Smart Farming Technologies for Groundwater Conservation in Transboundary Aquifers of Northwestern México

Abstract

This study evaluated the performance of a smart farming technology (SFT) and a climate-smart agriculture (CSA) approach for improving irrigation management in pecan (Carya illinoinensis) orchards in México through soil moisture monitoring, evapotranspiration estimation, and real-time data integration. Continuous monitoring allowed irrigation to be maintained at field capacity, preventing plant stress while avoiding total soil saturation or permanent wilting point. Calibration of soil moisture sensors showed a very strong correlation (R² = 0.99) between sensor reverse voltage and volumetric soil water content in predominant sandy loam soils, confirming the reliability of the monitoring system for irrigation scheduling. Seasonal analysis of reference evapotranspiration (ETo) and crop evapotranspiration (ETc) revealed increasing atmospheric water demand during summer months, with crop coefficient (Kc) values ranging from approximately 0.3 during dormancy to 1.0–1.3 during peak vegetative growth. After five years of field implementation of the technology, results showed water savings exceeding 50% compared with traditional flood irrigation practices. The optimized irrigation schedule reduced total seasonal irrigation depth from 216 cm to 128 cm, representing a 59% reduction in applied water while maintaining adequate soil moisture conditions for crop development at field capacity (FC). These results highlight the potential of integrating sensor-based monitoring, evapotranspiration modeling, and IoT platforms to enhance water-use efficiency and support sustainable pecan production under increasing climate variability.

1. Introduction

The state of Chihuahua, located in northern Mexico, is the largest producer of pecans [Carya illinoinensis] in the country, with 102,000 t (an average of 1.5 t per hectare), representing 65% of the country’s total production and a value of approximately 627 million dollars [1]. However, pecan trees are also a high water consumption crop, which increases water demands for this economic agricultural activity, requiring as much as 1700 mm of annual evapotranspiration (ET) as it relates to tree age, canopy, soil type, tree density per hectare, and agricultural practices [2,3,4,5,6]. Furthermore, losses in irrigation efficiency in this desert region occur through evaporation, deep percolation, or runoff [7], whereas the binational region of the Los Muertos/Pluvial Lake Palomas Basin and nearby regions are facing increasing water demands under stressed climate conditions, with drought intensifying [8]. Therefore, rural communities in Chihuahua are struggling with limited surface water resources for irrigation, which in turn causes demand for groundwater, generating water table depletion while providing for such surface water deficit; in turn, new groundwater resources are demanded through searching for groundwater in a complex geohydrological desert environment [9,10,11]. Thus, water availability and, most importantly, groundwater resources for agriculture will play a significant role in the binational region [12].

Under current drought conditions in northwestern Mexico/the southwestern US, the situation in the state of Chihuahua is very critical [13]. According to the North American Drought Monitor, 100% of Chihuahua was under a specific drought category in 2025, with the northwest transboundary region experiencing the most severe conditions [14]. In this region, severe drought affected 48% of the state, and exceptional drought affected another 28% [13]. Furthermore, flash droughts (FDs), which are defined in this paper as intensified, rapid changes in precipitation, temperature, wind, and radiation [15] in the state of Chihuahua, represent an extraordinarily dynamic and potentially devastating climatic phenomenon that emerges with little warning, leading to severe drought conditions within weeks. These conditions are characterized by high temperatures, rapid soil moisture depletion, and increased evapotranspiration (ET), collectively accelerating land drying. Such FDs, particularly when coinciding with heatwaves, can have disproportionate impacts on disadvantaged communities, as seen across many regions of the Global South. Chihuahua, a predominantly semi-arid region, is particularly susceptible to the effects of FDs given its already limited water availability. However, the combination of drought and extreme heat, known as compound heat-flash drought (CHFD) events, can significantly exacerbate constraints on plant photosynthesis due to the strong coupling between soil moisture and temperature [16]. This, in turn, limits ecosystem productivity and exacerbates the vulnerability of local populations and agricultural systems. In terms of socioeconomic impacts, these events can result in significant losses of agricultural land, livestock, and water resources in the region, thereby affecting communities’ food and economic security. Given that Chihuahua is part of the Global South, where the regions most affected by the decline in socio-ecosystem productivity due to CHFDs are located, local and national authorities must focus their efforts on monitoring, predicting, and mitigating the rise of these phenomena to minimize their adverse impacts [16]. Hence, to sustain the pecan industry and maintain productivity under extreme climatic conditions in this transboundary region, greater attention should be given to developing a better understanding of water management strategies for sustainable agriculture and food security [2].

In Mexico, more than 70% of water rights adjudications are allocated to agricultural production; however, water-use efficiency in this sector is relatively low (40–60%) [17]. Moreover, in the Mexican section of this transboundary watershed, nearly 300 groundwater rights adjudications are registered in the Public Registry of Water Rights (REPDA), where most groundwater wells employ inefficient irrigation technologies and draw from low-yielding aquifers. In addition, it has been documented that agricultural water demand will compete with other users under climate change, reducing food production in agricultural areas in the southwestern US and northern México [18,19,20]. Irrigation scheduling, which improves crop water-use efficiency (Actual Evapotranspiration, ETa), can be estimated using a specific crop coefficient (Kc) and the Reference Evapotranspiration (ETo) for the cultivar under consideration [21,22]. Real-time monitoring of water loss due to CHFD throughout the growing season (DGS) could improve traditional irrigation scheduling while saving a significant volume of water for agriculture by applying a Smart Farming Technology/Climate Smart Agriculture (SFT/CSA) approach [20,21]. Nonetheless, managing real-time information for agricultural applications can be challenging, as specialized technologies are required (e.g., computer servers, online connectivity, Internet of Things (IoT), which refers to the hyper-connections of devices, sensors, and data storage and transmission tools, soil moisture sensors, remote sensing data, and algorithms). In this process, vast amounts of digital data are collected to estimate agricultural parameters, which, when implemented using machine learning (ML) approaches, could improve decision-making and reduce farm risks. Therefore, technology for hosting digital data used in agricultural practices for water conservation could enable the capture of specific details across different situations during the DGS, thereby improving the selection of solutions that fit complex problems by leveraging advances in techniques such as real-time irrigation scheduling [20,21].

