Comparing Soil Moisture Data Across Brands: Five Checks That Must Match First
Two systems can both display a percentage and still describe different estimates of soil water. This guide gives five checks to run before calling two series comparable, a practical cross-brand overlap test, and the record structure that keeps a switched fleet auditable.
Disclosure: We may earn a commission if you purchase through links on this page, at no additional cost to you.
Short answer: usually not without an overlap test. Two systems can both display a percentage and still be describing different estimates of soil water. A buyer should first establish the quantity reported, the calibration path, the installation geometry, the data transformation, and the site conditions. If those controls are not aligned, retain separate baselines and compare trends rather than treating the values as interchangeable.
This guide is for farms, greenhouses, and integrators replacing a supplier, expanding into another controller ecosystem, or combining two sensor fleets in one dashboard. It addresses cross-brand comparability. It does not address whether one physical probe can be moved to a new location; that is a separate installation and re-baselining decision.
The decision to make before merging dashboards
A shared dashboard is not evidence that the underlying observations are comparable. Before making one irrigation threshold apply across brands, decide which of the following decisions the data must support: within-zone trend tracking, a simple wet-dry alarm, irrigation scheduling at a defined depth, or a documented comparison between blocks. The higher the consequence of the decision, the stronger the overlap evidence should be.
| Use case | Can brands be mixed? | Minimum evidence before acting |
|---|---|---|
| Within one brand and one established zone | Usually yes | Keep the original calibration, depth, and threshold logic. |
| Trend direction only across brands | Often, with caution | Same depth and irrigation cycle; compare rise and drawdown shape during an overlap period. |
| One common irrigation threshold | Only after validation | Parallel installation, matched logging, paired samples, and a documented conversion or a shared actionable range. |
| Historical yield, water-use, or agronomy comparison | Not by default | Traceable metadata and a stated harmonisation method; otherwise analyse each series separately. |
Why identical percentages can disagree
Most in-ground electronic probes do not directly weigh water. They sense an electrical or electromagnetic property of the soil and convert that signal to an estimate. The conversion may depend on the sensing method, probe geometry, frequency, factory equation, temperature treatment, electrical conductivity, and the soil for which the equation was evaluated. A screen label such as "moisture 0–100%" is therefore not enough to prove that two values share the same meaning.
The first question is whether both systems report volumetric water content, a relative moisture index, raw signal, or another calculated variable. The second is whether the number is produced by the same type of calibration. A value can be precise within one fleet yet offset from another fleet in the same soil. That offset may remain stable, change with wetness, or vary after installation.
Five checks before calling two data series comparable
| Check | What to obtain | Why it matters |
|---|---|---|
| Measured quantity and unit | Protocol register, unit definition, scaling and missing-value code. | A percentage, a decimal VWC value, and a 4–20 mA engineering conversion are not automatically equivalent. |
| Sensing method and model | Model number, sensing principle, probe dimensions and stated sensing volume. | Different geometries sample different soil volumes and respond differently to stones, air gaps, and wetting patterns. |
| Calibration and compensation | Factory equation, soil-specific option, temperature/EC handling, recalibration policy. | The relationship between signal and water content can shift with soil and operating conditions. |
| Installation context | Depth, orientation, soil contact, distance from emitter, tillage and crop zone. | A perfectly matched sensor model still observes a different process if installed in a different place. |
| Data path | Logger scaling, Modbus register map, averaging interval, timezone, reset behaviour. | A dashboard may alter values through scaling, rounding, filtering, or aggregation before export. |
How to run a practical cross-brand overlap test
The objective is not to declare a universal winner. It is to find out whether a new system preserves the management signal you need at this site. Run the comparison over wetting and drying events; a single dry-day reading is weak evidence.
- Freeze the comparison plan. State the soil zone, target depth, logging interval, reference time zone, and the management question to be protected.
- Install paired instruments as close as practical without allowing one probe or its disturbed installation hole to alter the other. Record their exact spacing, depth, orientation, and relation to emitters.
- Collect an overlap period that includes at least several irrigation or rainfall wetting events and subsequent dry-down. Export native timestamps and raw or least-processed readings where available.
- Check data integrity before comparing values. Look for time shifts, register scaling mistakes, flatlines, implausible jumps, communication gaps, and changes after maintenance.
- Plot paired observations and plot change over time. A constant offset may be manageable for some decisions; a changing slope or different response to wetting is a stronger warning that direct substitution is unsafe.
