The adoption of sensors and other technologies in modern agricultural systems has led to an increase in data availability for the optimization of crop management. Soil moisture (SM) forecasts, which ...heavily depend on the availability of SM sensor data, have the potential to reveal underlying processes that are not easily identified through day-to-day field management. The SM data’s complexity and the temporal and spatial dynamic interrelation with other environmental factors present challenges that make it necessary to implement preprocessing techniques and models capable of adapting to the changing characteristics of the data. Recent studies have suggested that combining data preprocessing techniques, such as input time series (TS) decomposition, with ensemble forecasting models can reduce data complexity, improve forecast accuracy, and provide the model with the necessary robustness against extreme input values. In this study, we assess the efficiency of Singular Spectrum Analysis (SSA) and Ensemble Empirical Mode Decomposition (EEMD) as input TS decomposition techniques to improve the preprocessing and forecasting performance of machine learning (ML) and deep learning (DL) algorithms. The results of this study suggest that the use of EEMD could have an average reduction of 44% in absolute errors while improving the interpretability of the data. In a case study on citrus production in Southeast Florida, we demonstrated that using EEMD as a preprocessing method significantly reduces prediction errors. This reduction could allow a substantial increase in water savings when users implement SM forecasts for irrigation scheduling.
•We evaluated SSA and EEMD for ML/DL soil-moisture (SM) forecasting.•EEMD incorporation led to a 44% error reduction compared to no preprocessing.•EEMD effectively isolated nonstructural components in the time series.•EEMD outperformed SSA due to high-frequency variation in SM-TS.•Adding EEMD could lead to significant water savings in citrus irrigation.
The development of cost-effective, digitally based decision support systems is a key challenge in the optimization of farm management. Yet, the majority of sensor-based decision tools which support ...fertiliser management have relied on simplistic mechanistic frameworks normally informed by a single sensor. This study used a 20-year nitrogen (N) experiment on winter wheat (Triticum aestivum L.) to test a range of approaches for N decision support systems, including commercial sensor-based options and a novel, multivariate, data-driven approach. The latter was based on a non-mechanistic framework in which various digital variables were trained directly against optimum N application rates using machine learning. It was hypothesized that such a method would enhance our ability to handle system complexity, resulting in higher accuracy for the decision, as compared to current farm management or to available sensor-based options, both of which are normally underpinned by mechanistic methods. Results showed that the proposed approach was able to predict the optimal N rate with an RMSE of 16.5 kg N ha–1 (R2 = 0.79). This method was also the only one that was statistically superior (p < 0.05) to the control scenario (the application of the historical average optimal N rate; RMSE =38.0 kg N ha–1). This proposed approach used a multivariate digital input including a spectral vegetation index (normalized difference vegetation index, NDVI), weather and soil moisture data and information from on-farm experimentation (the in-situ N response using a ‘N-rich’ strip) to guide the decision. When similar data input and modelling techniques were used to predict yield potential to then derive an N recommendation through a mechanistic decision framework – a nutrient mass balance – the recommendation error (RMSE) increased to 26.0 kg N ha−1 (R2 = 0.51). In summary, by forcing the input data through the mechanistic framework, the decision error increased. This study challenges the ideas that farm decisions should follow pre-established agronomic mechanistic frameworks and that digital technologies must necessarily be used to estimate specific crop and soil attributes so as to enable deployment of current decision systems at scale and site-specifically.
Florida is the largest producer of fresh-market tomatoes in the United States. Production areas are typically intensively managed with high inputs of fertilizer and irrigation. The objectives of this ...3-year field study were to evaluate the interaction between N-fertilizer rates and irrigation scheduling on yield, irrigation water use efficiency (iWUE) and root distribution of tomato cultivated in a plastic mulched/drip irrigated production systems. Experimental treatments included three irrigation scheduling regimes and three N-rates (176, 220 and 230
kg
ha
−1). Irrigation treatments included were: (1) SUR (surface drip irrigation) both irrigation and fertigation line placed right underneath the plastic mulch; (2) SDI (subsurface drip irrigation) where the irrigation line was placed 0.15
m below the fertigation line which was located on top of the bed; and (3) TIME (conventional control) with irrigation and fertigation lines placed as in SUR and irrigation being applied once a day. Except for the “TIME” treatment all irrigation treatments were controlled by soil moisture sensor (SMS)-based irrigation set at 10% volumetric water content which was allotted five irrigation windows daily and bypassed events
if the soil water content exceeded the established threshold. Average marketable fruit yields were 28, 56 and 79
Mg
ha
−1 for years 1–3, respectively. The SUR treatment required 15–51% less irrigation water when compared to TIME treatments, while the reductions in irrigation water use for SDI were 7–29%. Tomato yield was 11–80% higher for the SUR and SDI treatments than TIME where as N-rate did not affect yield. Root concentration was greatest in the vicinity of the irrigation and fertigation drip lines for all irrigation treatments. At the beginning of reproductive phase about 70–75% of the total root length density (RLD) was concentrated in the 0–15
cm soil layer while 15–20% of the roots were found in the 15–30
cm layer. Corresponding RLD distribution values during the reproductive phase were 68% and 22%, respectively. Root distribution in the soil profile thus appears to be mainly driven by development stage, soil moisture and nutrient availability. It is concluded that use of SDI and SMS-based systems consistently increased tomato yields while greatly improving irrigation water use efficiency and thereby reduced both irrigation water use and potential N leaching.
