Sensor-Based Automation of Irrigation on Bermudagrass, during Wet Weather Conditions

Taken from the Journal of Irrigation and Drainage Engineering @ASCE/March/April 2008

Abstract: New technologies could improve irrigation efficiency of turfgrass, promoting water conservation and reducing environmental impacts.  The objectives of this research were to quantify irrigation water use and to evaluate turf quality differences between (1) time-based scheduling with and without a rain sensor (RS); (2) a time-based schedule compared to a soil moisture sensor (SMS)-based irrigation system; and (3) different commercially available SMS Systems.  The experimental area consisted of common bermudagrass, located in Gainesville, Fla.  The monitoring period took place from July 20 to December 14, 2004 and from March 25 to August 31, 2005.  SMS-based treatments consisted of irrigating one, two or seven days a week, each with four different commercial SMS brands.  Time –based treatments with or without RS and a nonirrigated treatment were also implemented.  The treatment with the rain sensor resulted in 34% less water applied than that without the rain sensor (2-WORS) treatment.  Most SMS brands recorded irrigation water savings compared to 2-WORS, ranging from 69 to 92% for three of four SMSs tested, depending on the irrigation frequency.  Therefore, SMS systems represent a promising technology because of the water savings that they can achieve during wet weather conditions while maintaining acceptable turfgrass quality.


Turfgrass is the main cultivated crop in Florida with nearly four times the acreage as the next largest crop, citrus.  Irrigation of residential, commercial, industrial, and recreational turf areas is commonly employed to ensure acceptable turf quality.  As a consequence of problems related to droughts coupled with the steadily increasing demand for water the state of Florida has imposed restrictions on irrigation water use.  The development of best management practices (BMPs) for irrigation water use in turf has become an undeniable strategic, economic, and environmental issue for the state.  New landscape irrigation technologies could improve irrigation efficiency by promoting water conservation and reducing environmental impacts.

Florida receives an average of approximately 1,400 mm of rainfall a year, which varies depending on location in the state.  Although rainfall, typically, exceeds evapotranspiration (ET), irrigation is required because total annual rainfall, typically, varies both geographically and temporally, and lack of rainfall for even a few days causes depletion of moisture in the predominately sandy soils found in Florida.

Florida has the second largest withdrawal of groundwater for public supply in the United States.  In 1995, nearly 93% of the population in Florida used groundwater as a drinking water source.  Moreover, Florida has a fast-growing population with a net inflow of more than 1,100 people a day and is projected to be the third most populous state in the nation by 2025.  The USCB estimated that Florida accounted for approximately 11% of all new homes constructed in the United States in 2003, the largest amount in any single state; the majority of them with an inground irrigation system.  As urban populations swell, pressures on limited supplies of clean water are increasing.  Saltwater intrusion in the Floridian aquifer has been found in coastal Hillsborough, Manatee, and Sarasota counties.

The primary use of residential outdoor water is irrigation.  A study in the U.S. indicated that households that use automatic timers to control their irrigation systems use 47% more water outdoors than those without timers, homes with in-ground sprinkler systems use 35% more water outdoors than those without in-ground systems, and on average, 58% of household water is used outdoors.  In the Central Florida Ridge, the potable water used for landscape irrigation has been found to be as high as 74%, with an average of 64%, and even when irrigation is restricted to two days a week, typically, homeowners tended to overirrigate.

Overirrigation or underirrigation can negatively affect turfgrass quality.  It has been reported that deeper and less frequent irrigation improves turfgrass quality.  Augustin and Snyder (1984) concluded that this practice tended to reduce N leaching in sandy soils, increasing N utilization, resulting in a better color rating (better quality).  Bonos and Murphy (1999) reported an increase in drought stress was imposed, Jordan et al. (2003) found that bentgrass irrigated every four days produced a significantly larger and deeper root system, a higher shoot density, and high overall plant health – resulting in greater turf quality – than that watered every 1 or 2 days (even under golf putting green management conditions).  McCarty (2005) summarizes that drier soil conditions slow shoot growth, and increase root growth and leaf water content.  Moreover, limitations to the establishment and survival of some turfgrass weeds and reduction of some pathogen severity have been associated with deep, infrequent irrigation.  Hence, better irrigation scheduling by homeowners may lead to improved turfgrass quality coupled with potential savings in irrigation water use.