- Decide the operational rule. Options include keeping separate thresholds, applying a documented local conversion, or treating both systems as trend tools while retaining one reference fleet for absolute decisions.
What good comparison evidence looks like
A useful field record connects each comparison point to its context. Keep a small data dictionary that names the device, firmware if known, output register, unit, scaling equation, depth, installation date, soil zone, and every change made to the logger or dashboard. Pair that record with the raw data export. Without this traceability, a later analyst cannot distinguish a real soil response from a configuration change.
When an absolute value is important, add independent reference observations that fit the project's agronomic and budget constraints. The reference method and sampling depth must be stated. The purpose is not to promise laboratory-grade accuracy from a field installation; it is to learn whether the management threshold being carried forward remains defensible in this soil and installation.
Do not force a conversion when the evidence says no
A simple linear conversion is tempting: add a fixed offset or multiply one series by a constant. Use one only if the paired data show that the relationship is stable over the relevant wetness range, depth, season, and operating conditions. A conversion fitted during one irrigation cycle may fail when salinity, temperature, crop water uptake, or soil contact changes. Document its valid range, the date it was derived, and the trigger for rechecking it.
If the relationship is not stable, a more honest operating method is to preserve brand-specific alert bands. For example, define a zone's irrigation action from each system's observed response around known wetting and crop-stress conditions, rather than trying to make two display percentages identical. This protects the decision even when the numbers cannot be made interchangeable.
Keep calibration evidence separate from irrigation rules
Calibration asks how the instrument signal relates to a stated soil-water quantity. Irrigation scheduling asks when to act in a particular crop, root zone, soil, and irrigation layout. They are related, but they are not the same task. A supplier may provide a stated accuracy or a factory relationship; that does not establish the correct irrigation threshold for a new field. Conversely, a locally useful threshold does not prove that two sensor brands report interchangeable volumetric water content.
This distinction helps avoid a common procurement error: asking a new supplier to "match the old threshold" without giving the old system's unit, depth, soil zone, wetting geometry, historical response and action rule. A better request is to ask for the new model's measurement definition and documentation, then conduct a local overlap trial. The farm can then decide whether the old threshold remains valid, needs a locally documented translation, or should be replaced by a new zone-specific rule.
How to preserve a usable historical record after a switch
Do not silently splice the new values onto the end of an old data series. Keep both original series, tag the change date, and preserve the old system metadata. In reports, show the boundary where the sensor system, logger configuration, or calculation changed. Where a relationship has been shown only for a limited period, label it as a local operational conversion rather than rewriting the historical values as if they were directly measured by one instrument.
For dashboards, use fields that retain the source device and measurement method. A practical minimum is device family, exact model, probe depth, zone identifier, unit, scaling rule, calibration note, sampling interval, and a flag for values recorded during the overlap period. This makes later review possible when staff, suppliers, or controllers change, and lets one farm run a mixed fleet without pretending that every trend belongs to one homogeneous measurement population.
Questions to send with an RFQ
- Which physical quantity is reported by the configured model, and what unit/scaling is used at the RS485 or analogue output?
- Which sensing principle and factory calibration are supplied? Is there a soil-specific calibration option or procedure?
- What documentation defines the register map, byte order, resolution, averaging, warm-up, and error states?
- How are temperature and electrical conductivity handled in the stated measurement performance?
- What probe dimensions and installation method are required to avoid air gaps or disturbed-soil bias?
- Can the supplier provide the exact delivered configuration, current manual, calibration statement, and a sample data export before purchase?
Sourcing options for an overlap pilot
The following two listings give a buyer two different supplier paths for a controlled pilot. They are not presented as proof that the models match each other, nor as a recommendation to merge their values without the overlap process above. Marketplace listing identity and stated configuration require confirmation with the supplier before ordering.
System A — BGT RS485 soil moisture, temperature and EC sensor
System B — Renke 4-in-1 soil moisture, temperature, EC and pH sensor
Bottom line
Treat a brand change as a measurement-system change, not just a hardware replacement. Start with a parallel deployment in one representative zone, preserve the raw data and metadata, and judge whether the new system supports the same irrigation action across wetting and dry-down. If it does not, maintain separate local baselines. That is more useful—and more defensible—than producing a shared dashboard number that looks comparable but is not.
Share this article