This article aims at two main objectives: (i) to evaluate a methodological framework to implement SFT/CSA in the northwest region of Chihuahua, México; and (ii) to present preliminary results from the placement and operation phase of IoT SFT/CSA as they relate to real-time ET in a pecan orchard. This research is presented as a pilot study designed to share the technology with local small-scale farmers on both sides of the border between Mexico and the USA. We explore the potential benefits of an SFT/CSA approach for reducing water consumption and present our experience implementing these technologies in a drought-stressed binational agricultural region. It is essential to note that a cost–benefit analysis, the socio-economic impacts of such an experience, or an evaluation of farmers’ adoption of such an SFT/CSA are not reported; however, this will be addressed in a future paper as the project continues its current evaluation and development.

2. Materials and Methods

2.1. Study Area

The Chihuahuan Desert is a transboundary arid ecosystem located in NW-México and SW-US territory, encompassing the Los Muertos/Lake Palomas Basin (LM/LPB), which lies in Doña Ana County, NM, in the US, and in the Municipality of Ascensión, Chihuahua, México (Figure 1).

Within the LM/LPB, the Palomas-Guadalupe Victoria aquifer (0812) covers an area of 1830 km² [23], whereas the specific sample data site is located 40 km south of the border region of Columbus, NM, in the United States. The climate of the area is classified according to the Köppen classification, adapted by E. García for the conditions of the Mexican Republic [24]. This region corresponds to climate type BWk (x′), which resembles a very arid, temperate climate, whose general characteristics average an annual temperature between 12 °C and 18 °C. The coldest month ranges from −3 °C to 18 °C, and the hottest month averages less than 22 °C; precipitation is distributed throughout the year, and winter precipitation is estimated to be greater than 18% of the annual total. For the regional climatological analysis, information from the Ascensión, Palomas, and Bismarck climatological stations, which have records from 1903 to 2008, was used. Based on these datasets, the average annual precipitation in the study area is approximately 318 mm. The average annual temperature is 17.3 °C and exhibits a parabolic pattern, with higher values in May, June, July, August, and September and lower values in the remainder of the year. The lowest recorded temperatures occur in December, January, and February, whereas potential evaporation is 1686 mm/year during the summer time (June–August) [24]. For this research, climatological data from the weather station at Rancho El Regalo (RER) located at the study site were used to formulate calculations with data for the period from 2021 to 2025.

2.2. Socioeconomic Assessment with an Emphasis on Uses of Regional Groundwater for Agriculture

A socioeconomic assessment with an emphasis on the uses of regional groundwater for agriculture in the Palomas-Guadalupe Victoria aquifer (0812) was based on data from the federal Mexican resource in the Public Register of Water Rights (REPDA) provided by CONAGUA up to the year 2021, where land tenure and equity in the allocation of water rights adjudication volumes to different groundwater users were used to calculate the Gini Index [25]. To obtain the population of the study area, data from the Population and Housing Censuses [23] were used. These datasets are comprehensive, periodic studies that collect demographic, socioeconomic, and housing information on Mexico’s population at a specific point in time. Groundwater well identifications and their allocated water volumes were obtained from the Public Register of Water Rights (REPDA) provided by Conagua [26] for the Palomas-Guadalupe Victoria aquifer (0812). This data, updated as of 31 December 2021, provided details on the geographic locations of groundwater wells, the purposes of water use, the allocated pumping volumes per well, and the groundwater adjudication permits [27]. Additionally, a Geographic Information System (GIS) running under ArcMap v.10.3 by ESRI^® was used as the primary tool for processing and analyzing the collected data. In the initial stage, the data were converted from XLS to SHP format to determine the distribution and concentration of groundwater wells and population, as well as their relationships with other geospatial elements [27].

Using these data, the second stage of the process involved determining per capita groundwater availability in the Palomas-Guadalupe Victoria aquifer. Several key elements were considered for this calculation. First, the total amount of available groundwater equals the sum of all allocated groundwater volumes. To support our work, it is essential to have an accurate estimate of the amount of groundwater available based on hydrologic mass balance. Second, the total population represents the number of people living in the region comprising the study area, based on current, reliable demographic data. Third, the units for water and population were kept consistent so that the calculation was valid. Water volumes are typically expressed in cubic meters (m³) or liters, and population refers to the total number of people living in the region under consideration. Finally, the basic calculation was based on the total amount of available groundwater and the total population, usage, and distribution factors. This included the amount of groundwater allocated to different uses, such as agriculture, industry, and domestic consumption. Once these data were obtained, the calculation was carried out following the formula in Equation (1) [27].

Groundwater availability per capita = (Total amount of available water)/(Total population)

(1)

This calculation yielded an average value indicating the amount of water available to each person within the aquifer area. However, it is essential to recognize that groundwater availability is not the only factor in assessing water security; other factors, such as water quality, the sustainability of water sources, and the effective management of water resources, must also be considered.

2.3. Commitment to Our Work

Within the study area, most withdrawals from aquifer 0812 are for irrigated agriculture, whereas the remainder is for domestic use [22]. The absence of a Smart Farming Technologies (SFT) and Climate-Smart Agriculture (CSA) approach to improve water efficiency in irrigation practices is common in this transboundary watershed. Furthermore, many pecan orchards in this region rely on inefficient flood irrigation. At the same time, other farmers have lost their production potential due to drought and water scarcity, with groundwater depletion threatening their investments. This is relevant to a holistic, common strategy for groundwater sustainability in this binational region, as similar challenges are occurring on the US side at the Los Muertos Basin, where the Mimbres aquifer system is located. In this aquifer, the water table has declined by nearly 27 m over the past 40 years, resulting in changes to the transboundary water table surface [12]. Furthermore, extensive research has documented the need for binational groundwater information, with substantial evidence supporting the transboundary characteristics of these aquifer systems, particularly with respect to groundwater flow connections, which have been recognized by the International Shared Aquifer Resources Management Agency (ISARM) [28]. For example, the research region receives groundwater from the Las Palmas and Palomas-Guadalupe Victoria aquifers, highlighting the importance of ensuring sustainable groundwater use on both sides of the US-Mexico border [22].

The need to implement water-saving irrigation technologies while managing real-time digital data (e.g., climatic temperatures and soil moisture) and applying a Smart Farming Technologies/Climate-Smart Agriculture (SFT/CSA) approach at the farm is critical to the future of agriculture in this binational region. SFT/CSA is a new approach to agricultural production that uses real-time monitoring, gathering digital data to address the need for in-farm information to support decision-making for small-scale farmers [20]. Furthermore, the use of digital data in agriculture can play an important role when real-time monitoring practices are applied to evaluate soil moisture, soil temperature, and soil salinity, as well as changes in climatic temperatures to monitor climatic conditions for in situ decision-making at the farm, thereby estimating evapotranspiration (ETP) and other crop parameters, such as crop coefficients (Kc). These technical practices for collecting digital data can improve water efficiency and provide cloud-based Internet support for real-time information transfer to farmers and for research and academic purposes. These datasets can be used to model irrigation schedules, fill gaps in real-time information, complete the loop of necessary inputs for real-time irrigation scheduling, and support water conservation and decision-making at the farm. Furthermore, the online availability of digital data can be used to compare and evaluate binational regional climate changes and agricultural water demands, and to assess the implementation of such SFT/CSA approaches on both sides of the border.

