论文信息

Muhammad Rasool Al‐Kilani / Mohammad A. Okour / Qasem Abdelal 等 6 人

DYSONA – Applied Science被引 1 次

论文解读

AI 辅助整理
方法
30个地块的传感器测量、降雨、用水与能耗数据,核算水碳足迹。
发现
智能灌溉每100m²节水30m³并净减碳5475.6kg,集雨碳足迹最低但节水有限。
机会点
可结合减少蒸发等作物需水管理措施,量化更大尺度的水碳协同减排潜力。

论文摘要

节水需求日益迫切,但关于现有管理方案对碳排放影响的农场级数据却十分匮乏。本研究基于传感器测量、降雨数据、用水量和能耗信息,呈现了约旦马弗拉克一个拥有30个作业地块的农业综合体的相关洞察。年用水量为20,000 m³,碳足迹为3.47 kg eCO₂/m³。智能灌溉根据灌溉超过土壤田间持水量的频率,对应每个100 m²单地块季节性节水30 m³。综合考虑智能灌溉所需电子电气组件隐含碳(0.09 kg eCO₂/m³),这在所有地块上可折算为水碳足迹净减排5475.6 kg eCO₂。雨水收集因该地区干旱特性而节水潜力最小(260.6 m³),但因至灌溉区的运输环节,其碳足迹也最低(0.03 kg eCO₂/m³)。处理废水回用的可行性受限于所需辅助基础设施及处理阶段的隐含碳足迹,使其相对其他可用策略竞争力较弱。结果表明,现场技术可实现碳可行的节水,但管理作物需水(如减少蒸发)可能带来更大的水-碳减排。

全文

DYSONA – Applied Science

E-NAMTILA P ublishing DYSONA - Applied Science
DAS 8 ( 2027) 1-16 DOI: 10.30493/DAS.2026.010308
1
All published articles in DYSONA - Applied Science
journal are distributed under a Creative Commons Attribution 4.0
International License. © 2026 by the author(s).
DYSONA Applied Science ISSN: 2708-6283
Assessing water-carbon tradeoffs of smart
irrigation and non-conventional water
sources using sensor measurements and
water use information
Muhammad Rasool Al-Kilani 1*; Mohammad A. Okour 2; Qasem Abdelal 3; Khalid Bani-Melhem 4;
Rami Abu-Mousa 2; Temitope Samuel Egbebiyi 5, 6
1, Department of Smart Irrigation and Agriculture, Faculty of Agriculture, Irbid National University, Irbid, Jordan
2, Department of Research and Development, Biotechnology Research Unit, Oula for Seeds Production Company, Mafraq, Jordan
3, Department of Civil & Environmental Engineering, School of Sustainable Systems and Engineering, German Jordanian University,
Amman, Jordan
4, Center for Advanced Materials (CAM), Qatar University, P.O. Box 2713, Doha, Qatar
5, Department of Earth and Atmospheric Sciences, Cornell University, Ithaca, NY, USA
6, Climate System Analysis Group, Environmental and Geographical Science Department, University of Cape Town, South Africa
Ab stract
E-mail:
m.kilani@inu.edu.jo
R
eceived: 14/04/2026
Revised: 25/07/2026
Acceptance: 01/08/2026
Available Online: 03/08/2026
Published: 01/01/2027
The growing imperative for water conservation contrasts with a paucity of
farm‑level data on the concomitant carbon implications of available
management options. This work presents such insights from an agricultural
compound with 30 operational plots in Mafraq, Jordan using sensor
measurements, rainfall data, water use, and energy consumption information.
The annual water consumption was 20,000m 3 at a carbon footprint of 3.47 kg
eCO2/m3. Smart irrigation corresponded to seasonal water savings of 30m 3 per
single plot of 100m2, based on how often irrigation exceeded soil field capacity.
This scales to a net reduction of 5475.6 Kg eCO 2 of the water carbon footprint
across all plots, accounting also for imbodied carbon (0.09 Kg eCO 2/m3) from
the electronic and electrical components needed in smart irrigation. Water
harvesting showed the least water saving potential (260.6 m 3) due to the arid
nature of the region, but also the lowest carbon footprint (0.03 Kg eCO 2/m3)
resulting from transportation to irrigation area. The feasibility of treated
wastewater reuse at the site was constrained by the requisite ancillary
infrastructure and the embedded carbon footprint of the treatment phase,
rendering it less competitive relative to other available strate gies. The results
demonstrate how on‑si te techniques can have carbon ‑via ble water savings, but
managing crop water demand, like reduced evaporation, could result in much
larger water‑carb on reductions.
Keywords: Smart irrigation,
Water conservation, Carbon
footprint, Rainwater
harvesting

  1. Introduction
    Climate change is expected to have significant impacts on agricultural production systems around the world, and
    rainfall variability may have the most direct impacts on crop yields [1]. Under these conditions, it is expected that a
    20% increase in the fraction of irrigated lands is necessary to meet global food demand [2]. Researchers in vulnerable
    regions have largely focused on reducing the pressure on fresh water through exploring technological advancements
    in non‑co
    nventional water resources such as treated wastewater reuse and desalination [3][4]. Research efforts have
    also been directed towards improving monitoring of water resources and water accounting through novel techniques
    [5][6]. However, improving water use efficiency and reducing losses remains to be a priority, particularly in arid