Over the last decade, the soil moisture sensor (SMS) industry has advanced dramatically.  Two basic reasons can explain this advancement.  The first has been the major development of computer technology, with more powerful, smaller, and more economical integrated circuits.  The second phenomenon has been the significant advances in the application of electromagnetic methods to the measurement of soil water content.

Since 1991 Florida law has required a rain sensor (RS) device or switch hooked up to all automatic lawn sprinkler systems.  A rain sensor is a piece of equipment designed to interrupt a scheduled cycle of an automatic irrigation timer when a specific amount of rainfall has occurred.  Benefits and advantages of its use are similar to those of SMSs, and have been summarized by Dukes and Haman (2002a).  Even though RSs have been mandated on all automatic irrigation systems installed after 1991, and have been commercially available for many years, little evidence related to their usefulness and/or to quantify their water savings exists.

The goal of this research was to find out if different SMS systems (sensor with a proprietary controller) could reduce irrigation water application – while maintaining acceptable turf quality – compared to various time-based irrigation schedules to simulate common homeowner practices.  The objectives of this experiment were to quantify irrigation water use and to evaluate turf quality differences between: (1) time-based scheduling with and without an RS; (2) a time-based schedule compared to an SMS-based irrigation system; and (3) different commercially available irrigation SMS systems.

Materials and Methods

The experimental area was located at the Agricultural and Biological Engineering Department research facilities, University of Florida, Gainesville, Fla.

The site consists of 72(3.7 m x 3.7 m) plots on a field covered with well-established common bermudagrass.  Each plot was sprinkler irrigated by four quarter-circle pop-up spray heads.  Plots were mowed twice weekly at a height of 5.5 cm.  Agrochemicals were applied as needed to control weeds and pests, with no visual toxicity signs on the bermudagrass after the applications.  Nutrient applications were made using ammonium sulfate (21-0-0), at an N rate of 50 kg/ha, in April and May of 2004, before the beginning of the experiment.  Then, a granulated 16-4-8 controlled- release fertilizer (Professional Turf Fertilizer, TurfGro, Phoenix. Ariz.) was applied at an N-P-K rate of 180-45-90 kg/ha, in July 2004 and April 2005.

Four commercially available SMS were selected for evaluation: Acclima Digital TDT RS-500 (Acclima Inc., Meridian, ID.), Watermark 200SS-5 (Irrometer Company, Inc., Riverside, Calif.), Rain Bird MS-100 (Rain Bird International, Inc., Glendora, Calif), and Water Watcher DPS-100 (Water Watcher, Inc., Logan, Utah), codified as AC, IM, RD, and WW, respectively.  Each SMS system included a sensor to be buried in the soil and a controller, which could be adjusted to different soil water thresholds.


Two basic types of treatments were defined:  SMS-based treatments, and time-based treatments (Table 1).  All four SMS brands were tested at three irrigation frequencies: one, two, and seven days per week (1, 2, and 7 day/week, respectively), result in 12 SMS/frequency combinations.  The 1 and 2 day/week watering frequencies represent typical day of the week irrigation restrictions imposed in Florida (FDEP 2006; SJRWMD 2006).  Within the time-based treatments, a frequency of 2 day/week was defined (the most common in Florida and current watering restriction in the study area).  To simulate requirements imposed on homeowners by Florida Statutes (Chap. 373.62), two time-based treatments were connected to a rain sensor; with-rain sensor (2-WRS) and deficit-with-rain sensor (2-DWRA).  The rain sensor (Mini-click II, Hunter Industries, Inc., San Marcos, Calif) was set at a 6 mm rainfall threshold.  A without-rain-sensor treatment (2-WORS) was also included, in order to simulate homeowner irrigation systems with an absent or nonfunctional rain sensor.  Finally a nonirrigated treatment (0-NI) was implemented as a control for turfgrass quality.  Experimental treatments were replicated four times, in a completely randomized design.

