**1. Introduction**

The state of California, and in general the U.S. west, has some of the largest cities across the nation, making urban water demand a vital component of any integrated water resources management plan. Statistics indicate California's population rose to nearly 40 million in 2016 and projections by the California Department of Finance show an increase to more than 45 million by 2060 [1]. It is expected that, in the future, competition for water and land resources among urban, environmental, and agricultural uses will intensify as a result of increased population, coupled with changes in land use and climate [2]. In addition, climate change is altering precipitation and temperature patterns, making drought severity likely to increase in the American Southwest [3].

Irrigation demand is a significant component of total water use in the urban sector in California [4]; therefore, improving irrigation water use efficiency is a crucial water conservation strategy. Considerable water savings have been reported as a result of implementing emerging technologies for landscape irrigation management [5–7]. Soil moisture sensor-based and evapotranspiration (ET)-based smart irrigation controllers can help increase irrigation water use efficiency by maintaining the root zone soil moisture status within a programmed desired range and scheduling irrigation based on crop coefficient and reference ET (ETo) data, respectively.

Water agencies in California often incentivize the adoption of smart irrigation controllers in residential areas. According to Singh et al. [8], in 2018–2019, approximately half of the major water agencies across the state provided rebates, ranging from \$45 to \$300, for installing ET-based smart irrigation controllers. However, the scientific research on the efficacy of smart irrigation controllers for autonomous irrigation scheduling and water conservation is limited in the region. Most scientific work on smart irrigation controllers has focused on avoiding over-irrigation in humid areas with abundant precipitation [9]. Davis et al. [6] compared ET-based smart controllers with a time-based treatment in Florida and reported, on average, a 43% reduction in applied water. Compared with timer-based fixed irrigation, the water-saving potential of ET-based controllers was also reported in other case studies in Florida and Nevada [10,11]. Recently, we [7] used an ET-based smart irrigation controller for autonomous irrigation scheduling of 'Tifgreen' hybrid bermudagrass in southern California and obtained promising results. In another study conducted in Southern California, Bijoor et al. [12] investigated the water budgets of lawns under three management scenarios, including the use of a smart soil moisture sensor-based controller. They concluded that the implementation of smart sensors was a more significant option than the choice of turfgrass species in irrigation efficiency.

An urban feature for potential water conservation is the turfgrass landscape, as it is a large component of urbanized land area. Turf has become an important crop based on the acreage planted, including residential, commercial, and institutional lawns, parks, golf courses, and athletic fields. Beyond recreational use, turf provides valuable ecosystem services that are in high demand by society, like capturing runoff, contributing to the abatement of the heat island, reducing dust and noise, and fostering biodiversity [13]. However, the turfgrass irrigation water requirement could be more significant than some alternative landscape species [14]. Consequently, it is crucial to precisely identify the minimum water requirement of commonly planted turfgrass species and study the water use of alternative warm-season turfgrass species that are more resistant to heat, drought, and salinity.

In a recently published study [7], we introduced the turfgrass water response function (TWRF) as an empirical statistical model to estimate the response of turfgrass species (based on the aesthetic values) to varying irrigation scenarios and ETo demand. We used NDVI as the response variable owing to its overall high correlation to turfgrass health and growth. We estimated the response of hybrid bermudagrass to varying irrigation and water conservation scenarios in inland Southern California using long-term ETo data.

This study was carried out to (i) determine the response of hybrid bermudagrass and tall fescue to varying irrigation scenarios (level and frequency) in central California, (ii) evaluate the use of an ET-based smart irrigation controller for autonomous irrigation scheduling, (iii) monitor and assess the dynamics of near-surface soil moisture over time under the imposed irrigation scenarios, and (iv) develop regression-based TWRFs and use them to estimate the response of turfgrass species to irrigation scenarios based on long

term mean ETo demand in the study region. Part two of this study focuses on applying ground-based remote sensing for turfgrass irrigation management [15].

#### **2. Materials and Methods**

## *2.1. Study Area*

A two-year turfgrass irrigation research project (2018–2019) was conducted at the University of California Agricultural and Natural Resources Kearney Research and Extension Center (36◦36 02.2 N 119◦30 38.8 W) in Parlier, California. Figure 1 depicts the long-term mean (1983–2021) ETo and precipitation data for the study site obtained from a nearby California Irrigation Management Information System (CIMIS) weather station (station number 39). The long-term mean annual ETo and precipitation for the region were 1367 and 273 mm yr−1, respectively. ETo demand exceeds the natural precipitation, indicating the need to irrigate urban landscape species, particularly over the summer. The long-term data show that peak ETo of 7 mm per day occurred in early July (Figure 1). The annual precipitation was 192 mm and ETo was 1452 mm in 2018. The annual precipitation was 268 mm and ETo was 1462 mm in 2019.

**Figure 1.** The long-term mean reference evapotranspiration (ETo) and precipitation trends at the study site obtained from a nearby CIMIS weather station (**a**). The lightbox used in this study to collect digital images for the visual rating (**b**).

The soil at the research site is classified as Hanford fine sandy loam (websoilsurvey.sc. egov.usda.gov; accessed on 18 August 2021). Figure 2 shows the laboratory-determined soil water retention and hydraulic conductivity curves for the experimental site. Four undisturbed soil samples were taken from approximately the top 20 cm layer at the beginning of the trial. The samples were analyzed to determine the soil water retention and hydraulic conductivity curves using HYPROP and WP4C laboratory instruments [16]. The soil water retention (based on the composite data) at soil tensions of 10, 33, and 1500 kPa was 0.25 m3 m<sup>−</sup>3, 0.17 m3 m−3, and 0.06 m3 m−3, respectively. The saturated hydraulic conductivity was 11 mm day<sup>−</sup>1.

**Figure 2.** The soil water retention (**a**) and hydraulic conductivity (**b**) curves of the surface soil (0–20 cm) at the study site.