DYSONA – Applied Science 8 (2027) 1-16 Al-Kilani et al.
2

regions with scarce water resources [7]. In this context, agriculture is a major focal point as the largest global water
consumer.
Over the past decade, studies have shown that smart irrigation systems, combining soil moisture sensing, weather ‑
based scheduling, and data ‑driven control can significantly reduce agricultural water use while maintaining or
enhancing productivity [8][9]. These technologies enable irrigation to be scheduled according to real‑time crop needs,
thereby minimizing overwatering, runoff, and evaporation losses [10]. A substantial body of research published over
the past decade shows that smart irrigation can reduce water consumption by approximately 20– 40% under typical
field conditions, with higher savings; sometimes exceeding 50% achieved in optimized or water ‑scarce environments
[11][12]. These gains are largely driven by the system’s ability to respond dynamically to changing environmental
conditions, such as rainfall, temperature, and soil variability, which are often overlooked in traditional irrigation
practices [12]. In addition to conserving water, smart irrigation can maintain or even improve crop yields by
preventing both water stress and excess irrigation, leading to better overall water use efficiency. Early studies on
sensor‑based irrigation, with the integration of IoT and artificial intelligence, already showed notable improvements
in water ‑use efficiency compared to conventional scheduling, particularly through reducing over‑ irrigation and
matching application to crop demand, with typical water savings ranging from 20‑ 40% under field conditions and
reaching 30– 50% in optimized or drought ‑prone environments [13]. Experimental and modelling studies highlight
that real ‑time monitoring and adaptive control are key drivers of these gains, enabling systems to respond
dynamically to soil moisture deficits, evapotranspiration, and weather variability [14]. Advanced approaches using
machine learning, multi ‑sensor fusion, and predictive analytics have reported water savings of up to 30 –50%
compared to traditional irrigation, alongside improvements in water productivity and yield stability [15]. Also, other
studies emphasize that smart irrigation reduces uncertainty in irrigation decision‑making and minimizes losses due to
runoff and deep percolation [16]. However, research has shown that actual savings are context‑ dependent, influenced
by crop type, climate, soil properties, and system design. Overall, there is strong consensus that smart irrigation
represents a critical technological pathway for enhancing water ‑use efficiency and supporting sustainable agriculture
under increasing water scarcity and climate variability.
Extensive advancements and insights have been revealed on the role of different management systems in water
conservation, such as wastewater treatment, water harvesting, and smart irrigation [17‑ 19]. This also extended to life
cycle analysis of these various options, revealing insights related to their carbon footprint and environmental impacts
[20‑23]. However, a knowledge gap remains on the tradeoff between these different systems. There have been no
experimental works comparing the potential for these water conservation pathways on a single site, or across sites
under similar conditions and management. Accordingly, this work explores the water conservation potential of smart
irrigation compared to alternative water conservation options at an agricultural compound in a semi ‑arid region in
Jordan. The agricultural compound hosts numerous cropping plots under similar management conditions. Such
experiments can support decision making on the choice of water conservation technology. Specifically, they can
promote improved understanding and commitment to targets related to several sustainable development goals (SDG)
including SDG 13 ‘Climate Action’ and SDG 15 ‘Life on Land’.
2. Material and Methods
2.1. Study area
This study was conducted at an agricultural compound in the rural areas of Mafraq, Jordan (Fig. 1). At this site, 30
farm plots are commonly operational, these comprise greenhouses and open fields at non ‑fixed ratios. The
arrangement of greenhouses and open fields changed every season based on the crops of choice for that season and
based on the production schedule for the operators. Different types of crops are grown at the site, which vary by
season, but are mainly comprised of vegetables. The region is c haracterized by a semi ‑arid Mediterranean climate
pattern, with wet cold winters, warm summers, and most rainfall occurring between November and March.
Meteorological records for the study site show that annual rainfall was around 215 mm, and average temperature was
around 18.3oC. National soil maps show that the area is dominated by Regosols; however, the soil at the specific site
for this study was reported as Calcisols. Water sources at the site included rainwater harvesting and purchased water
from local suppliers (licensed groundwater reselling points).

DYSONA – Applied Science 8 (2027) 1-16 Al-Kilani et al.
3

Figure 1. Study area comprising the agricultural compound where the experiments and data
collection was done for this work (Satellite image taken during the site-preparation phase).
2.2. Soil and irrigation assessment
Data relevant to soil and irrigation assessments includes a wide range of physical and chemical properties such as
water‑h
olding capacity, fertility, salinity, and texture [24]. However, obtaining all these variables for a specific site can
be costly, and are rarely readily available [25]. The current analysis included the minimal data needed to enable
precise irrigation scheduling, zone‑s
pecific water application, and identifying any notable soil issues. For that purpose,
soil samples were obtained from different points at the site. The samples were collected from the top layer (30cm) of
the soil, which was a suitable depth to represent root growth conditions for vegetables. Air ‑d
ried soil samples were
sieved using a 2 mm mesh and homogenized prior to analysis. Chemical properties were determined using a
standardized soil‑w
ater extract method (Fig. 2). Field capacity (FC) was determined as the amount of water remaining
in the soil after three days from saturation, while permanent wilting point was assumed to comprise 50% of FC for the
purposes of this work. Soil analysis that was done in the lab included texture, bulk density, pH, and salinity. Irrigation
water quality was done using samples from the emitters to account for any effects from the irrigation network, the
samples were collected during the irrigation uniformity test. The results of the soil and irrigation system assessments
are shown in Table 1 and Table 2, respectively.
Table 1. Physical and chemical soil properties determined for the study area
Parameter Range Notes
Texture Silty loam balanced texture that supports favorable water holding capacity, workability, aeration, and drainage
Bulk density (g/cm3) 1.1 – 1.2 Moderate values, consistent with medium texture
Field capacity (FC) (%) 33 ‑37 Amount of water the soil can hold against gravity (Ideally not exceeded)
Plant welting point (PWP) (%) 17‑19 Estimated based on FC
pH 7.8 – 8.4 Relatively good compared to common soils in Mafraq which range around 8.3
Salinity (µS/cm) 433‑512 Low salinity; minimal risk to plant health and nutrient uptake.

DYSONA – Applied Science 8 (2027) 1-16 Al-Kilani et al.
4

Table 2. Irrigation assessment report for the study area
Parameter Value Assessment details
Irrigation system type Drip irrigation / inline Excellent choice for the selected crop High water conservation ability
Water storage system Closed water tank at a higher altitude Excellent choice to avoid system failure risks Avoids water losses from evapotranspiration
Water conservation options Mulch application Excellent choice to avoid pests and reduce water losses
Irrigation layout
Lateral lines connected
(16mm) connected to a
subline (3 Inches)
Ideal for the production system
Irrigation uniformity 87.7% Excellent, ideal for implementing sensor‑based irrigation scheduling
Pressure in the network <0.6 bar * Suboptimal, the experiments were held following maintenance to improve pressure
Flowrate ~4 L/h Ideal for the crop of choice (eggplant) at the experiment plot.
*Increased to 0.9 following maintenance prior to conducting experiments.

