2015 journal article
Residential Irrigation Water Use in the Central Piedmont of North Carolina. II: Evaluation of Smart Irrigation Technologies
JOURNAL OF IRRIGATION AND DRAINAGE ENGINEERING, 141(4).
A study was conducted in Cary, North Carolina, in the spring and summer of 2009 with the purpose of evaluating the effectiveness of two “smart irrigation” controllers based on the amount of irrigation applied and resulting turf quality in residential settings. Twenty-four residential sites were selected, in clusters of four, representing six geographical areas within the town. Each geographical cluster included one site of each treatment. The treatments were standard irrigation controller with an add-on soil moisture sensor system (SMS); standard irrigation controller with an add-on evapotranspiration-based adjustment system (ET); standard irrigation controller using seasonal runtimes based on historical climate data (ED); and a control group which used a standard irrigation controller with no intervention (CON). Weekly water usage was obtained from irrigation meter readings and turf quality was characterized using a visual rating and a normalized difference vegetation index (NDVI) meter. Maximum water savings were achieved by the SMS treatment (42% less than CON), followed by ET and ED treatments. No statistical difference in average weekly water use was found between the ET group and the ED group provided with controller run-time guidance. The mean weekly visual turf quality index was highest for the SMS treatment, but only statistically different from the ED group. Average weekly NDVI was greatest for the ED group, although average NDVI values were not statistically different among any of the groups. Although water use was less during the 2009 study period contrasted against the three previous years for those receiving some form of intervention (ED, ET, and SMS), the same trend in water use was found by the CON group, rendering any findings in change in behavior inconclusive. Variability in water application by cooperator groups receiving an intervention decreased in the study period compared to the three previous years, suggesting an impact.