This research addressed the lack of scientific research and strengthened different research lines aligned with these interests to promote water conservation, binational data sharing, and the use of real-time digital data in agriculture. To delineate the research site of this paper, we selected a 4-ha parcel, planted in January 2020 in a 10 m by 10 m tree distribution, with a micro-sprinkler irrigation system at a flow rate of 36 lph. This is a multiyear project (10 years) managed annually, starting its sixth year in 2026, during which planned tasks will be carried out once all computers and sensors are installed and connected to the cloud. We also expect multiannual results across the different stages of the project, including data collection and processing, nut production, and product quality. Installing digital piezometric dataloggers for groundwater monitoring, continuing to install soil moisture sensors to assess water availability at sites with changing soil textures, and expanding the use of automated micro-irrigation (drip) systems to improve water-use efficiency are part of the project’s ongoing agenda. Therefore, digital data can play an important role in real-time monitoring of soil degradation, transboundary groundwater depletion, and climatic conditions, thereby supporting in situ decision-making at the farm.

2.4. Data Collection and Processing Procedures

Pecan trees are one of the most important crops in northern Mexico, with annual acreage increments above other types of crops. To help reduce the challenges of producing more food with fewer resources, especially water, in situ monitoring is a key tool for evaluating potential solutions in these rural communities, which face significant criticism for their high-water consumption in pecan production. Therefore, IoT capacity at the farm is essential for implementing and enhancing technological solutions. In México, the approach to humanitarian technology (innovation for all) is considered an important factor in supporting the implementation of such an SFT/CSA approach. Therefore, addressing the technological burden of adopting real-time digital data in agricultural practices and training small farmers to adopt technology holistically are essential to the success of Mexico’s agriculture. Practical applications of capturing digital data at the research site of this study will significantly contribute to a cloud-based “Data Lake” for agricultural production, benefiting the surrounding farms. Such a vision requires computing capacity and real-time solutions to enhance the regional economy and, most importantly, to improve family income among local farmers, thereby reducing the gap between farm investments and productivity while implementing the required technological infrastructure logistics (Figure 2).

Demand for irrigation water for pecan orchards has increased significantly over the past 20 years, whereas irrigation practices remain deficient due to limited or nonexistent irrigation technologies. Irrigation scheduling is a key tool for water-efficient agricultural practices; therefore, it is important to estimate real-time ETP using meteorological data and to install tensiometers to measure soil moisture and map soil physical parameters. In this project, these data were available via a cloud-based protocol for accessing modeling results after real-time regional ETP and Kc were estimated, and the preliminary results were shared with local farmers. Computer-based tools and cloud computing were used to develop the model for real-time ETP and Kc estimates, which was improved while integrating computer programming protocols, enhancing the data architecture while “cleaning data” through an ETL (Extraction-Transformation-Loading) a process that was applied to better manage our integrated data base from the “Data Lake” reservoir, using an SQL protocol running descriptive and predictive analysis to our process. Despite the advantages of implementing IoT-based monitoring systems for irrigation management, several sources of uncertainty may influence the reliability of the collected data and derived irrigation recommendations. Soil moisture measurements obtained from tensiometers and other in situ sensors are affected by soil heterogeneity, installation conditions, and calibration procedures. Tensiometers measure soil matric potential rather than volumetric water content; therefore, the conversion to plant-available water depends on soil water retention curves that vary with soil texture, structure, and organic matter content (Figure 3).

Spatial variability within pecan orchards may also lead to localized differences in soil moisture, meaning that a limited number of sensors may not fully capture field-scale variability. Additional uncertainty arises in estimating reference evapotranspiration (ETo or ETP). Meteorological measurements used in evapotranspiration models, such as air temperature, relative humidity, solar radiation, and wind speed, contain inherent sensor errors that may propagate into evapotranspiration calculations [29]. These measurement uncertainties may result in deviations of approximately 5–15% in daily evapotranspiration estimates. Furthermore, crop coefficients (Kc) used to estimate crop evapotranspiration may vary depending on orchard management practices, tree age, canopy density, and local microclimatic conditions, which may not be fully captured by standardized coefficients reported in the literature [30]. The use of IoT communication networks and cloud-based data systems introduces additional operational limitations. Wireless data transmission may experience intermittent connectivity or packet loss, leading to temporary gaps in the monitoring dataset. Similarly, long-term sensor performance may be influenced by battery degradation, environmental exposure, and maintenance conditions. Finally, uncertainties may also arise during data processing, particularly during extraction, transformation, and loading (ETL) procedures, where missing data, time-synchronization issues, or sensor outliers may affect modeling results. Consequently, the irrigation scheduling framework developed in this study should be interpreted as a decision-support tool rather than an exact prediction system. Continuous calibration, periodic sensor validation, and the future integration of multiple monitoring points across the orchard are recommended to reduce uncertainties and improve the reliability of real-time irrigation management strategies [29].

Additionally, agroclimatology enabled the prediction of the behavior of climatological variables and their effects on agricultural production, thereby facilitating decision-making. These practices, such as managing irrigation schedules and estimating crop water stress, help improve water-use practices while pursuing SFT/CSA. To estimate reference evapotranspiration (ETo), Hargreaves and Samani [5] developed an equation (Equation (2)) using minimal climatological data, accounting for variations in temperature and solar radiation (R0). For this purpose, climatological “data lakes” generated at the Rancho El Regalo (RER) climatological station (Figure 3) were used, along with a network of humidity and temperature sensors installed in the pecan orchard at the research site.

ETo = 0.0135 KT (Tmed + 17.78) R0 × (Tmax − Tmin) 0.5

(2)

Subsequently, Equation (2) was used, with a coefficient of transpiration (KT) of 0.162 for inland regions, following the modification proposed by Allen [31] and accepted by Samani [4]. Values for solar radiation (R0) were taken from the extraterrestrial solar radiation table, using the average range value for latitudes between 30° and 32°, in the Northern Hemisphere [4], obtaining the results of the ETo for the years 2021, 2022, and 2023 from Equation (2). In addition, other climatological variables, such as precipitation and relative humidity, were examined to identify trends in their relationships with ETo.