Figure 2. Soil sampling (A) and analysis (B and C), and irrigation assessment (D and E) done for the study site

DYSONA – Applied Science 8 (2027) 1-16 Al-Kilani et al.
5

2.3. Sensor assessment
Prior to selecting a sensor to conduct the irrigation
experiments, several types were evaluated to determine
the most suitable choice. Previous researchers have
shown that different soils can have different sensor
performances [26]. Thus, site
‑s
pecific assessment is
crucial in selecting a proper sensor.
O
verall, three types of sensors were evaluated: SMT100,
T
‑H
igrow, and TZT Module. Each model has a unique
price range, measurement mechanism, and dimensions,
but they are all widely available in the market. For each
sensor type, 3 sensors were tested to confirm sensor ‑to‑
s
ensor variability. The SMT100 is a time domain
transmission (TDT) sensor manufactured by Truebner
GmbH in Germany (Fig. 3 A). The T ‑Hi
grow sensor is a
capacitance‑b
ased soil moisture probe developed by
LilyGo in China (Fig. 3 B), which integrates an ESP32 microcontroller, allowing for wireless data transmission and
easy integration into IoT systems. The TZT soil moisture module (Fig. 3 C) is a resistivity ‑b
ased sensor produced by
TZT, a Chinese electronics supplier.
The sensors’ readings were assessed based on actual soil moisture in samples prepared in the lab. Three samples with
10%, 20%, and 30% soil moisture were prepared by adding controlled volumes of water to dried soil samples from
the experimental site. For each assessment run, three sensors were used to examine repeatability and reproducibility
across different moisture ranges. Evaluation criteria included measurement error, correlation, repeatability, and
reproducibility. Measurement error was represented by mean absolute error (MAE), which expresses the average of
the absolute differences between the sensor readings (y i) and actual soil moisture (ŷ i) across the samples (n), is
calculated based the following equation:
MAE = 1n �|yi − y ı�|
n
i=1

Correlation refers to the statistical relationship between the sensor’s readings (x) and the actual soil moisture content
(
y). Correlation was represented by Pearson’s correlation coefficient (r) based on the following equation:
r = 1n − 1 � �x − x�Sx � . �y − y�Sy �
Repeatability refers to a sensor’s ability to produce consistent measurements when the same procedure is repeated
under identical conditions. High repeatability indicates more reliable readings in controlled environments, which is
crucial for detecting small changes in soil moisture over time. Reproducibility, on the other hand, assesses how
consistent the sensor’s measurements are when conditions vary (different operators, different instruments of the
same model, or different locations), reflecting the sensor’s robustness and reliability across broader applications.
2.4. Water carbon assessment
This assessment included a breakdown of all operations and relevant energy consumption related to irrigation. The
entire process (from the water source to irrigation) was identified, and the energy and fuel use associated with all
pathways was identified. The contribution of smart irrigation to water conservation was done using a data
‑d
riven
approach. Smart irrigation would conserve water by avoiding the addition of excess water beyond field capacity (FC),
after which water is not retained in the rootzone and not beneficially used by plants. Thus, soil moisture
measurements were used to calculate the impact smart irrigation would have in conserving water. Excess irrigation
(ExcIrr) from a single irrigation application, which would have been saved through smart irrigation, was estimated
through the following equation:
Figure 3. Soil moisture sensor types tested for this study:
the SMT100 sensor (A), the T-Higrow sensor (B), and the
TZT conductivity sensor (C)

DYSONA – Applied Science 8 (2027) 1-16 Al-Kilani et al.
6

�
𝜃𝜃𝑣𝑣 > 𝐹𝐹𝐹𝐹, 𝐸𝐸𝐸𝐸𝐸𝐸𝐼𝐼𝐼𝐼𝐼𝐼(𝑚𝑚3) = 𝜃𝜃𝑣𝑣 𝑚𝑚𝑚𝑚𝑚𝑚−𝑑𝑑 − 𝐹𝐹𝐹𝐹∆𝜃𝜃𝑖𝑖𝐼𝐼𝐼𝐼 × 1𝑚𝑚3.
(𝜃𝜃𝑣𝑣 ≤ 𝐹𝐹𝐹𝐹, 𝐸𝐸𝐸𝐸𝐸𝐸𝐼𝐼𝐼𝐼𝐼𝐼(𝑚𝑚3) = 0

Where θ v max ‑d is the maximum volumetric soil moisture
recorded during a day in which irrigation was applied, and
Δθirr is the amount of soil moisture increase from adding
1m3 of irrigation (Fig. 4). A preliminary controlled
experiment showed that adding 1m 3 of irrigation water to
the experimental site increased θ v sensor measurement by
12% (±1%). Accordingly, it was possible to link sensor
measurements from irrigation monitoring experiments to
the amount of excess water that was added. Four SMT100
sensors were then installed in the monitoring plot, one at
each row. The experimental plot was under greenhouse
conditions and normal irrigation management done at the
site by the same team that normally manages it. This
approach was used to determine the contribution of smart
irrigation over existing management. This procedure of
examining water added beyond FC would show the amount
of water saved assuming an ideal smart irrigation
application, since the water saved from smart irrigation
could be low on the first tests after installation, but would
be much higher after repeated tests, optimization, and
integration into existing management.
To assess the contribution of water harvesting to water conservation and reducing the related fuel and energy use, the
seasonal rainfall for the past 3 years was obtained for Mafarq. The seasonal rainfall was multiplied by the total surface
area of the rooftop designated for rainwater harvesting (~1500m 2), assumed to be 90%. These assumptions were
based on feedback from the maintenance team to avoid overestimation. The water conservation from potential
wastewater treatment was determined based on the amount of freshwater used for non ‑irrigation purposes. This was
the fraction of water that can be retrieved as grey or black water, while irrigation was assumed to be non ‑retrievable
due to lack of drainage or other recovery systems. Besides smart irrigation and non ‑conventional sources, the amount
of water saved from using greenhouses was also considered to show how all these systems compare to a reduction in
evaporation and reduced crop water needs. The amount of water saved from using greenhouses (Reduced
evaporation) was determined based on existing records of irrigation consumption at greenhouses and open fields. The
fuel consumption associated with transporting 1m 3 of fresh water from local suppliers was calculated based on
distance from the water source (D), fuel consumption per trip (Fuel trip), volume of water delivered per trip (Water del),
and transportation efficiency (Efftrs) as follows:
Fuel per water (L m3⁄ ) = 2 × D × Fueltrip
Waterdel × Efftrs

