ResearchTechnology and Digital PracticeIntegration with IoT and Smart City Systems

Smart irrigation controllers and soil moisture sensors

What weather-based controllers and soil moisture sensors do, what WaterSense and California's MWELO say, measured savings and limits, and data questions.

Field Method · AI-assisted research roundupPublished Sources checked

A smart irrigation controller changes or withholds scheduled watering using weather data or a soil moisture reading. A clock timer repeats the schedule it was given. In the evidence we read, savings were largest where sites had been over-watered, and use rose at many sites that had been watering less than the theoretical requirement. The label and the regulation describe what the device must be able to do. Neither measures what it saves on a given site.

Sources were read on November 15, 2026. This is not legal or engineering advice.

Why controls get attention

EPA states that residential outdoor water use in the United States is nearly 8 billion gallons a day, mainly for landscape irrigation, and reports an expert estimate that as much as 50 percent is wasted through overwatering. EPA WaterSense, Labeled Controllers University of Florida extension authors cite end-use research finding that homes with in-ground automatic systems applied 47% more water than homes without, largely from timers set once and left. UF/IFAS AE446

Two kinds of control

Weather-based controllers. These are also called evapotranspiration (ET) controllers. ET is water lost from soil and plant surfaces to the air. The controller estimates water used since the last irrigation and replaces all or part of it. One University of Florida publication groups them in three types:

  • Signal-based. Weather data are collected from public sources or station networks and an ET value is sent to controllers by wireless communication. Wi-Fi models with smartphone apps are a newer trend.
  • Historical ET. A stored crop water use curve for the region, derived from historical averages and sometimes modified by a temperature or solar radiation sensor.
  • On-site weather measurement. Weather measured at the controller is used to calculate ET continuously.

UF/IFAS AE442 A companion publication adds a fourth type: add-on devices that work with an existing timer and allow or block events from a soil water balance. It notes that a weather station may not reflect the site where rainfall is localized, and that signal-based models fall back to historical averages or the last value when the signal is lost. Historical controllers may respond poorly to unusual weather. Many on-site controllers use simplified ET methods, some the temperature-based Hargreaves equation, which may be less accurate across diverse climates. UF/IFAS AE446

Soil moisture sensors. Most commercial sensors estimate volumetric water content from the dielectric constant of the soil, which rises with water content. In the common bypass arrangement the sensor system is an add-on to a timer. It skips a scheduled cycle when the reading is wetter than a set threshold. An on-demand system has a low and a high threshold and starts and stops irrigation itself. UF/IFAS AE437 A bypass sensor only allows or blocks programmed events, so the base schedule still matters. UF/IFAS AE442

The Irrigation Association’s Smart Water Application Technologies (SWAT) testing program lists categories including weather-based and soil moisture-based controllers, rain sensors and flow sensors. Irrigation Association, SWAT

What the WaterSense label requires

WaterSense is a voluntary EPA label. Two product specifications apply.

Weather-based controllers: Version 1.1, September 2, 2021. Version 1.0 was released on November 3, 2011. EPA describes Version 1.1 as a minor revision that references ANSI/ASABE S627, and says the test method is based on the SWAT protocol as modified by EPA. EPA WaterSense, Weather-Based Irrigation Controllers It covers stand-alone controllers and add-on devices that use current weather data. A rain device or a soil moisture sensor used alone does not qualify under this specification. The performance test runs for 30 consecutive days across six test zones:

  • Irrigation adequacy must be at least 80 percent for each zone.
  • Irrigation excess must be no more than 10 percent for each zone, and no more than 5 percent averaged across the six zones.

The controller must also keep its settings through a power loss without a battery, tell the user when signal or sensor input is lost, connect to a rain or soil moisture sensor, accommodate watering restrictions, include percent adjust, fall back to historical data or percent adjust if its weather source is lost, and return to weather-based mode after a manual test. Packaging may not encourage use in standard mode. A licensed certifying body certifies conformance. WaterSense WBIC Specification v1.1

S627 also has an optional second test method with a longer period. EPA kept the first and states that it has no test data from the second. WaterSense Performance Overview, WBIC