Site-specific irrigation decisions require information about variations in soil moisture within the active rooting depth of the crop. Producers have been using soil moisture sensors to make ...irrigation decisions, and soil moisture sensors have been shown to help reduce water usage without reducing yields. There are still unanswered questions on improving efficiency with soil moisture sensors based on density and location of sensors within a field. This three-year study used sensors to evaluate the spatio-temporal variability of soil moisture across an 18-ha production field in a corn/soybean rotation. A 55 m by 55 m grid was laid on the field, resulting in 44 sampling points that fell either underneath the center-pivot irrigation or the end gun. At each point location, two Watermark granular matrix sensors were installed at depths of 31 and 61 cm for 2018 and 2020 and an additional 76 – cm sensor in 2019. Analysis of soil samples collected in year one revealed fairly homogeneous soils across the field with silty clay loam as the major soil type and only eight percent silt loam. Plant height and leaf area index (LAI) were measured weekly at each of the 44 sampling points. Inverse distance weighted (IDW) interpolation methods were used to predict soil water tension (SWT) values for locations between known points and aid in sensor density and placement within the field. Linear regression was used to model the relationship of LAI and plant height with soil matric potential to find surrogate methods for predicting SWT. The IDW results show that when uniform irrigation applications are made to the field, fewer sensors that are placed in better locations throughout the field can be as useful as a densely gridded array of sensors. Results showed that, while not strong, plant height had a better relationship to SWT than LAI.
•Plant height and LAI cannot be used to predict temporal variability of SWT.•Soil water tension (SWT) variability is highest when the soil is driest.•Placement of sensors within a field is more important than the density of sensors.•Spatial variability of SWT can exist even with homogeneous soils.
•An advanced, flexible and energy efficient all-optical self-heated platform for soil VWC monitoring is here proposed.•The sensor proved to be able to provide highly stable, energy efficient, fast ...(time response ≤ 10 s) and repeatable (repeatability error of 1.5%VWC) measurements, with resolution values lower than 0.25%VWC, with performances that are comparable with the ones commercial devices.•The possibility to tune the injected power value according to the application field makes the proposed platform extremely flexible and versatile for its exploitation in large areas and/or long-distance monitoring applications with relevant benefits in terms of energy consumptions, costs and system complexity.
The determination of the amount of moisture in soil is crucial in several application fields such as agricultural, geotechnical, hydrological, and environmental engineering. In this paper, we propose the use of an advanced all-optical self-heated sensing platform for highly sensitive and repeatable monitoring of the soil volumetric water content (VWC). The proposed platform is realized by integrating, inside a standard metallic needle, a fiber optic heating device integrated with a fiber Bragg grating sensor (FBG) acting as thermal monitor. The all optical active heating device consists of two main components: i) a core offset fusion splice between two single mode optical fibers to spread light from the core to the cladding and ii) a metal-coated region of the fiber where the FBG is inscribed. A full characterization of the proposed device has been performed for different VWC values from 9 to 36 %VWC in the soil temperature range between 5 and 40 °C. Collected results demonstrate that the proposed platform is able to perform highly stable, fast, robust and energy efficient measurements of the soil water content, with a repeatability error of 1.5 %VWC, time response ≤ 10 s and resolution ≤ 0.25 %VWC. The possibility of tuning the value of the injected power according to the application makes the proposed sensor extremely flexible and versatile for its exploitation in large areas and/or long-distance monitoring applications with relevant advantages in terms of energy savings, costs and system complexity.