Water‑energy use (WEU) (kW.hr/m 3) at the delivery site (selling point) was calculated using the hydraulic power
formula based on the density of water (ρ), gravitational acceleration (g), total dynamic head (TDH), and pump
efficiency (η):
WEU (kW. hr/m3) = ρ × g × TDHη × 3.6 × 106
The local water supplier was estimated to have a pumping efficiency of 70% and a water depth of 400m, based on an
assessment by local expert engineers and utilizing information such as pump model and date of most recent
installation.
Figure 4. Illustration of the protocol sensor
measurements interpretation to derive excess water from
applied irrigation “ExcIrr” (FC: Field capacity).

DYSONA – Applied Science 8 (2027) 1-16 Al-Kilani et al.
7

The carbon footprint as equivalent CO 2 in kg (eCO 2) for different components was calculated using national
conversion factors for each water, energy, and fuel type and based on values reported in the literature (Table 3).
Wastewater treatment, as a potential option, was assigned values reported from previous works due to lack of values
reported for small‑s
cale decentralized treatment systems in Jordan.
Table 3. Carbon footprint conversion factors used in this study
Parameter Unit Factor (kg eCO2) Notes
Freshwater
1 m3 of freshwater from the
Jordanian national water
grid.
3.47
This is an average value reported for
the carbon footprint resulting from
water pumping, treatment, and supply
through the Jordanian national water
grid [27].
Electrical energy 1 kWh.hr from the Jordanian national gridline. 0.54
Equivalent CO2 emission factor for 1
kW.hr used from the national energy
grid as reported in national reports
[27].
Fuel consumption 1 L of diesel 2.6 Equivalent CO2 emission factor for 1 L of diesel in Jordan [27]
Greenhouses Amount of plastic to cover a 200m2 greenhouse 200
Based on a previous life cycle analysis
in the literature. This number would be
divided over the lifespan of the plastic
(3 years) and the amount of water a
single greenhouse saves [28][29].
Wastewater treatment Per 1m3 of wastewater 1
Average value estimated from the
literature, assuming primary and
secondary treatment sufficient to meet
irrigation standards [30
‑3
3].
Grey water treatment Per 1m3 of greywater 0.4
Average value estimated from the
literature, assuming less intensive
treatment sufficient to meet irrigation
standards [30‑33
].

  1. Results
    3.1. Sensor selection results
    Sensor assessment results considered accuracy, represented by correlation, and reliability, represented by
    repeatability and reproducibility (Table 4). Correlation for all sensors was excellent, showing that all sensors
    responded well to changing soil moisture levels.
    H
    owever, standard error, deviation and variance for the TZT sensor were substantially higher across both
    repeatability and reproducibility. This means that the same TZT sensor resulted in different measurements under the
    same conditions, and that different sensors of the same model also produced inconsistent readings. The high
    correlation shows that the readings of the TZT sensor increase with moisture level, but were not precise or
    reproducible. While the SMT100 and T‑H
    igrow sensors showed acceptable performance across all previous indicators
    when the experiment was replicated using the same sensor or using different sensors. This was followed by an error
    assessment, i.e. sensor readings compared to lab ‑m
    easured values (Fig. 5), which demonstrates the suitability of
    SMT100 calibration settings and its measurement principle (time domain transmission) over the T ‑H
    igrow
    capacitance principle. Based on these results, SMT100 was selected for further experiments on the contribution of
    smart irrigation for water conservation under the investigated conditions.

DYSONA – Applied Science 8 (2027) 1-16 Al-Kilani et al.
8

Table 4. Sensor assessment conducted to select the suitable sensor type for irrigation monitoring experiments
Test Sensor Standard Error Standard Deviation Sample Variance Confidence level (95%) Correlation test
Repeatability
test
SMT100 <0.001 <0.001 <0.001 <0.001 0.999
T‑Higrow 0.006 0.01 <0.001 0.02 0.991
TZT 4.91 8.5 72.33 21.13 0.984
Reproducibility
test
SMT100 0.014 0.024 <0.001 0.06
T‑Higrow 0.019 0.032 <0.001 0.08
TZT 16.05 27.79 772.33 69.04

3.2. Smart irrigation prospect
The potential water savings from smart irrigation were determined
through a data
‑dr
iven approach based on sensor measurements of soil
moisture. The readings were used to identify the excess moisture added
to the soil through manually controlled irrigation (Fig. 4). Soil moisture
level was considered in excess if it surpassed FC, which was identified
based on previous soil analysis for the same site where the experiments
were conducted.
T
he results demonstrate that soil moisture was generally below FC, but
never decreased to reach welting point (~18% on average) (Fig. 6). This
meant that only a limited amount of water was lost to drainage, and crops
would generally remain safe from water stress under the current
management (Irrigation without sensor assistance). The time series
overall indicate proper irrigation management, despite the absence of any
feedback mechanisms for the irrigation managers, who did not have
access to sensor readings. To estimate the volume of water that could
potentially be saved through smart irrigation (defined as irrigation that
avoids exceeding field capacity while also preventing the soil from
reaching the wilting point) the time series were used to calculate the
amount of water corresponding to soil moisture exceeding FC. Since the
rootzone depth had not been previously determined, a straightforward
physical calculation was not feasible. Instead, an experiment was conducted to determine the contribution of each
cubic meter of irrigation water to increasing volumetric soil moisture (Fig. 7).
A
nalysis of the full time series demonstrated that a total of 5m 3 could potentially be saved monthly within the
experimental plot (100m 2). This corresponds to 10 ‑15 m3/month for other plots, which range between 200 and
300m2.

