed Is Actually Dry
Search "soil water potential sensor" and you land on pages that either sell you a probe or lecture you on the physics of tension. Almost none of them answer the question you actually have standing over a bed: what reading means "water now"? There are two different numbers in play — tension in centibars and volumetric percent — and they measure soil water in genuinely different ways. Here's what each one tells you, and the honest threshold that should trigger a watering.
Two numbers, two different questions
Capacitive probes — the kind in most home weather-station ecosystems — report volumetric water content, a percentage of how much of the soil is water. It answers "how much water is in the ground?" Tensiometers and Watermark-style sensors report soil water potential in centibars (cb) of tension — how hard a root has to pull to get that water free. That answers a slightly different, more plant-centric question: "how much effort does a root have to spend?"
The distinction matters because the same soil can read "20% moisture" and mean very different things in sand versus clay. Tension normalizes for that — a plant in clay and a plant in loam both start to struggle at similar centibar values even though their percentages differ. Neither number is "right." Percent is cheaper to measure and intuitive; tension maps more directly to plant stress. What both share is that a single reading is a poor trigger.
The threshold nobody prints on the box
The SERP is full of tension charts — "0–10 cb saturated, 10–30 cb field capacity, 30–60 cb time to water" — and vendor moisture tables that imply a magic percentage. Treat those as orientation, not gospel; your soil sets the real line. Here's the number Siltwell actually acts on, and it's deliberately conservative:
- Capacitive percent: a bed is flagged dry when its two most recent readings are both below the threshold — default 25%, which you tune per bed.
- Watermark tension: a bed is flagged dry when the two most recent readings are both above 60 cb.
The 25% and 60 cb figures are starting points, not laws of nature. A fast-draining sandy bed and a moisture-holding clay bed will cross those lines at different real-world dampness, which is exactly why the per-bed threshold is adjustable. You watch how a bed behaves for a couple of weeks, then nudge the number so "alert" lines up with "the soil my plants actually want water at."
Why "two readings, not one" is the whole trick
Soil sensors spike. A probe can catch a splash from the hose, a pocket of air after cultivation, or just electrical noise, and throw a single reading that looks alarmingly dry (or wet) for one cycle. If you trigger on any one reading, you get nagged constantly and stop trusting the alerts — the fastest way to end up ignoring a sensor entirely.
Siltwell's soil job runs every 10 minutes but only fires a "this bed is low" warning and push when the two most recent readings are both past the line. One odd blip won't do it; a genuine drying trend will. That's the piece the physics-heavy articles skip — they'll explain what 60 cb means but never tell you how to keep a noisy sensor from crying wolf.
One honest boundary worth stating plainly: this alerts you which bed is dry — it does not run your irrigation. Siltwell tells you "water this bed, skip that one"; the hose is still yours to turn on. If you want the per-bed reading walkthrough, the per-bed probe guide covers mapping a sensor to a bed, and meter vs. bed sensor compares the jab-in dial to a logging probe.
Which sensor to actually buy
For most gardeners, a logging capacitive probe is the practical choice: cheaper than a tensiometer, no reservoir to refill, and it feeds a continuous trend rather than a one-off dial reading. Siltwell works with a specific set:
- Ecowitt WH51 — capacitive percent, best value, pairs to any Ecowitt gateway or Wittboy console.
- Davis 6345 — for WeatherLink setups.
- Ambient WH31SM — needs a WS-2000 or WS-5000 console.
If you're on Tempest, note it has no soil sensor of its own — you'd add an Ecowitt WH51 alongside it. AcuRite and La Crosse aren't supported here, since they lack a soil probe and an open API.
Mapping a soil sensor to a bed and getting per-bed dry alerts is part of Siltwell Pro ($22/year, with a 14-day free trial). The free tier still lets you log watering by hand and gives forecast-based, garden-level guidance. Gear we mention is linked with disclosed affiliate links on the setup page — here's how to connect a station and pick a probe.
One honest caveat: a sensor narrows the guesswork, it doesn't repeal the weather. A probe reads one spot at one depth in one bed — not the whole bed, and not the roots two feet over. Tension and percent both help you stop the obvious mistakes, like watering ground that's still damp or letting a fast-draining bed dry out unseen. Use the two-reading rule to earn your trust in the number, then let your own eyes handle the rest.
Common questions
- What should your soil moisture percentage be?
- There's no single correct number, because it depends on your soil, your plants, and the sensor type. As a working starting point for a capacitive probe, Siltwell treats roughly 25% volumetric moisture as the "getting dry" line, then lets you tune that threshold per bed once you've watched how that soil actually behaves. A sandy bed and a clay-heavy one will sit at very different percentages while feeling equally "ready to water."
- How do I check soil moisture?
- The finger test works — push a finger a couple of inches down and feel for dampness — but it's late and easy to second-guess. A buried sensor turns the guess into a number: capacitive probes report volumetric percent, while tensiometers and Watermark sensors report tension in centibars (how hard roots must pull). Reading a few inches down, at root depth, beats judging by the dry surface.
- Do soil moisture meters actually work?
- The cheap two-prong "meters" you jab in and read on a dial are unreliable — they drift and mostly react to salts. Buried capacitive probes and tension sensors that log continuously are far more trustworthy, because you're watching a trend over time rather than one instant reading. The honest move is to trust the direction and repeated readings, not a single number.
- What is the most accurate soil moisture meter?
- For home gardens, a logging capacitive probe or a tension sensor beats a handheld dial meter, because it reads the same spot repeatedly and shows a trend. Siltwell works with the Ecowitt WH51 (capacitive percent, best value), the Davis 6345, and the Ambient WH31SM. Accuracy also comes from how you use the reading — two consistent readings past your threshold mean more than one spike.
Related guides
What a Soil Moisture Probe Actually Tells You — and How to Turn One Reading Into "Water This Bed"
A soil moisture probe gives you one number, at one spot, at one moment. Here's the threshold that should trigger watering — and how to log it per bed.
Soil Moisture Meter vs. a Bed Sensor: Which Bed Actually Needs Water?
A stick-in moisture meter gives one reading, one spot, one moment. Here's how a mapped soil probe watches every bed continuously and tells you which one is dry.
Waterproof Soil Moisture Sensors: What a Probe Buried in the Bed Year-Round Actually Decides
Buyers want a soil probe they can leave buried, not a handheld spot-check. Here's what a continuously buried reading actually tells you to do — bed by bed.
Free planting tools
- When to plant — the last date to sow each crop before your fall frost
- Microclimate checker — how much colder your yard runs than the forecast
- GDD estimator — whether a crop will ripen where you are