The reduction of soil suction and consequent loss of shear strength due to infiltration is known to trigger shallow landslides during periods of concentrated rainfall. In the mountainous terrain of ...Nepal, the risk of shallow rainfall-induced landsliding is further exacerbated by non-engineered hillslope excavation for local roads construction. To better understand the combined effect of rainfall and road cutting on landsliding, a detailed investigation of a shallow landslide was conducted, on a site with a steep road cut that failed due to rainfall infiltration in July 2018. An integrated investigation approach was adopted, combining field and laboratory testing and field monitoring with a series of coupled hydro-mechanical analyses with the finite element code PLAXIS 2D. The field and laboratory tests were performed to characterise the subsoil condition and determine the soil parameters for the hydro-mechanical analyses. The field monitoring program was set up to obtain real-time measurements of rainfall and volumetric water content of the soil. The monitored data was used to calibrate the numerical model and assess the reliability of its predictions. Results of the numerical back-analysis suggest that the investigated landslide was triggered by rainfall infiltration causing a gradual reduction of soil suction at shallow depths of ≤1.7 m and the presence of the steep road cut promoted slope failure by allowing larger displacements to occur in the hillslope. Without the road cut, the slope was found to have ~35% greater initial factor of safety and under the landslide-triggering rainfall, the undisturbed slope was found to remain stable with ~170% greater factor of safety than that in the slope with the road cut. This indicates that the presence of a road cut increases the likelihood of landslide during rainfall and lowers the minimum level of rainfall needed for landslide initiation. Hence, rainfall-induced roadside slope failures could become more frequent and extensive if roads continue to be built by informal slope excavation, without adopting suitable interventions, some examples of which are presented in this study.
•Combined effect of rainfall and road cut on landsliding was investigated.•Integrated methodology with field and laboratory testing and numerical modelling was adopted.•Reliability of numerical modelling was improved by calibration against field measurements.•Two ways by which road cuts promote rainfall-induced landslides were identified.
In drylands, salt accumulation due to excessive irrigation and poor drainage negatively affects agricultural production. Water saving and drainage system improvement can effectively prevent salt ...accumulation. However, drip irrigation and subsurface drainage require initial and maintenance costs, making them difficult for farmers in developing countries.As a low-cost and simple technique to improve drainage function, a new subsoil breaker called cut-soiler, which fills channels at 40-60 cm depth in soil with crop residues and functions as a filter, was developed in Japan. In a previous study, cut-soiler contributed to the improvement of fields with poor drainage in Japan. However, its salinity control benefits for drylands are unknown. Therefore, this study aimed to experimentally demonstrate cut-soiler on salt-affected soil in India.A lysimeter experiment (2 m squares) was conducted to evaluate the saline soil remediation with and without cut-soiler plots during the dry and rainy seasons from October 2018 to August 2019. The calculated saturated EC (EC(c)) was continuously monitored by dielectric soil moisture sensors (5TE) to investigate its dynamics. The results showed that the EC(c) peak with cut-soiler at 12 cm depth, after irrigation during the dry season, was 18.7% lower than that without cut-soiler. In the rainy season, the EC(c) with cut-soiler at 50 cm depth decreased in response to rainfall and was 38.2% lower than that without cut-soiler. These results indicate that salts were dissolved by irrigation or rainfall and the infiltration water containing dissolved salts flowed through the outlet pipe of the cut-soiler.
Real time soil moisture monitoring using sensors has potential to save irrigation water and improve water productivity. Field experiments were carried out for two successive years (2016-17 and ...2017-18) to produce wheat crop at the Water Management Research Center, Postgraduate Agricultural Research Station, University of Agriculture, Faisalabad. Field irrigation methods included flood irrigation (canvas pipe), perforated pipe irrigation, and drip irrigation under different planting geometries and irrigation designs. The sensor-based irrigation systems were developed using locally available material to minimize the cost of equipment development and energy consumption for crop irrigation. Seven wheat crop treatments used in this experiment were T1-flood irrigation flat sowing by rabi-drill, T2-flood irrigation bed furrow planting with 0.254 m furrow, T3-perforated pipe irrigation bed furrow planting with 0.254 m furrow, T4-perforated pipe irrigation bed furrow planting with 0.203 m furrow, T5- perforated pipe irrigation bed furrow planting with 0.152 m furrow, T6-drip irrigation flat with 0.914 m lateral spacing and T7- drip irrigation on beds with 0.914 m lateral spacing. An IT-based web server was developed for monitoring soil moisture status to serve as decision support system for applying irrigation to the crops. The developed sensors sent soil moisture signals on cloud for data storage, reuse and sharing purpose using coding. The irrigation was applied based on soil moisture status. The system based on micro-controller was tested for irrigating wheat crop. Raspberry Pi-3 (Model B) controlled hardware in distribution box (DB) made excellent use of indigenized soil moisture sensors for calibration and irrigation water management. Type-I (Single probe) and Type-II (Double probe) steel sensors performed best due to high R2 values of about 0.99 and RMSE in the range of 3.30% - 3.50% during calibration. The calibration further improved the accuracy of both steel and copper sensors. Since the sensors were designed, developed, and calibrated during the 1st year (2016-17) and properly installed in 2nd year (2017-18), therefore, have affected crop and soil parameters positively. Drip irrigation treatments (T6 = 359.56 mm and T7 = 358.65 mm) required significantly lowest mean amount of water than those by all the other treatments and the flood irrigation treatments (T1 = 431.55 mm and T2 = 424.95 mm) required significantly greatest (α = 0.05) amount of mean irrigation depth. Drip irrigation treatments (T6 and T7) produced high mean water productivity values (14.30 and 14.20) than those under flood irrigation treatments (T1 = 9.6 and T2 = 10.30) and perforated pipe irrigation treatments (T3 = 12.66, T4 = 12.43 and T5 = 12.30). The mean yield of wheat grain over two years was greater under drip irrigation treatments (T6 = 5145.1 kg/ha and T7 = 5091 kg/ha) than those under flood (T1 = 4139 kg/ha, T2 = 4371 kg/ha) and perforated pipe irrigation treatments (T3 = 4969 kg/ha, T4 = 4872 kg/ha, T5 = 4775.7 kg/ha). Perforated pipe irrigation treatments had significantly greater (α = 0.5) wheat grain yield than those under flood irrigation treatments.