Soil moisture-based controllers: Version 1.0, February 2021. EPA released the final specification on February 11, 2021. EPA WaterSense, Soil Moisture-Based Irrigation Controllers It covers bypass-type products sold with an interface device. It excludes on-demand sensors, sensor mechanisms sold alone and products meant only for agriculture. Testing follows ANSI/ASABE S633 with EPA’s changes: three replicate units, three water depletion levels (20, 40 and 60 percent), one moderately coarse test medium, and water with an electrical conductivity of 3 dS/m. To pass:

  • Every unit must enable and disable irrigation at each depletion level.
  • The relative average deviation among replicates must be 10 percent or less.
  • Readings must change with soil moisture (a regression slope of at least 0.01 in absolute value).
  • Every unit must still work after three days in a freezer.

The supplemental capabilities mirror the controller specification, including an indication when sensor input is lost and reversion to percent adjust. WaterSense SMS Specification v1.0

Our interpretation: both labels certify capability under a defined test. Neither is a field savings test, and EPA’s own supporting statement says savings depend on design, installation, maintenance, programming and monitoring. WaterSense SMS Supporting Statement

What California’s model ordinance says

The Model Water Efficient Landscape Ordinance (MWELO) is in Title 23 of the California Code of Regulations, Division 2, Chapter 2.7. DWR states that all local agencies must adopt, implement and enforce the MWELO or a local ordinance at least as effective, and that local ordinances are usually more stringent. The current amendment was approved by the Office of Administrative Law and took effect on January 2, 2025. DWR, MWELO Check the adopted local ordinance; the wording below is the state model only. Our article on tracking regulatory updates lists where California rules are published, and our article on site-scale strategies for heat, stormwater and drought covers MWELO applicability and documentation.

The following is from DWR’s PDF of the amended text. DWR, MWELO Regulations and Appendices

Scope. The model ordinance applies to new construction with a landscape area of 500 square feet or more and rehabilitated landscapes of 2,500 square feet or more, where a building or landscape permit, plan check or design review is required. New construction between 500 and 2,500 square feet may use a prescriptive option. Projects of 2,500 square feet or more use the performance option (sections 491 and 491.2).

Definitions. An automatic irrigation controller can self-adjust and schedule irrigation using ET (weather-based) or soil moisture (sensor-based) data. A soil moisture sensor signals the controller to interrupt or initiate irrigation; a rain sensor interrupts it when it rains. A flow sensor at the supply point must be connected to a compatible controller or flow monitor (section 490.2).

Prescriptive option (section 492). Automatic irrigation controllers “are required.” They must use ET (weather-based) or soil moisture (sensor-based) data “and utilize a rain sensor.” Controllers “shall use non-volatile memory.”

Performance option (section 493.2.2). Controllers using ET or soil moisture sensor data, with non-volatile memory, “shall be required for scheduling irrigation events.” Sensors for rain, freeze, wind and similar conditions “shall be required on all irrigation systems, as appropriate for local climatic conditions.” The next sentence is advisory: irrigation “should be avoided” in wind, freezing weather or rain. Flow sensors that detect high flow from damage or malfunction “are required” for all non-residential landscapes and for residential landscapes of 5,000 square feet or larger. The irrigation design plan must show the location, type and size of controllers, moisture sensing devices and rain sensors.

Scheduling and handover (sections 493.4 to 493.6 and 494). Scheduling “shall be regulated by automatic irrigation controllers” using current reference ET data, with CIMIS given as the example, or soil moisture sensor data. Total annual applied water must not exceed the Maximum Applied Water Allowance. The parameters used to set the controller are submitted with the Certificate of Completion for the establishment period, the established landscape and temporarily irrigated areas. For the performance option, the Certificate of Completion Package requires a diagram of the irrigation plan showing hydrozones to be kept with the controller. A maintenance schedule is submitted, and irrigation audits are done by a certified landscape irrigation auditor who did not design or install the landscape.

The model ordinance names WaterSense only in its definitions of a certified landscape irrigation auditor and an irrigation audit (section 490.2). Its controller and sensor provisions state the required function, not a product label. Local agencies and rebate programs may set their own product conditions.