Reference evapotranspiration (ETo) is a key parameter for irrigation scheduling and water balance studies in arid and semi-arid regions, where among the most widely used methods for estimating ETo are the FAO Penman–Monteith and Hargreaves–Samani methods. The FAO Penman–Monteith equation [31] is considered the standard approach because it incorporates multiple meteorological variables, including solar radiation, air temperature, humidity, and wind speed. In contrast, the Hargreaves–Samani method [5] is a simplified empirical model that primarily relies on maximum and minimum air temperature and extraterrestrial radiation. The Penman–Monteith method generally provides more accurate estimates of reference evapotranspiration because it accounts for both aerodynamic and energy-balance processes that control evapotranspiration. However, its application requires a complete set of meteorological measurements, which were unavailable at the research site. Therefore, the Hargreaves–Samani method was applied as an alternative, since only temperature data were available, producing ETo estimates that were generally consistent with those obtained using the Penman–Monteith approach. However, because the region experiences periods of high wind speed or elevated vapor pressure deficit, the Hargreaves–Samani method tends to underestimate evapotranspiration because it does not explicitly account for wind speed and humidity. However, the simplified Hargreaves–Samani mode provided reliable estimates using the limited meteorological data available for this study. In arid regions like the study area, previous studies have reported mean absolute errors ranging from approximately 0.5 to 1.5 mm day⁻¹ when Hargreaves–Samani estimates are compared with Penman–Monteith results [2,3,4]. In the present case, the error magnitude remained within this range, suggesting that the temperature-based method provided an acceptable approximation for operational irrigation management.

The RER meteorological station collected data at 5 min intervals, and maximum (Tmax) and minimum (Tmin) temperatures were recorded daily from January 2021 to December 2025, in contrast to the official temperature estimates for the region [24]. Climatological data were obtained from a Davis Vantage Pro meteorological station (Figure 3), which provided precipitation and temperature data to estimate potential evapotranspiration (ETo) using Equation (2). During this period, the temperature at the RER station where the research site is located averaged −2.7 °C in winter, with the highest temperatures recorded in July, peaking at 37 °C (Figure 4). The meteorological dataset collected at the study site includes continuous monitoring of air temperature from 2021 to 2025, a primary variable for estimating reference evapotranspiration (ETo). Temperature data are essential in widely used evapotranspiration models where air temperature directly influences vapor pressure deficit and the atmospheric demand for water. The temperature monitoring graph for the 2021–2025 period (Figure 4) illustrates the seasonal variability of maximum, minimum, and average monthly temperatures, showing a consistent climatic pattern typical of arid and semi-arid regions of northern Mexico. The dataset shows a clear seasonal cycle with maximum temperatures reaching approximately 37–40 °C during June–August, which corresponds to the peak evaporative demand period, minimum temperatures frequently ranging between 3 and 8 °C during December–January, reflecting winter dormancy conditions for pecan orchards and average temperatures during the growing season (April–September) generally ranging between 24 and 30 °C. This seasonal behavior strongly influences evapotranspiration dynamics because higher temperatures increase saturation vapor pressure and atmospheric water demand, thereby increasing ET rates. The dataset indicates moderate interannual variability in temperature patterns between 2021 and 2025. For example, the 2023 summer period shows some of the highest observed maximum temperatures, exceeding 40 °C, suggesting enhanced evaporative demand that year. Slightly lower maximum temperatures were observed during 2021 and 2022, although seasonal patterns remained consistent. Such variability is important when modeling irrigation requirements because small temperature differences can lead to significant differences in seasonal crop water demand. The temperature range observed in the dataset (often 12–18 °C between daily maxima and minima during summer) indicates strong diurnal variability, which typically leads to higher evapotranspiration rates in arid climates.

During the summer months (May–September), when pecan trees are in active growth and nut-filling stages, the combination of high temperatures and large diurnal temperature ranges suggests that daily ETo values may exceed 6–8 mm day⁻¹, depending on radiation and wind conditions.

This dataset included 5 years of observations (2021–2025), with 12 months per year, for a total of 60 observations. The variables analyzed included the maximum temperature (Tmax), minimum temperature (Tmin), and average temperature (Tavg), with the statistical analyses shown in

Table 1. Statistical analysis which contains monthly maximum, minimum, and average temperatures for 2021–2025 (n = 60 observations).

Statistic	Tmax (°C)	Tmin (°C)	Tavg (°C)
Sample size (n)	60	60	60
Mean	26.26	9.26	17.99
Median	26.50	9.25	18.55
Minimum	14.00	−5.90	4.40
Maximum	39.40	22.00	31.10
25th percentile	19.73	1.40	10.45
75th percentile	33.05	18.38	25.48
standard deviation (σ)	7.80	8.58	8.32

. The probability distribution of the monthly average temperature for the period 2021–2025 was derived from the CeTraTecAI monitoring station dataset. Because evapotranspiration is strongly temperature dependent, the probability distribution helps estimate the likelihood of high atmospheric water demand. The standard deviation is approximately σ ≈ 8.3 °C. The fitted normal distribution illustrates the seasonal variability in temperature conditions that influence evapotranspiration demand in pecan production systems in northern México. The interpretation of these data shows that the large temperature range confirms the strong seasonal climate variability typical of semi-arid northern Mexico, whereas the difference between Tmin and Tmax indicates a strong diurnal and seasonal thermal amplitude that strongly influences evapotranspiration processes. For the standard deviation (σ), which quantified temperature variability, the largest variability occurred in Tmin, reflecting strong winter cooling events; Tmax variability was slightly lower but still significant due to seasonal warming.

The statistical reliability of the monitoring dataset helped demonstrate consistent multi-year climatic patterns, statistically stable temperature averages, and significant seasonal variability that influence irrigation demand, whereas the 60-observation dataset provided a statistically robust basis for evapotranspiration modeling and irrigation scheduling. The potential uncertainty associated with energy balance closure in evapotranspiration estimation was addressed by combining meteorological-based evapotranspiration calculations with soil moisture monitoring and water balance validation. Although surface heat flux measurements in arid and semi-arid environments may introduce systematic deviations of approximately 10–20% in the surface energy balance [16], the reference evapotranspiration used in this study was estimated using standardized methods based on meteorological observations rather than direct eddy-covariance measurements of energy fluxes. To minimize potential bias, evapotranspiration estimates were cross-validated against changes in soil water content within the root zone measured by the in situ sensors. This approach allowed the consistency of evapotranspiration estimates to be evaluated against observed soil moisture depletion during irrigation cycles [32]. By integrating soil moisture measurements with evapotranspiration modeling, the potential impact of energy balance closure errors on irrigation scheduling was reduced, ensuring that calculated crop water demand remained consistent with observed soil water dynamics under the arid climatic conditions of the study area.