Error (mm/mm)
SMT100 T-Higrow
Figure 5. Measurement error of the SMT100
sensor compared to the T-Higrow sensor

DYSONA – Applied Science 8 (2027) 1-16 Al-Kilani et al.
9

Figure 6. A sample of soil moisture readings at the experimental site (for 1 month out of a 3 -
month test, data points were averaged over scale to improve visualization with each peak
standing for an irrigation similar to the visualization in Figure 4)

Figure 7. Volume of irrigation water corresponding to soil moisture measurements and
irrigation events (data shown for a 3-week duration)
3.3. Other water conservation approaches
Total seasonal water usage was recorded to be 20,000m 3, Irrigation usage was estimated to comprise 18,000m 3 and
the rest was allocated for other usage such as cleaning and hygiene maintenance. This provided the first potential
pathway for water conservation, which was treated wastewater reuse (Fig. 8). Another possible supplementary water
source is water harvesting. The storage tank designated for collecting rooftop rainfall is also used for storing water
from other sources, and so, the amount was estimated using seasonal rainfall in the area (Fig. 9).

0%
1%
2%
3%
4%
5%
6%
7%
8%
15%
20%
25%
30%
35%
40%
45%
Excess moisture
Soil moisture readings
Volumetric soil moisture Excess moisture beyond FC
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
0
0.2
0.4
0.6
0.8
1
1.2
7 8 9 10 11 12 13 14 15 16 17 18 19 20
Cumulative excess water (m3)
Excess irrigation water (m3)
Day

DYSONA – Applied Science 8 (2027) 1-16 Al-Kilani et al.
10

Figure 8. Water consumption and wastewater production
potential (m3) assuming a 90% recovery from treatment
Figure 9. Rainfall water harvesting, assuming 90% effective
rainfall for collection

The utilization of greenhouses in the study area was not particularly intended for water conservation, but rather to
achieve ideal temperatures for growth conditions, avoiding genetic contamination of the desired breeds, and
minimizing biotic stress. The main utilization for greenhouses in the region is to avoid the risk of cold stress, due to
minimum temperature dropping below 5 oC at nighttime. However, greenhouses also affect water consumption by
limiting evaporative factors, mainly solar radiation and wind speed, which increase evaporative demand in
accordance with the Pennman ‑M
onteith energy model. While a significant temperature increase is noticed inside
greenhouses, this increase is somewhat buffered by scheduled ventilation around 12:00 pm to avoid heat stress.
I
rrigation water applications did not have a dynamic feedback mechanism
to optimize water addition, i.e. there was no specific mechanism for the
workers to identify crop water demand and apply water accordingly.
Instead, the workers relied on experience, i.e. limiting or increasing water
applied and monitoring the crops’ physiological responses. Thus,
monitoring water consumption on greenhouses and open fields provides
insights into the contribution of greenhouses to water conservation at the
site (Fig. 10). Water consumption was different based on crop type,
growing season, plot area, amongst other factors. The results indicate that
agricultural production in greenhouses requires on average around
160m3, while open fields required around 315m 3 on average. This
indicates that greenhouses require almost half the water requirement on
average, this would vary based on crop type, season, plotting area, and
other factors. However, this observation is not necessarily a consistent
water conservation approach, since greenhouse crops should be tested in
open‑f
ield conditions at the site, and greenhouses are used based on
designated experimental plans.
A
ll the possible options can be compared when considering that 30
agricultural plots are usually active within the company perimeter. This
means that the total water harvesting volume and total treated
wastewater would be separated between these plots. Conversely, water
savings from the greenhouses and smart irrigation (determined at plot
18000
1260
540
2002000
Irrigation Domestic / industrial
Greywater blackwater
Non-recovered
Figure 10. Seasonal water consumption for
the period 2023-2025 in different
greenhouses and open fields at the study
site for different crops for plot sizes of
250m2. Upper and lower limits represent
upper and lower quartiles.

0
50
100
150
200
250
300
350
400
450
500
0
50
100
150
200
250
300
2021 2022 2023 2024 2025
Water volume (m3)
Rainfall (mm)
Water harvesting volume Rainfall (mm)
Seasonal water use (m3)
0
100
200
300
400
500
600
Open field Greenhouse

DYSONA – Applied Science 8 (2027) 1-16 Al-Kilani et al.
11

level) would be multiplied by the number of plots for their common operational duration in the study site (6 months)
(Fig. 11).

Figure 11. Volume of water saved from different water conservation options (assuming 30
operational plots)
3.4. Carbon footprint
The carbon footprint associated with each cubic meter of water savings was determined based on the energy and fuel
consumption corresponding to its use. While this is not holistic, it provides a general idea on the variations between
different water conservation options. The carbon footprint of the water supply from the national grid (3.47 kg
eCO2/m3) was considered as the benchmark.
The carbon reduction from each water conservation method was estimated based on its ability to reduce water
consumption and the embodied carbon footprint of conserved water (Table 5). Water harvesting resulted in moderate
water savings but required transporting the water from the harvesting and storage site to the irrigation tank, which
requires diesel‑fueled delivery tank. Smart irrigation potentially allowed for reducing irrigation water consumption
but was associated with using electricity ‑operated sensors, controllers, and pumps. Energy consumption from
pumping was attributed to smart irrigation because the site did not normally use pumps to maintain pressure. The
pressure was maintained in the network through the hydraulic head of the irrigation storage tank being higher than
the cultivation plots. This hydraulic head is concentrated by a network of valves which deliver the water to one zone
at a time by keeping only that one zone’s valve open, a process which is done manually based on a fixed schedule.
However, with regards to smart irrigation, the process would be triggered and stopped based on sensor readings and
so a dedicated pump would be needed for each zone. With regards to potential savings from wastewater treatment, it
was based on the amount potentially generated on site, while carbon footprint was based on average values obtained
from previous studies. The carbon footprint values for wastewater reuse have the highest uncertainty in this work,
but the volumes were determined for the site, so they can be considered reliable for this work.
The results demonstrate that rainwater harvesting had the lowest contribution to reducing the carbon footprint, while
smart irrigation had the largest potential impact on carbon footprint reduction (Table 5). Despite rainwater
harvesting having the lowest embodied carbon footprint, its contribution to carbon reduction was small due to its low
water savings potential on the site. Wastewater reuse had a larger water savings potential; however, its high
embodied carbon footprint resulted in smaller net reduction compared to smart irrigation option.