The weekly irrigation depth was programmed to replace 100% of the monthly historical net irrigation requirement, based on recommendations by Dukes and Haman (2002b) for the area where this experiment was carried out.  All treatments were programmed to apply the same amount of irrigation per week, except for treatments 2 –DWRS (60% of this amount), and 0-NI.  Therefore, differences in water application among treatments would be the result of sensors bypassing scheduled irrigation cycles.

Table1.  Irrigation Treatment Codes and Descriptions
Treatment Codes Irrigation frequency (days/week) Soil moisture sensor brand or treatment description
(a) Time based
2-WORS 2 Without rain sensor
2-WRS 2 With rain sensor
2-DWRS 2 Deficit with rain sensor,   60% of 2-WRS
(b) SMS based
1-AC 1 Acclima
1-RB 1 Rainbird
1-IM 1 Irrometer
1-WW 1 Water Watcher
2-AC 2 Acclima
2-RB 2 Rainbird
2-IM 2 Irrometer
2-WW 2 Water Watcher
7-AC 7 Acclima
7-RB 7 Rainbird
7-IM 7 Irrometer
7-WW 7 Water Watcher
0-NI 0 No irrigation
Note:  SMS=soil moisture sensor.

We are now skipping over several pages of scientific explanations of exactly what was done and how and resuming the discussion with the results and conclusions.

If you want the entire information, use this link,

Water Savings

The results show the water savings (%) of each treatment compared to the time-based treatments 2_DWRS, 2-WRS, and 2-WORS.  Treatments 7-AC, and 7-RB achieved the highest amounts of water savings throughout this experiment and, as expected, 2-WORS applied more water than all the other treatments.  On the other hand, the IMs always allowed more water to be applied compared to the other brands in every frequency tested.  This could be due to their reported limitations to timely sense differences in soil water content, their hysteric behavior, the high variability of readings, and their limitations in sandy soils, where low tension values are necessary to prevent plant stress.

When compared to the water conservative 2-DWRS treatment, brands AC, RB,, and WW showed water savings that ranged from 44 to 80%, 55 to 76%, and 26 to 57%, respectively.  On the other hand, al IM frequencies applied more irrigation than 2-DWRS, with values that ranged from 15 to 77% more water.

Treatment 2-IM was the only SMS-based treatment that applied more water than the time-based 2-WRS (11%).  Conversely, 1-IM and 7-IM reduced water application 20 and 28%, respectively, compared to 2-WRS: AC sensors recorded irrigation water savings ranging from 65 to 88%, RBs from 72 to 85%, and WWs from 54 to 73%, depending on the irrigation frequency tested.  It is important to remark that these water savings were on top of those already achieved by 2-WRS.  Therefore, these results show that, in general, SMSs can also act as rain shut-off devices, although with a superior performance than rain sensors in terms of water savings.

When the irrigation treatments were compared to more than 75% of the surveyed homeowners in Florida, with a nonfunctional or absent rain sensor (2-WORS), the difference in water savings increased, ranging from 77 to 92% for the ACs, 81 to 90% for RDs, 69 to 82% for WWs, and 27 to 53% for IMs.  Even 2-IM (which applied 11% more water the 2-WRS) showed water savings (27%) with respect to 2- WORS, indicating that this sensor was operative but did not bypass as many scheduled irrigation cycles as other SMS-based treatments.

These results clearly demonstrate that the use of SMSs (along with traditional timers in residential irrigation systems) could lead to water savings more than twice as much as a rain sensor device alone, even when the time schedule is programmed to provide 60% of net irrigation requirements.