What the savings evidence shows

EPA’s planning figures. EPA’s general controller page says replacing a clock-based controller with a labelled controller can save an average home up to 15,000 gallons a year, and gives a national figure conditional on every home with an automatic system installing and properly operating one. EPA WaterSense, Labeled Controllers Its product pages give nearly 7,600 gallons for a weather-based controller EPA WaterSense, Weather-Based Irrigation Controllers and more than 15,000 gallons for a soil moisture sensor. EPA WaterSense, Soil Moisture-Based Irrigation Controllers For soil moisture sensors the arithmetic is published: average household outdoor use of 50,500 gallons a year multiplied by a 30 percent savings factor gives 15,150 gallons. EPA reports that field and plot studies showed savings from 30 to 83 percent, with a weighted average of 49 percent. It chose 30 percent because most of the studies were in Florida, where rain is frequent, and because several were controlled plot studies that likely inflate savings. WaterSense SMS Supporting Statement

University of Florida plots and homes. The researchers summarise their own studies. Soil moisture sensors on field plots saved 35% to 54% in dry weather and up to 70% to 90% under normal Florida rainfall. Reclaimed-water homes retrofitted with sensors applied 44% less than monitored-only homes over 32 months. ET controllers on plots ranged from 14% more water applied to a 68% reduction, compared with treatments without sensor feedback. In homes selected for excessive irrigation, properly installed and programmed ET controllers reduced water use by 21% to 31%. The authors state that results vary with brand, weather and site. UF/IFAS AE442

California billing data, 2009. The largest California evaluation we found compared a year of billing data before and after installation for 3,112 controllers at 2,294 sites, weather-normalized. Of those sites, 1,987 were single-family residential, 296 multi-family, commercial or other non-residential and 11 irrigation-only. About 60% of sites were self-installed and 40% professionally or utility installed. MWD and EBMUD, 2009

  • Average outdoor use fell by 47.3 thousand gallons per site, or 6.1%, which was statistically significant at the 95% level.
  • Southern California sites fell by 5.6% (significant). Northern California sites fell by 6.8% (significant at 90%, not at 95%).
  • 56.7% of sites had a significant decrease in application ratio and 41.8% had a significant increase.
  • Sites that increased had a median prior application of 95% of the theoretical requirement. Sites that decreased had a median of 137%.
  • For 384 sites with three years of data, use rose 6% in year 1, then fell 7.8% in year 2 and 16.4% in year 3 against the pre-installation year. More than 90% of those controllers were one product family (HydroPoint/Irritrol/Toro), so the report says this result largely reflects that technology.

The report’s own explanation is that a controller designed to meet the theoretical requirement will raise use where the irrigator applied less than that before. It also found residential sites more likely than non-residential sites to increase application after installation.

Our interpretation: the Florida plot results and the California program results measure different things. A plot study compares a device with a fixed schedule under frequent rain. A program evaluation measures a mixed population, including sites that were under-watering. The 2009 study also predates both current specifications and Wi-Fi controllers. For a project, the prior or assumed baseline schedule can matter as much as the device type.

Upkeep

The University of Florida guidance for bypass sensors says to bury the sensor in the root zone in representative soil, in the driest zone if it serves several, about three inches deep for turfgrass, and at least 5 feet from sprinkler heads and impervious surfaces. Automatic calibration starts by saturating the soil and reading it 24 hours later; the setpoint then falls between 50% and 75% of field capacity depending on brand. The authors advise calibrating after lawn establishment, which they put at 30 to 60 days. UF/IFAS AE437 Follow the manufacturer’s instructions for the product specified; these points are from one extension program in sandy Florida soils and do not cover clay soils or drip zones.

EPA lists three conditions for sensors to reach intended savings: a properly designed, installed and maintained system, proper programming, and monitoring of water use after installation. WaterSense SMS Supporting Statement The MWELO maintenance schedule includes routine inspection, auditing, adjustment and repair of the irrigation system and its components. DWR, MWELO Regulations and Appendices

Data, connectivity and privacy

NIST’s consumer profile describes an IoT product as the device plus components such as a companion app and a backend cloud service, and notes that those components have access to the device and its data. It recommends capabilities that include asset identification, data protection, interface access control and software update. The profile is a set of recommendations. NIST IR 8425

One maker’s documentation shows what this means in practice. We use Hunter because its documentation was openly readable, not on merit. Its Hydrawise settings page says firmware updates happen automatically when the controller is connected, that an optional Offline Mode runs static watering times from the most recently synchronized schedule, and that an offline adjustment applies when the controller has been disconnected for an extended period. Hunter, Hydrawise Controller Settings (maker documentation) Its Hydrawise security statement says traffic between the controller and the service is encrypted, that no user-identifiable information passes between them, and that no home-network details are sent to the service. Hunter, Hydrawise Security Statement (maker documentation) Neither page says what controller and usage data are kept or for how long. For any maker, read the current privacy notice and data terms. This is not a product assessment.