0 100 200 300 400 500
0 1000 2000 3000 4000 5000
Greenouses
Treated wastewater reuse
Treated greywater reuse
Smart irrigation
Water harvesting
Water saved per plot (m3)
Total volume of water saved (m3)
Total Per plot

DYSONA – Applied Science 8 (2027) 1-16 Al-Kilani et al.
12

Table 5. Embodied carbon footprint and footprint reduction from different water conservation methods
Conservation method Carbon footprint (kg eCO2/m3) Reduced footprint per m3
(kg eCO2/m3)**
Estimated reduced
water consumption
(m3)
Net emission
reduction
(Kg eCO2)
Smart Irrigation 0.09 3.38 1620 5475.6
Wastewater treatment 1.00* 2.47 1800 4446
Rainwater harvesting 0.03 3.44 260.6 897

  • Estimated based on average values from previous works [31‑33].
    ** Estimated by subtracting carbon footprint from the carbon footprint of the national water supply (3.47 kg eCO2/m3).
    The carbon footprint from greenhouses was not considered in this section, since its construction carbon footprint can
    complicate the calculations.
  1. Discussion
    Limited works have explored the water‑c arbon tradeoffs between different water conservation approaches. This work
    compared smart irrigation to two non ‑c
    onventional water options, rainwater harvesting and wastewater treatment
    and reuse, at a commercial agricultural compound in Northern Jordan. The study utilized soil moisture sensor
    measurements and water use and management information to derive reliable estimates. The carbon footprint of each
    water conservation method was driven mainly by pumping energy, not by treatment chemistry or materials.
    Wastewater treatment and reuse was assigned the highest footprint (1.00 kg eCO2/m 3), largely due to energy ‑
    i
    ntensive aeration during treatment; conventional activated sludge processes require far more energy than anaerobic
    alternatives, which operate at 0.01– 0.05 kWh/m3 [34]. Smart irrigation's footprint (0.09 kg eCO 2/m3) was calculated
    from powering sensors, controllers, and pumps. Rainwater harvesting had the lowest footprint (0.03 kg eCO 2/m3),
    derived mainly from transportation, since collection relied mostly on gravity rather than pressurized pumping. This
    matches previous findings that pumping operations can produce around 6.5 times higher CO2 ‑e
    quivalent emissions
    than mains water delivery, and that pumping accounts for roughly 70% of the net life
    ‑c
    ycle impact of rainwater
    harvesting systems where a pump is used [35]. Recycling water through non ‑p
    umped rainwater harvesting has been
    shown to cut global warming impact by 19 –51% relative to full life ‑cy
    cle emissions [35], consistent with the low
    footprint observed here.
    S
    mart irrigation, on the other hand, showed potential savings of 5m 3 of water per month per 100m2. Over the site's 6 ‑
    m
    onth operational season, this corresponds to roughly 30m 3 per 100m2 per season. This demonstrates how savings
    per plot can be modest, but it accumulated substantially once extended across a full farm or estate. It is important to
    note that these numbers are site
    ‑s
    pecific and only reflect the water management conditions at the site. However, some
    aspects would likely be reflected in other similar sites, such as how smart irrigation gave the largest net reduction in
    carbon footprint, even though its per ‑m3 footprint was not the lowest. This agrees with a previous study on solar ‑
    p
    owered smart irrigation, which reported a 28.1% reduction in combined water and energy consumption compared
    to conventional irrigation, along with a drop in carbon footprint from 0.252 to 0.181 kg CO 2/m2/year [36]. This
    suggests that water savings, not the footprint of the technology itself, mainly drives the climate benefit. Rainwater
    harvesting's small contribution to overall carbon reduction reflects the same pattern: even where a commercial
    rainwater harvesting system was shown to outperform municipal water supply in nearly every environmental impact
    category, its advantage depended on the system's energy intensity staying below a specific threshold (an energy
    saving of at least 0.86 kWh/m3 relative to the municipal baseline) [37]. This highlights how sensitive its benefit is to
    water volumes and system design rather than to the technology's inherent footprint. Wastewater reuse had greater
    water‑s
    aving potential than rainwater harvesting, but its higher footprint reduced its net benefit. This is similar to
    previous reports showing that demand ‑s
    ide water conservation could improve wastewater treatment plant eco ‑
    e
    fficiency by up to 189%, reducing GHG emissions by 1.67 million tons CO2 ‑e
    quivalent nationally compared to relying
    on treatment‑te
    chnology improvements alone [38].
    S
    everal uncertainties should be noted regarding the findings presented in this work. The wastewater reuse footprint
    was based on published average values rather than site measurements, making it the largest source of uncertainty;
    treatment energy use varies with technology, influent load, and electricity mix, ranging from as little as 0.01 kWh/m 3
    for anaerobic systems to substantially more for conventional aerobic treatment [34]. Recent advancements have also