For this study, daily values and monthly average ETo were estimated from digital data transfer and administered under ETL and SQL protocols, which were developed to enhance the project’s “Data Lake” infrastructure. ETL (Extract, Transform, Load) is a data management process/workflow, while SQL (Structured Query Language) is the programming language used within that process to manage and manipulate the databases. The study covered a total area of 4 ha, with two wired tensiometers installed in sandy loam soil at depths of 25 cm and 45 cm (in this area, 375 two-inch-diameter pecan trees were established in 2021 with 10 m × 10 m spacing). ETP calculations using the Hargreaves-Samani Equation [5] were incorporated into an algorithm to support real-time ET calculation. Additionally, digital piezometric sensors were installed at nearby wells within the watershed to monitor and evaluate groundwater depletion, thereby informing future transboundary management of the binational aquifers. Equipment with cutting-edge technology was installed in 2020 to enable real-time monitoring of soils irrigated under a TORO^® microsprinkler system (36 lph) (The TORO Co. Micro-Irrigation Business, El Cajon, CA, USA), while measurements of temperature to calculate evapotranspiration (ETP) and atmospheric conditions at the pecan orchard were taken using a Davis Weather Station^® and two LSE01 Dragino soil sensors (Dragino Technology, Co., Shenzhen, China), these being installed to measure soil relative humidity, soil temperature, and soil EC. The LSE01 LoRaWAN is a wireless environmental monitoring device designed primarily for precision agriculture and Internet of Things (IoT) applications, and it transmits data via LoRaWAN networks to remote servers or cloud platforms. These characteristics enabled the development of the two sensor installations to monitor soil relative humidity at different depths on the same site. The soil sensors use FDR (Frequency-Domain Reflectometry) to measure the soil’s dielectric constant. The sensors were calibrated according to Hrisko’s methodology [29], which enabled calibration of the volumetric humidity of the field-collected sample at the specified depths against the sensor voltage using a linear equation (Figure 5). The sensors were installed on 15 November 2023, and collected daily data until 30 December 2025. Because sandy loam soils exhibit spatial variability in texture and structure, sensor placement was carefully selected to represent typical soil conditions within the monitored orchard. Nevertheless, it is recognized that point-based soil moisture measurements may not fully capture field-scale heterogeneity.

To visualize real-time soil moisture, temperature, and salinity using the Internet of Things (IoT), Dragino sensors were placed at a strategic location within the pecan orchard (Figure 3), which has greater canopy homogeneity. The sensor’s digital data was transmitted 100 m to the LG308 Dragino gateway, enabling connection to the server and integration with the Dragino Sensor platform, allowing real-time data viewing on the Things Board internet platform. Crop evapotranspiration (ET_c) was estimated through the variation in soil moisture, without the presence of rain. Since the sensor collects information every 5 min, it was decided to average over the period from 24:00 to 14:00. The formula for estimating ET_c is given in Equation (3) [31].

$${\mathit{E}\mathit{T}}_{\mathit{c}}=\sum -{\mathit{Z}}_{\mathit{r}\mathit{i}}{\displaystyle \frac{{\Delta \mathit{S}\mathit{W}\mathit{C}}_{\mathit{i}}}{\Delta \mathit{t}}}$$

(3)

ET_c is the daily crop evapotranspiration (mm); Z_ri is the depth of the sensor (mm); ΔSWC_i is the variation in the sensor humidity (%); and Δt is the validation time (one day).

Estimation of the crop coefficient and its relationship with remote sensors was calculated using Equation (4) [31], where Kc is the crop coefficient; ET_c is the crop evapotranspiration calculated with Equation (3) (mm d⁻¹), and ET_o is the potential reference evapotranspiration measured with climate data from the RER meteorological station and Equation (2).

$$\mathit{K}\mathit{c}={\displaystyle \frac{{\mathit{E}\mathit{T}}_{\mathit{c}}}{{\mathit{E}\mathit{T}}_{\mathit{o}}}}$$

(4)

Data for the ET_c calculation were collected with the RER weather station at the research site (Figure 3), whereas ET_c (Crop Evapotranspiration) is crop-specific and represents the water needed by a specific crop and is used for scheduling irrigation, calculating crop water requirements, and water management by multiplying ET_o by a crop coefficient (Kc) (Equation (5)) [31].

ET_c = ET_o × Kc

(5)

3. Results

The results indicate that in the 5th year of the study, water savings exceeded 50% compared with traditional flood irrigation practices in the region (

Table 2. Pecan Kc and irrigation schedule.

Schedule	Days/Irr#	Pecan Kc	Traditional Depth of Irrigation (cm)	Net Irrigation Under SFA/CSA (cm)	Water Savings (cm)
15 March	1/1	0.30	12	7	5
15 April	30/2	0.53	12	7	5
30 April	45/3	0.68	12	7	5
15 May	60/4	0.79	15	9	6
30 May	75/5	0.86	15	9	6
15 June	90/6	0.92	15	9	6
30 June	105/7	0.92	15	9	6
15 July	120/8	0.92	15	9	6
30 July	135/9	0.92	15	9	6
15 August	150/10	0.92	15	9	6
30 August	165/11	0.92	15	9	6
15 September	180/12	0.60	12	7	5
30 September	195/13	0.50	12	7	5
15 October	210/14	0.40	12	7	5
30 October	225/15	0.30	12	7	5
15 November	240/16	0.30	12	7	5
Total	240/16		216	128	88 (59% reduction)

). Furthermore, farmers are becoming familiar with this technology and are interested in implementing these tools in their own orchards. Continuous monitoring enabled us to understand water demand in the pecan orchard as trees mature and begin nut production; this will require more water as they become economically viable, producing more with fewer water resources. For calibration of the soil moisture sensors and measurement of crop evapotranspiration, the sensors detected trends in soil water content during the established period (Figure 5), with a high correlation (R² = 0.99), indicating a highly reliable soil moisture monitoring system. This correlation indicated that the sensors were measuring the correct soil moisture at the field where they were installed, which had a sandy loam soil texture, primarily sand (typically <35%) with significant silt and some clay. This information enabled irrigation scheduling and monitoring of plant stress due to soil moisture deficiency in the field while controlling water availability between field capacity, thereby avoiding inefficiencies in water application to the orchard (Figure 6).