Automation of Irrigation Systems

Complete automation of a residential irrigation system, based on SMSs, could be achieved by programming the timer to run every day as a scheduling strategy.  Then, SMSs will allow the system to initiate the scheduled irrigation cycles only when it is actually needed by the turfgrass (or other irrigated plant type), and override cycles when the sensed water content is over a preset threshold.  In this experiment, this type of control was confirmed when the 7 day/week irrigation frequency applied significantly less water than the other frequencies, and when two of the SMS-based treatments, programmed to run 7 day/week, consistently applied the smallest amount of water.  In effect, treatments 7-AC and 7-RB recorded total water savings of 85% or more, when compared to 2-WRS and 90% or more when compared to 2-WORS.

This concept (with a potential irrigation frequency of seven days a week) seems contradictory to the water use regulations and restrictions imposed by the Water Management Districts and/or municipalities in Florida (where irrigation is allowed only one or two days per week).  However, during wet weather conditions these results suggest that setting the correct threshold, and programming the automatic irrigation system to run everyday for a short period of time (allowing the SMS to decide whether to irrigate), could save large amounts of water and may be a more effective water conservation strategy than day of the week watering windows.  Moreover, this concept is not in opposition to the general recommendation for deeper and less frequent irrigation for turfgrass, because these treatments (7-AC and 7-RB) overrode almost every scheduled irrigation cycle, resulting in a low actual irrigation frequency.


The three time-based treatments (2-WORS, 2-WRS, and 2-DWRS) were significantly different from each other during the study period.  The treatment with a functional rain sensor (2-WRS), at a 6 mm threshold, applied significantly less water (34%) than the without rain sensor treatment (2-WORS), showing the importance of a well-maintained rain shut-off device In all automated irrigation systems in Florida.  On the other hand, treatment 2-DWRS, applied close to the desired 60% of the water applied by 2-WRS.  These time-based treatments were established to mimic the operation of irrigation systems carried out by different homeowner profiles.  However, according to the results of this research, these treatments were fairly well managed compared to homeowners’ actual operation practices in the Central Florida Ridge.  Therefore, results in water use from this experiment can be considered conservation and differences for actual homeowners could be even larger.

For the SMS treatments, all three irrigation frequencies tested (1, 2, and 7 day/week) were significantly different.   The 2 day/week frequency applied the highest volume of water, followed by the 1 day/week frequency, and the 7 day/week was the one that applied the least amount of water.  These results suggest that scheduling high-frequency irrigation cycles (7day/week) in closed control loop irrigation systems appears to be a viable strategy regarding water conservation for turfgrass irrigation in Florida’s sandy soils.

The results showed that, on average, the SMS-based treatments were significantly more efficient as a means to save water than the time-based treatments.  However, not all SMS treatments tested performed the same.  The 2-IM treatment was the only SMS-based treatment that applied significantly more water than 2-WRS (11%).  The other two IM treatments, 1-IM and 7-IM, applied less water than 2-WRS (20 and 28%, respectively), but always applied more water than the other brands/treatments in every frequency tested.  The other brands (AC, RB, and WW) resulted in irrigation water savings compared to 2-WRS, which ranged from 54 to 88%, depending on the irrigation frequency.  These results showed that most SMSs can also act as rain sensors, with superior performance in terms of water savings.  When these last brands were compared to 2-WORS, the differences in water savings increased, and ranged from 69 to 92 % over the 308-day study period.

It should be noted that the specific performance of the individual sensors largely depends on the threshold setting and the sensor burial depth.  Even when sensor burial depths were as similar as practically possible in this experiment, the sensors thresholds might have varied slightly, hence, affecting the results to some extent.  In any case, soil moisture sensor systems appear to be a promising technology that could lead to a complete automation of residential irrigation systems, to substantial savings in irrigation water, and to sound environmental and economic benefits to the state.  Testing this technology with actual irrigation systems on homes is recommended to validate these results.