DYSONA – Applied Science 8 (2027) 1-16 Al-Kilani et al.
13

shown that treatment process can be optimized to produce minimal environmental impact through circular
applications [39]. Smart irrigation water savings were based on sensor estimates, which may not capture seasonal or
crop‑specific variation. Additionally, performance and calibration functions will vary according to soil type and sensor
module [40], highlighting the importance of site/plant type calibration in each case. Rainwater harvesting estimates
depended on assumptions about catchment area, rainfall, and transport emissions, which are also site‑ specific. The
national grid baseline factor also reflects Jordan's current electricity mix, which will shift as renewable energy use
grows. Greenhouses were left out of this comparison because their main role is agronomic, not water conservation;
allocating their embodied footprint to water savings alone would not reflect their overall function on site.
5. Conclusions
There is a lack of field ‑based studies that compare smart irrigation with other non ‑conventional water conservation
options under the same farming conditions and within a shared water –carbon framework. This study assessed the
possible contribution of smart irrigation to sustainable water conservation compared to other options within a
commercial agricultural compound in Northern Jordan. The results indicate that smart irrigation can improve
irrigation efficiency even where water management is already reasonably maintained. In the present case, soil
moisture slightly exceeded FC several times during the monitoring period. However, converting these losses into
water volumes throughout the season and across all growing plots resulted in significant potential water savings. This
point is important because it shows that, in well‑ managed semi‑arid systems, the benefit of smart irrigation may lie
less in dramatic savings at the plot level and more in the steady reduction of small, repeated losses. While the volume
saved within a single plot was modest, the cumulative effect became far more meaningful when considered across the
larger number of operational plots within the agricultural compound. From a managemental perspective, this
observation gives smart irrigation practical significance as a tool for improving overall efficiency at farm scale rather
than only as a plot‑ scale intervention. Wastewater reuse and rainwater harvesting showed contrasting profiles.
Wastewater reuse showed greater water ‑saving potential, estimated by the fraction of water consumption not
directed towards irrigation. However, it also carried a potentially high embodied carbon footprint. Rainwater
harvesting had the opposite profile: the lowest footprint, but the smallest saving potential, given the site's arid
environment. This demonstrated how water demand management can offer a substantially higher reward compared
to technology integration and non ‑conventional water sources. Together, these results show that the most carbon ‑
effective conservation option is context‑ dependent. It is shaped as much by pumping energy and system design as by
the technology category itself. For farm managers and policymakers in Mafraq, and in similar semi‑arid parts of Jordan
and the wider region, this has a practical implication. Conservation strategies should be selected based on local energy
demand and realistic water ‑saving potential. They should not be chosen based on assumed environmental
performance alone. Smart irrigation offers a particularly accessible entry point in this regard. It required minimal
infrastructure investment compared to options like wastewater treatment. Despite this lower investment, it still
demonstrated a significant potential carbon footprint reduction in this study.

Conflict of interest statement
The authors declared no conflict of interest.
Funding statement
The authors declared that no funding was received in relation to this manuscript.
Data availability statement
The authors stated that experimental data used in this study will be made available upon reasonable request from the
corresponding author.

DYSONA – Applied Science 8 (2027) 1-16 Al-Kilani et al.
14

References

  1. Rezaei EE, Webber H, Asseng S, Boote K, Durand JL, Ewert F, Martre P, MacCarthy DS. Climate change impacts on
    crop yields. Nat. Rev. Earth Environ. 2023;4(12):831‑46. DOI
  2. Rosa L, Chiarelli DD, Sangiorgio M, Beltran‑Peña AA, Rulli MC, D’Odorico P, Fung I. Potential for sustainable
    irrigation expansion in a 3 °C warmer climate. Proc. Natl. Acad. Sci. 2020;117(47):29526‑34. DOI
  3. Al‑Hazmi HE, Mohammadi A, Hejna A, Majtacz J, Esmaeili A, Habibzadeh S, Saeb MR, Badawi M, Lima EC, Mąkinia J.
    Wastewater reuse in agriculture: Prospects and challenges. Environ. Res. 2023;236:116711. DOI
  4. Zegait R, Kouadri S, Chelgham M, Vucinic L, Ajia F, Sieni E, Al‑Kilani MR. AI‑based prediction of total dissolved
    solids in reverse osmosis desalination for arid regions. Desalin. Water Treat. 2025;323:101411. DOI
  5. Al‑Bakri J, Al‑Kilani MR. Drought Risk Management in the MENA Region: The Role of Earth Observations and
    Open‑Source Tools. In: NATO Science for Peace and Security Series C: Environmental Security. Springer Nature
    Switzerland. 2025. DOI
  6. Al‑Omoush RA, Al‑Bakri JT, Abdelal Q, Al‑Kilani MR, Hamdan I, Aljarrah A. Developing a Remote Sensing‑Based
    Approach for Agriculture Water Accounting in the Amman–Zarqa Basin. Water. 2025;17(14):2106. DOI
  7. Abdelal Q, Al‑Kilani Mr, Al‑Bakri J, Keblouti M, Sakna A. Reduced Frequency of Soil Moisture Measurements for
    Cost‑effective Irrigation Management in Arid Regions. J. Agrometeorol. 2026;28(1):99‑102. DOI
  8. Vallejo‑Gómez D, Osorio M, Hincapié CA. Smart Irrigation Systems in Agriculture: A Systematic Review. Agronomy.
    2023;13(2):342. DOI
  9. Bwambale E, Abagale FK, Anornu GK. Smart irrigation monitoring and control strategies for improving water use
    efficiency in precision agriculture: A review. Agric. Water Manag. 2022;260:107324. DOI
  10. Kim Y, Evans R, Iversen W. Remote Sensing and Control of an Irrigation System Using a Distributed Wireless
    Sensor Network. IEEE Trans. Instrum. Meas. 2008;57(7):1379‑87. DOI
  11. Adeyemi O, Grove I, Peets S, Domun Y, Norton T. Dynamic Neural Network Modelling of Soil Moisture Content for
    Predictive Irrigation Scheduling. Sensors. 2018;18(10):3408. DOI
  12. Kumar KA, Chaitanya K, Neelima TL, Madhukar B. Advances in Smart Irrigation Technologies for Enhancing Water
    Use Efficiency and Crop Productivity under Changing Climate Conditions. Agriculture Reviews 2024;3(1):01‑5.
    DOI
  13. Hamdouni A. Artificial Intelligence‑Driven Integrated Water Management and Agricultural Sustainability:
    Evidence from Saudi Arabia. Resources. 2026;15(3):38. DOI
  14. Martelli A, Rapinesi D, Verdi L, Donati IIM, Dalla Marta A, Altobelli F. Smart irrigation for management of
    processing tomato: a machine learning approach. Irrig. Sci. 2024;43(6):1407‑24. DOI
  15. Huang Z, Wang Y, Hui C, Xiao C. An Intelligent Water‑Saving Irrigation System Based on Multi‑Sensor Fusion and
    Visual Servoing Control. In: 2025 12th International Forum on Electrical Engineering and Automation (IFEEA).
    IEEE. 2025:1023‑7. DOI
  16. Touil S, Richa A, Fizir M, Argente García JE, Skarmeta Gómez AF. A review on smart irrigation management
    strategies and their effect on water savings and crop yield. Irrig. Drain. 2022;71(5):1396‑416. DOI
  17. Christou A, Beretsou VG, Iakovides IC, Karaolia P, Michael C, Benmarhnia T, Chefetz B, Donner E, Gawlik BM, Lee Y.
    Sustainable wastewater reuse for agriculture. Nat. Rev. Earth Environ. 2024;5(7):504‑21. DOI
  18. Frimpong F, Asante MD, Peprah CO, Amankwaa‑Yeboah P, Danquah EO, Ribeiro PF, Aidoo AK, Agyeman K, Asante
    MOO, Keteku A. Water‑smart farming: review of strategies, technologies, and practices for sustainable agricultural
    water management in a changing climate in West Africa. Front. Sustain. Food Syst. 2023;7:1110179. DOI
  19. Danboos A, Sharil S, Mohamad Hamzah F, Yafouz A, Huang YF, Ahmed AN, Ebraheem AA, Sherif M, El‑Shafie A.
    Water budget‑salt balance model for calculating net water saving considering different non‑conventional water
    resources in agricultural process. Heliyon. 2023;9(4):e15274. DOI