The results of the calculated crop evapotranspiration (ETc) and estimated reference evapotranspiration (ETo) are presented in Figure 7 and show a downward trend of mm/day in the winter months and an upward trend until July–August, consistent with the development of the tree’s foliage. ETo (Reference Evapotranspiration), which is climate-driven, represents the atmospheric demand for water based solely on climate factors (sun, wind, humidity). This analysis quantifies the amount of moisture the air can absorb, estimates the overall drying power of the atmosphere, and indicates that compound heat-flash drought (CHFD) events are particularly water-demanding for the crop. These findings are relevant to planning irrigation schedules using the generated data and to applying an SFT/CSA approach.

The calculated Kc is shown in Figure 8, which indicates maximum values ranging from 1.0 to 1.3 from April to August, consistent with values reported by others [2,3] for the same species (black dots in Figure 8). These Kc values were used to estimate pecan orchard-specific water requirements throughout the growing season, accounting for differences in plant morphology, physiology, and growth stages relative to the reference crop when planning irrigation schedules. The irrigation schedule for this specific real-time Kc is shown in Table 2 and Figure 9, along with an IoT display of soil moisture during the irrigation period.

The net irrigation depth per irrigation event under an SFT/CSA approach is compared with a traditional irrigation schedule, showing the advantage of estimating irrigation calendars using digital data to calculate crop water consumption and potential savings, with a significant difference in orchard water demand (Table 2).

The theoretical model shown in Figure 8 and Table 2 clearly outlines the four classic phenological stages of the crop. An initial Kc of approximately 0.3 is observed, corresponding to the dormancy and bud break stage, dominated by soil evaporation, and an average Kc that reaches a plateau of 0.92 during the stage of maximum leaf cover and fruit filling. Analysis of the observed data reveals distinct patterns with low-demand periods occurring at the initial and final stages of the irrigation calendar. There is a high correlation between the observed values and the theoretical curve during the winter and autumn months (January–March and October–November). The low dispersion suggests that the water balance adequately captured actual evapotranspiration, which is primarily driven by soil evaporation. Significant variability is observed during the development and mid-growth stages, particularly in the summer months (June–September). Negative deviations (Kc < 0.6 in June–July) suggest possible episodes of water stress that limit crop transpiration to levels below theoretical potential. Conversely, periods exceeding 1 (Kc > 1.0 in August) are likely attributable to errors in estimating deep drainage following precipitation events, leading to an overestimation of water consumption in the mass balance.

To place the results of this study in a broader context, the irrigation water savings achieved through real-time evapotranspiration monitoring and soil moisture sensing were compared with results reported in previous studies evaluating precision irrigation technologies in different cropping systems (

Table 3. Comparison of irrigation water savings reported in precision irrigation studies.

Study	Crop/System	Technology Used	Water Savings (%)	Key Findings
Pereira et al. (2012) [33]	Irrigated agriculture	ET-based irrigation scheduling	20–45%	Crop evapotranspiration improves irrigation management
Evans and Sadler (2008) [34]	Various crops	Soil moisture sensors + automated irrigation	20–40%	Precision irrigation improves water-use efficiency in irrigated systems
Fereres and Soriano (2007) [35]	Orchard crops	Deficit irrigation strategies	15–35%	Controlled water stress can reduce water consumption without yield losses
Kim et al. (2008) [36]	Irrigated crops	Wireless sensor network irrigation system	~30%	Real-time monitoring improves irrigation decision-making
Fernández et al. (2018) [37]	Orchards	Soil moisture monitoring + precision irrigation	25–40%	Sensor-based irrigation increases water productivity
This study	Pecan orchard	ET monitoring + soil moisture sensors + IoT platform	>50%	Real-time irrigation scheduling significantly reduces water demand

).

4. Discussion

Groundwater represents the most critical water resource for agricultural production in northern México, particularly within the transboundary aquifer systems shared with the southwestern United States. In these semi-arid regions, irrigated agriculture accounts for most groundwater withdrawals, and crops such as pecan orchards play a central role in regional economic development. However, rising agricultural water demand and prolonged droughts have intensified pressure on these shared aquifers. Previous studies have emphasized that improving irrigation efficiency is essential for reducing groundwater depletion in arid agricultural systems [2,3]. The results obtained in this study support these findings and demonstrate that integrating real-time evapotranspiration (ETP) estimation with soil moisture monitoring can significantly improve irrigation efficiency in pecan orchards. Traditional irrigation management practices in northern Mexico frequently rely on fixed irrigation schedules or farmer experience rather than on real-time measurements of crop water requirements. As reported in previous studies, such approaches often result in excessive water application, leading to inefficient groundwater use and unnecessary pumping costs [2]. In contrast, the monitoring system implemented in this study integrates meteorological observations, soil moisture sensors, and an IoT-based data platform to estimate real-time evapotranspiration and crop water demand. By adjusting irrigation schedules based on measured soil moisture and atmospheric demand, irrigation volumes were reduced by more than 50% without adversely affecting vegetative development in a five-year-old pecan orchard.

These results are consistent with previous research demonstrating that precision irrigation techniques based on evapotranspiration monitoring can substantially reduce water consumption while maintaining crop productivity [4]. Maintaining soil moisture levels near field capacity is particularly important in perennial crops such as pecan orchards, as it ensures adequate water availability and sufficient soil aeration for root development. In the present study, increasing irrigation frequency while reducing irrigation duration helped avoid conditions of excessive soil saturation or water stress near the permanent wilting point. This strategy proved effective in reducing plant stress during periods of high atmospheric demand and compound heat–flash drought (CHFD) events. The implementation of IoT-based monitoring technologies also represents a significant advancement compared with traditional irrigation decision-making approaches. Real-time visualization of soil moisture, soil temperature, and climatic conditions allows farmers to respond dynamically to environmental variability. Similar digital agriculture systems have been shown to improve irrigation management and water-use efficiency across several cropping systems [37]. In this study, access to real-time data via an online platform enabled farmers and researchers to continuously monitor soil moisture and adjust irrigation schedules, reducing reliance on subjective assessments based solely on farmers’ experience.