•A ten years real farm case study is reported and studied.•Digital agriculture solutions improved corn yield on average by 31% on a ten years basis.•Applied nitrogen reduced by around 23% while the ...total yield was improving.•Precision and digital agriculture approaches evolve with changing conditions and scenarios.
Farmer’s management decisions and environmental factors are the main drivers for field spatial and temporal yield variability. In this study, a 22 ha field cultivated with corn for more than ten years using different prescription maps of nitrogen application rates was investigated. Prescription maps were developed based on archived yield maps, soil analysis and recently integrated with Sentinel 2 satellite images. In addition, farmer experience and availability of variable rate application (VRA) requirements had an influence on the development of the homogeneous management zones. The initial approach with VRA was quite simple, based on a simple partitioning of the field into three rectangular zones (defined mainly based on previous yield maps and farmer experience). The partitioning changed with time and knowledge, evolving to the final five irregularly shaped zones (defined based on Farm works decision support software). Furthermore, since 2010 the farmer began using soil moisture sensor for irrigation decisions. Results of the present study highlight an improvement in corn yield and a reduction in total applied nitrogen. Corn yield improved on average by 31% on a ten years basis to reach more than 14 ton/ha dm. in 2018. At the beginning of VRA, yield maps showed a high spatial variation between field zones compared to reduced variation in the following seasons. In addition, the nitrogen applied reduced by around 23% while the total yield was improving. These results showed an increase in the partial factor productivity from less than 54 to around 87 kg of corn grain per kg of nitrogen applied. This promising result shows that farmer management decisions can improve every season by continuous monitoring of crop performance, understanding field variability and taking advantage of recently developed decision support software tools.
•Sensor positioning and accuracy effect on irrigation efficiency is investigated.•A mathematical model incorporating a system-dependent boundary condition is used.•Soil moisture sensors positioning ...and accuracy considerably affect irrigation efficiency.•Models are efficient tools for irrigation scheduling characteristics investigation.
Recent advances in electromagnetic sensor technologies have made automated irrigation scheduling a reality using state-of-the-art soil moisture sensing devices. However, many of the available guidelines for sensor placement were empirically determined from site and crop specific experiments. Sensors accuracy could be also an important factor affecting irrigation efficiency. This study investigates how soil moisture sensors positioning and accuracy may affect the performance of soil moisture based surface drip irrigation scheduling systems under various conditions. For this purpose several numerical experiments were carried out using a mathematical model, incorporating a system-dependent boundary condition in order to simulate soil moisture based irrigation scheduling systems. The results of this study provided clear evidence that soil moisture sensors positioning and accuracy may considerably affect irrigation efficiency in soil moisture based drip irrigation scheduling systems. In specific cases the effect of soil moisture sensors positioning was as high as 16%; however, when nearby sensor positions were examined, the observed differences were generally low. The effect of sensors accuracy was even clearer. For the lower sensor's error level studied (±0.01cm3cm−3) the effect on irrigation efficiency ranged between 2.5% and 6.4%, while for the higher error level (±0.03cm3cm−3) the effect ranged between 10.2% and 18.7%. These results highlight the importance of a detailed study taking into account the characteristics of specific crops, irrigation, and scheduling systems as well as soil moisture sensors in order to provide a sound basis for improved irrigation scheduling. The need for soil specific calibration of the sensors used in such systems is highlighted as well. Lastly, a significant outcome of this study is the ability of computer models to serve as efficient tools for the detailed investigation of sensors positioning and accuracy, or other automated scheduling system characteristics.