DYSONA – Applied Science 8 (2027) 1-16 Al-Kilani et al.
15

  1. Kalboussi N, Biard Y, Pradeleix L, Rapaport A, Sinfort C, Ait‑mouheb N. Life cycle assessment as decision support
    tool for water reuse in agriculture irrigation. Sci. Total Environ. 2022;836:155486. DOI
  2. Raimondi A, Quinn R, Abhijith GR, Becciu G, Ostfeld A. Rainwater Harvesting and Treatment: State of the Art and
    Perspectives. Water. 2023;15(8):1518. DOI
  3. Efthimiou N, Giotis T, Ragkos A. Applications for Non‑Conventional Water Resources in the Mediterranean Basin:
    A Literature Review. Sustainability. 2025;17(11):4964. DOI
  4. He X, Li Z, Xing C, Li Y, Liu M, Gao X, Ding Y, Lu L, Liu C, Li C. Carbon footprint of a conventional wastewater
    treatment plant: An analysis of water‑energy nexus from life cycle perspective for emission reduction. J. Clean.
    Prod. 2023;429:139562. DOI
  5. Al‑Khreisat A, Al‑Bakri J, Atiyat M, Al‑Kilani MR, Farhan I, Zucca C, Khudairat W. Soil Analytical Capabilities for
    Sustainable Land Management Across National Soil Services in the Mediterranean. Sustainability.
    2025;17(18):8228. DOI
  6. Al‑Bakri JT, Atiyat M, Al‑Kilani MR, Al‑Khreisat A, Farhan I, Zucca C. Towards harmonized soil intelligence: insights
    from a multi‑country survey of mediterranean soil information systems. Euro‑Mediterr. J. Environ. Integr.
    2026;11(2):97. DOI
  7. Abdelal Q, Al‑Kilani MR, Al‑Shishani G. Impact of Soil Particle Size on Soil Moisture Measurements through
    Dielectric and Electrical Resistance Properties. Eurasian Soil Sci. 2025;58(3):32. DOI
  8. Ministry of Environment. Updated submission of Jordan’s 1st nationally determined contribution (NDC).
    Hashemite Kingdom of Jordan. UNFCCC. 2021.
  9. Wyngaard SR, Kissinger M. Tomatoes from the desert: Environmental footprints and sustainability potential in a
    changing world. Front. Sustain. Food Syst. 2022;6:994920. DOI
  10. CIDB MC. Construction industry development board Malaysia. Kuala Lumpur. 2007.
  11. Gupta R, Lee S, Lui J, Sloan WT, You S. Carbon footprint assessment of water and wastewater treatment works in
    Scottish islands. J. Clean. Prod. 2024;450:141650. DOI
  12. Gallagher J, Gill LW. The Life Cycle Environmental Performance of On‑Site or Decentralised Wastewater
    Treatment Systems for Domestic Homes. Water. 2021;13(18):2542. DOI
  13. Masmoudi Jabri K, Nolde E, Ciroth A, Bousselmi L. Life cycle assessment of a decentralized greywater treatment
    alternative for non‑potable reuse application. Int. J. Environ. Sci. Technol. 2019;17(1):433‑44. DOI
  14. Rodríguez C, Sánchez R, Rebolledo N, Schneider N, Serrano J, Leiva E. Life cycle assessment of greywater
    treatment systems for water‑reuse management in rural areas. Sci. Total Environ. 2021;795:148687. DOI
  15. Jayabal R, Sivanraju R. Wastewater Treatment for Energy Conservation and Zero Carbon Footprint: A Review.
    Energy Sci. Eng. 2025;13(7):3806‑20. DOI
  16. Rashid ARM, Bhuiyan MA, Pramanik B, Jayasuriya N. Life cycle assessment of rainwater harvesting system
    components – To determine environmentally sustainable design. J. Clean. Prod. 2021;326:129286. DOI
  17. Abdelhamid MA, Abdelkader TK, Sayed HAA, Zhang Z, Zhao X, Atia MF. Design and evaluation of a solar powered
    smart irrigation system for sustainable urban agriculture. Sci. Rep. 2025;15(1):11761. DOI
  18. Ghimire SR, Johnston JM, Ingwersen WW, Sojka S. Life cycle assessment of a commercial rainwater harvesting
    system compared with a municipal water supply system. J. Clean. Prod. 2017;151:74‑86. DOI
  19. He Z, Wang R, Wang J, Chen H, Zhang S, Wang K, Lai J, Ren N, Ho S. Water conservation strategies reduce
    greenhouse gas emission from wastewater treatment plants: A domino effect. Environ. Sci. Ecotechnol.
    2025;26:100574. DOI
  20. Al‑Kilani MR, Abdelal Q, Bani‑Melhem K. Exploring the Circular Reuse Potential of Wastewater and Scrap Metal in
    Crop Irrigation at Urban–Rural Intersections Through Electrocoagulation. Water Environ. Res.
    2026;98(4):e70364. DOI

DYSONA – Applied Science 8 (2027) 1-16 Al-Kilani et al.
16

  1. Al‑Kilani MR, Abdelal Q, Al‑Shishani G. Are conductivity sensors useless for irrigators? Exploring measurement
    consistency around soil moisture thresholds relevant to different applications. Irrig. Drain. 2024;74(3):1018‑30.
    DOI

论文原文与出版信息

DYSONA – Applied Science
查看论文原文

继续探索