Despite these advantages, several limitations must be considered. Soil moisture measurements can vary significantly depending on soil texture, bulk density, and spatial variability across agricultural fields. Consequently, careful calibration and strategic placement of soil sensors are necessary to ensure that measurements accurately represent field conditions. Previous studies have also highlighted that point-based soil moisture measurements may not fully capture field-scale variability in heterogeneous soils [37]. Additionally, the installation and maintenance of digital monitoring infrastructure may represent financial and technical challenges for small-scale farmers, particularly in rural areas with limited technical support. Another limitation concerns the technological infrastructure required to support real-time monitoring systems. Installing computing facilities and communication equipment on farms may involve high initial costs. However, increasing Internet connectivity in rural regions of México provides opportunities to overcome these barriers. In the present study, the IoT monitoring platform had an effective communication radius of approximately 5 km, potentially allowing soil sensor signals from nearby irrigation fields to be received and processed by the central computing system at CeTraTecAI. This capability suggests that shared digital infrastructure could support regional irrigation-monitoring networks, enabling multiple farms to benefit from a centralized data-processing system.

Beyond farm-scale irrigation management, the results of this study have broader implications for transboundary groundwater management. Many aquifers in northern Mexico and the southwestern United States are hydraulically connected, and groundwater extraction in one region may influence water availability across the international boundary. Improving irrigation efficiency in agricultural areas that depend heavily on groundwater can therefore help reduce pumping rates and slow the depletion of shared aquifer systems. The integration of real-time evapotranspiration monitoring and precision irrigation technologies may thus represent an important strategy for promoting sustainable groundwater use in transboundary basins. The original contribution of this research lies in integrating real-time evapotranspiration estimation, soil moisture monitoring, and IoT-based data sharing within a transboundary water management context. While previous studies have evaluated precision irrigation technologies independently, relatively few have examined their potential to mitigate groundwater depletion in shared aquifer systems. By demonstrating that digital monitoring tools can significantly reduce irrigation water demand in pecan orchards, this study contributes to the development of technological solutions that support both agricultural productivity and groundwater conservation. Furthermore, implementing a cloud-based data-sharing platform represents a step toward collaborative water management, providing remote access to real-time measurements of soil moisture, soil temperature, and soil salinity, enabling communication among researchers, farmers, and water managers. Such platforms can facilitate knowledge exchange and promote the adoption of precision irrigation practices across agricultural communities facing increasing water scarcity. Overall, the findings highlight the potential of digital agriculture technologies to improve irrigation efficiency, reduce groundwater extraction, and enhance the resilience of agricultural systems in water-scarce environments. Future research should focus on expanding the monitoring network across multiple orchards, integrating groundwater monitoring data, and evaluating the long-term effects of optimized irrigation management on aquifer sustainability and agricultural productivity.

5. Conclusions

The need to integrate SFT/CSA technologies into food production is undeniable, making agricultural activities a viable, scalable, and effective approach to a sustainable pathway that reduces water consumption and improves irrigation efficiency. This, in turn, promotes groundwater conservation in transboundary, drought-prone regions such as northwestern Chihuahua, México, and the southwestern United States. The integration of technology tools, such as real-time soil moisture sensors, localized climatic data, Internet of Things (IoT) infrastructure, and evaporation-based irrigation scheduling, as presented in this study, can help address structural inefficiencies in traditional irrigation practices in the research region. Preliminary results of this research from a five-year pilot study implemented on a pecan orchard within the Palomas-Guadalupe Victoria Aquifer have improved irrigation water savings above 50% compared to the traditional volume use with flood irrigation without applying an SFT/CSA approach for water conservation. These savings could be achieved when irrigation decisions are aligned with actual crop water demands rather than fixed schedules or empirical experience alone. The practice of automated, continuous monitoring of soil moisture, along with real-time estimation of reference and crop evapotranspiration (ETo and ETc), enables a more frequent and flexible irrigation schedule with smaller volumes, keeping soil moisture near field capacity and avoiding both water stress and excessive saturation. This, in turn, will not only reduce groundwater pumping but also lower associated energy consumption, supporting healthy vegetative development across the orchard’s different growing stages, demonstrating how water conservation can be achieved without compromising productivity.

We have also generated a localized database of estimated crop coefficients (Kc) for arid and semi-arid environments, derived from our measurements, showing that the observed seasonal dynamics of Kc values fluctuated in accordance with the established phenological patterns in the pecan orchard. We also note deviations from the associated compound heat-flash drought (CHFD) events, which are representative of extreme summer conditions. These findings reinforce the value of real-time, site-specific monitoring to adjust irrigation management when extremely variable climatic conditions are a factor, if generalized or statistical coefficients are not representative of a specific region. Furthermore, we also highlight the emphasis of the broader socio-environmental relevance the application of SFT/CSA for transboundary aquifers has, such as in the study area where agriculture remains as the main economic activity in which groundwater is the key source of water and where well permits are unevenly distributed and aquifer depletion threatens long-term water scarcity for all users, including agriculture and domestic uses. Therefore, while improving irrigation efficiency represents a “win-win” strategy, as it reduces pressure on shared groundwater resources while promoting the economic resilience of small farmers, this growing interest in understanding and adopting such technologies underscores the role of transfer and capacity building in scaling up water-saving practices.

We also demonstrated the feasibility of data sharing through a cloud-based structure for binational collaboration between project partners, where the use of an IoT platform allowed for the sharing of digital “data-lakes” with real-time access to climatic and soil moisture information, promoting transparency and sharing of learning experiences while generating comparative analysis across the U.S.-México border. These data-sharing infrastructures promoted not only on-site decisions but also evaluated regional water-planning options, groundwater monitoring, and the development of cooperation and coordination strategies for the regional transboundary aquifer. Nonetheless, even though the results are promising, several limitations remain. We still need to evaluate the economic costs and benefits of adopting technology, as well as the long-term impact on yields and farm profitability, and the social factors that might influence farmers from accelerating the adoption of SFT/CSA tools. Likewise, soil heterogeneity and soil sensor calibration might be important for site-specific implementation and technical support. Future research should therefore focus on comprehensive cost-specific analysis, technology adoption dynamics, and integration to evolve toward a machine-learning and artificial-intelligence scenario that further automates irrigation scheduling and predicts water stress with greater accuracy under evolving climatic scenarios.