Our interpretation, as questions to settle before specifying a connected controller:

  • What does the controller do when the network or service is down, and does that suit the planting?
  • Who holds the account after handover: owner, maintenance contractor or installer?
  • What data leaves the site (address, schedules, flow records), who can see it, and how long is it kept?
  • How are firmware updates delivered, and for how long does the maker commit to them?
  • If the flow sensor and controller also serve as a privately owned submeter, as the MWELO definition allows, who can export the record?

What we could not settle

  • Current California field savings. We found no California evaluation of Wi-Fi era controllers with a sample comparable to the 2009 study.
  • Soil moisture sensors outside Florida. We found no primary California or arid-climate home study.
  • The standards themselves. ANSI/ASABE S627 and S633 were not read. The SWAT protocol pages did not display content.
  • Large landscapes and central control. Apart from the mixed 2009 sample, we found no field evidence for commercial central-control systems or for drip and low-volume zones.
  • Maker data terms. We did not read any maker’s data processing terms, so data retention for connected controllers is open.
  • Local rules. We did not read any city or county ordinance, including those in San Diego County, or any water supplier rebate terms.
  • The official code site. We did not open the official online California Code of Regulations to confirm the codified wording.
  • Connected-device law. California’s connected-device security statute did not open, so it is not described.

Using this in Field Method

Site Lens, for San Diego County sites, has an Irrigation Planning tab for selecting candidate planting groups for hydrozone review. You enter dated ETo, effective rainfall, a landscape coefficient and irrigation inputs, and download an editable Excel worksheet with demand calculations and MWELO scheduling parameters. It covers one candidate hydrozone and one period, does not determine MWELO compliance, and does not select or program a controller. For the energy and emissions attached to irrigation water, see our article on estimating a landscape carbon footprint.

Sources

  1. US EPA WaterSense — WaterSense Labeled Controllers
  2. UF/IFAS Extension — Smart Irrigation Controllers: Operation of Evapotranspiration-Based Controllers (AE446)
  3. UF/IFAS Extension — Smart Irrigation Controllers: What Makes an Irrigation Controller Smart? (AE442)
  4. UF/IFAS Extension — Smart Irrigation Controllers: How Do Soil Moisture Sensor (SMS) Systems Work? (AE437)
  5. Irrigation Association — Smart Water Application Technologies (SWAT)
  6. US EPA WaterSense — Weather-Based Irrigation Controllers
  7. US EPA WaterSense — Specification for Weather-Based Irrigation Controllers, Version 1.1 (September 2, 2021)
  8. US EPA WaterSense — Performance Overview: Weather-Based Irrigation Controllers (April 2022)
  9. US EPA WaterSense — Soil Moisture-Based Irrigation Controllers
  10. US EPA WaterSense — Specification for Soil Moisture-Based Irrigation Controllers, Version 1.0 (February 2021)
  11. US EPA WaterSense — Specification for Soil Moisture-Based Irrigation Controllers, Supporting Statement, Version 1.0 (February 2021)
  12. California Department of Water Resources — Model Water Efficient Landscape Ordinance
  13. California Department of Water Resources — MWELO Regulations and Appendices, effective January 2, 2025
  14. Metropolitan Water District of Southern California and East Bay Municipal Utility District — Evaluation of California Weather-Based "Smart" Irrigation Controller Programs (July 1, 2009)
  15. NIST — IR 8425, Profile of the IoT Core Baseline for Consumer IoT Products (September 2022)
  16. Hunter Industries (maker documentation) — Hydrawise App: Controller Settings
  17. Hunter Industries (maker documentation) — Hydrawise: Web and App Security Statement

These tools support site review and design studies. Read each tool’s stated capabilities and limits before use.

Explore Site Lens
Back to Research