Data-driven irrigation management, implemented through real-time decision-making via SFT/CSA, is a powerful tool and strategy for addressing groundwater depletion, evaluating climate variability, and promoting agricultural sustainability in transboundary arid ecosystems. Drought frequency, heat extremes, intensifying climate change, and water competition will demand investments in smart irrigation technologies and hands-on farmer training, and cross-border collaboration will be crucial to ensure resilient food production systems and the long-term sustainability of shared groundwater resources. This research presents a novel approach to improving irrigation efficiency and groundwater conservation in semi-arid agricultural systems by integrating real-time evapotranspiration estimation, soil moisture monitoring, and Internet of Things (IoT) technologies within a transboundary groundwater context. While previous studies have investigated evapotranspiration-based irrigation scheduling or soil moisture monitoring independently, the originality of this work lies in the development and application of a fully integrated digital monitoring framework designed to optimize irrigation practices in pecan orchards located in groundwater-dependent agricultural regions of northern Mexico. One of the primary contributions of this research is the implementation of real-time evapotranspiration (ETP) estimation, combined with in situ soil moisture monitoring, to dynamically regulate irrigation scheduling. In many agricultural areas of northern Mexico, irrigation management still relies on fixed irrigation intervals or empirical knowledge derived from farmer experience. Such practices frequently lead to over-irrigation and inefficient groundwater extraction. By integrating meteorological observations, soil sensors, and cloud-based data processing, this study demonstrates that irrigation volumes can be adjusted to match the crop’s actual water demand at different stages of development. The results indicate that this approach can reduce irrigation water use by more than 50% in a pecan orchard without compromising vegetative growth during the middle stage of orchard maturity.

A second innovative aspect of this work is the deployment of an IoT-based monitoring system that enables real-time visualization and remote access to soil moisture, temperature, and salinity data. This technological framework facilitates data sharing among researchers, farmers, and water managers and represents a shift from traditional irrigation management to data-driven decision-making. The implementation of a cloud-based platform enables continuous monitoring of field conditions and provides a practical tool for supporting irrigation decisions under variable climatic conditions, including extreme heat and compound heat–flash drought (CHFD) events. Such digital integration is particularly important in regions where farmers often lack access to advanced monitoring tools or technical expertise. Another important contribution of this research is its application to transboundary groundwater management. Many aquifers in northern Mexico are hydraulically connected to aquifer systems in the southwestern United States, and agricultural pumping in these regions directly influences groundwater availability on both sides of the international border. By demonstrating that precision irrigation technologies can substantially reduce groundwater withdrawals at the farm scale, this study provides evidence that digital agriculture solutions may play an important role in mitigating groundwater depletion in transboundary aquifer systems. This perspective extends the research’s relevance beyond farm-level irrigation management and highlights its implications for regional water sustainability and binational water resource governance. The study also contributes to the emerging field of smart farming technologies for water-scarce environments. By combining environmental sensing, cloud-based data storage, and automated data-processing workflows, the research demonstrates how digital infrastructure can create a scalable agricultural monitoring system. The IoT platform’s ability to transmit sensor signals over approximately 5 km suggests that centralized computing facilities could support irrigation monitoring across multiple farms, potentially creating regional irrigation networks that improve water-use efficiency at the landscape scale. Although the results of this study demonstrate the potential of real-time evapotranspiration monitoring and soil moisture sensing to significantly improve irrigation efficiency in pecan orchards, several limitations should be acknowledged. First, the monitoring system was implemented in a single reference orchard at the midstage of its development (approximately 5 years of age). As pecan trees mature, their canopy structure, root depth, and water requirements increase substantially, which may influence evapotranspiration rates and irrigation demand. Therefore, the results obtained in this study may not fully represent the long-term water requirements of fully mature orchards. Future research should evaluate the performance of this irrigation management framework across orchards of different ages and production stages to better understand how crop water demand evolves over time.

Another limitation relates to the spatial variability of soil properties within agricultural fields. Soil moisture measurements obtained from point-based sensors may not fully represent the heterogeneity of soil texture, structure, and hydraulic properties across the orchard. Variations in soil characteristics can influence water retention capacity and infiltration rates, potentially affecting irrigation efficiency if sensor placement does not adequately capture field-scale variability. Future studies should consider deploying multiple soil moisture sensors across different soil zones to improve the spatial representation of soil water conditions and refine irrigation scheduling strategies. In addition, implementing IoT-based monitoring systems requires technical infrastructure, including sensors, communication networks, and cloud-based data processing platforms. Although Internet connectivity is increasingly available in rural areas of Mexico, installing and maintaining such systems may still pose financial and technical barriers for small-scale farmers. Future efforts should focus on developing cost-effective monitoring systems and training programs that facilitate the adoption of precision irrigation technologies among agricultural producers. Another important limitation is that this study primarily evaluated the impact of optimized irrigation scheduling on water use reduction and vegetative growth, but it did not assess long-term crop yield responses or nut production under the modified irrigation regime. Because pecan productivity is influenced by both water availability and climatic variability, future research should evaluate the effects of evapotranspiration-based irrigation management on yield performance, nut quality, and long-term orchard productivity. From a hydrological perspective, this research focused on irrigation efficiency at the farm scale, but the broader implications for groundwater sustainability were not directly quantified. Although reduced irrigation demand likely decreases groundwater extraction, additional studies are needed to evaluate how widespread adoption of precision irrigation technologies may influence groundwater levels and recharge dynamics within transboundary aquifer systems. Integrating irrigation monitoring with groundwater monitoring networks and hydrological modeling could provide a more comprehensive assessment of the long-term benefits of these technologies for aquifer conservation.

Future research should also explore integrating additional environmental monitoring tools, such as remote sensing data, satellite-based evapotranspiration estimates, and machine learning algorithms, to support predictive irrigation management. Combining these technologies with ground-based sensor networks could enhance the accuracy of evapotranspiration estimates and improve the scalability of precision irrigation systems across larger agricultural regions. Despite these limitations, the results of this study highlight the significant potential of integrating real-time evapotranspiration monitoring, soil moisture sensing, and IoT-based data platforms to improve irrigation efficiency and reduce groundwater extraction in semi-arid agricultural regions. Expanding the implementation of these technologies across multiple orchards and agricultural communities could contribute to more sustainable water management practices and support the long-term resilience of agriculture in groundwater-dependent regions.

Finally, the research advances the practical implementation of precision irrigation strategies for perennial orchard systems, which are often underrepresented in the literature compared with annual crops. Pecan orchards require careful irrigation management due to their deep root systems, long production cycles, and sensitivity to water stress during key phenological stages. By demonstrating that real-time evapotranspiration monitoring can effectively regulate irrigation in a developing pecan orchard, this study provides valuable insights for improving irrigation practices in similar perennial cropping systems in semi-arid regions. Overall, the novelty of this research lies in integrating real-time evapotranspiration monitoring, soil moisture sensing, IoT-based data management, and groundwater conservation strategies within a transboundary agricultural context. This multidisciplinary approach not only advances scientific understanding of precision irrigation systems but also provides a practical framework for improving water-use efficiency and supporting sustainable agricultural development in water-scarce regions.