Monitoring Of Soil Water Content Research Articles

Monitoring of soil water content using spherical smart aggregates based on electromechanical impedance (EMI) technique

Real-time monitoring of soil water content is of great significance to prevent many engineering accidents, such as landslide, water seepage of foundation dam, and reduction of foundation bearing capacity. Electromechanical impedance (EMI) technique based smart aggregates (SAs) have showed excellent monitoring capability in various engineering applications. However, there are limited reports on the application of EMI based SAs in the monitoring of the soil water content. In this paper, the EMI based spherical SAs (SSAs) were investigated in the monitoring of soil water content. Firstly, the SSAs were designed and fabricated using the monolithic concrete encapsulation method, and their stable performance in air were evaluated, and confirmed by testing and analyzing the impedance spectrum. Then, the fabricated SSAs were immersed in the water environment for 28 d to ensure the stability under the working conditions, such as the soil with high water content and the hydration process of early-age concrete. Secondly, the monitoring experiments of soil water content were carried out based on the SSAs and the traditional SAs using the EMI technique. The measured impedance signatures under different water contents in soil were quantified by three types of statistical indexes, including root mean square deviation, mean absolute percentage deviation, and correlation coefficient deviation. The experimental results show that compared with the traditional SAs, the SSAs are more sensitive and stable to monitor the soil water content. Finally, the effect of temperature on the performance of SSA based on the EMI technique were conducted experimentally. The results demonstrate that temperature has influence on the monitoring results of the SSAs based on the EMI technique.

Monitoring of Thick Vadose Zone Water Dynamics Under Irrigation Using a 48 m Deep Caisson at the Luancheng Critical Zone Observatory

AbstractUnderstanding soil water dynamics in the deep vadose zone (DVZ, below the root zone) of irrigated farmland is critical for estimating groundwater recharge and assessing the impacts of agricultural activities on groundwater quality. However, the state and movement of soil water in the DVZ remain poorly understood owing to the lack of in‐situ observations. Here, based on a 48‐m deep caisson (Critical Zone Observatory, CZO) established in a typical irrigated farmland in the piedmont region of the North China Plain, we extended the monitoring of soil water content and matric potential across the entire vadose zone profile. We found that a high soil water content and matric potential (close to or slightly higher than the field capacity) was maintained in the DVZ of typical irrigated cropland. The response of soil water in the DVZ to the water inputs at the ground surface was predominantly sequentially delayed with an increase in depth in the DVZ. As a result, the wetting front propagated into the soil layer to a depth of 38 m within 228 days. Moreover, high soil moisture contents and matric potentials in the DVZ resulted in a dramatic but explicable difference between the estimated average pore water velocity (0.73 m year−1) and the wetting front propagation rate (0.16 m day−1). This study demonstrates the unreported rapid response of soil water in the DVZ to water input at the ground surface under intensively irrigated cropland and could promote a better understanding of groundwater recharge processes with a thick unsaturated zone.

A pipeline robot system for monitoring soil water content distribution

Monitoring soil water content (SWC) distribution is crucial for better understanding soil water dynamics and accurate hydrological modeling. In this study, a pipeline robot system for monitoring SWC distribution is designed to improve the current technical problems of insufficient measurement distance and low positioning accuracy of the existing SWC monitoring platform. The system consists of a pipeline robot, a fixed base station and a PVC pipe. The robot could fit into a 55-mm-diameter PVC pipe to measure SWC distribution horizontally, vertically, or at any angle to the horizontal direction. The base station serves as a data collector and also supplies power and distance information for recharging and positioning the robot, respectively. A series of tests for evaluating the performance of the pipeline robot system were conducted. Field experiments were also conducted to monitor soil water infiltration and horizontal distribution of SWC of crop root zone at three plots. The results showed that the minimum radius of volume of sensitivity of the dielectric sensor is about 2.25 cm. The robot for mobile measurement performed stable both in horizontal and vertical directions and the response of the dielectric sensor output to different dielectric materials is significant. The accumulated error of the encoder is effectively decreased from 4.3% to 1.2% by correcting the position error using the laser ranging sensor. The measured SWCs during the calibration experiment were highly correlated with those obtained by the drying method (R2 = 0.990 and RMSE = 0.0181 cm3 cm−3). The results of the field experiments demonstrated that the system is feasible for monitoring soil water infiltration in the vertical direction and SWC distributions of crop root zone horizontally at the three plots, indicating that the potential application of the developed system for long-term monitoring of SWC distribution under field conditions in the near future.

Model-based simulation-optimization of irrigation scheduling – A field evaluation with processing tomatoes

Due to climate change, increased regulation of water resources and competition from other beneficial uses, the agricultural sector is under pressure to use water more efficiently. This paper reports the field evaluation of two model-based simulation-optimization approaches for irrigation scheduling: deterministic optimization and stochastic optimization. The field experiments were conducted at the University of California Davis research farm with a processing tomato crop. The crop growth simulation model used in the study was DSSAT-CROPGRO processing tomato. In order to mitigate the impact of weather forecasts inaccuracies, irrigation schedules were updated every 5 to 10 days, depending on operational constraints. These updates were performed via custom graphical user interfaces that enabled the user to visualize the expected outcomes of various decisions or scenarios before choosing which irrigation schedule to implement. An irrigation treatment based on manual monitoring of soil water content with a field-calibrated neutron probe served as benchmark. The model-based treatments achieved yields and water use efficiencies that were not significantly different from those obtained in the neutron probe-based treatment. These results demonstrate the high potential of model-based simulation-optimization approaches for in-season adaptive irrigation scheduling, especially since neutron probes, which are considered one of the most accurate indirect methods of measuring soil water content, are not commonly used by growers.

Toward Large‐Scale Soil Moisture Monitoring Using Rail‐Based Cosmic Ray Neutron Sensing

AbstractCosmic ray neutron sensing (CRNS) has become a promising method for soil water content (SWC) monitoring. Stationary CRNS offers hectare‐scale average SWC measurements at fixed locations maintenance‐free and continuous in time, while car‐borne CRNS roving can reveal spatial SWC patterns at medium scales, but only on certain survey days. The novel concept of a permanent mobile CRNS system on rails promises to combine the advantages of both methods, while its technical implementation, data processing and interpretation raised a new level of complexity. This study introduced a fully automatic CRNS rail‐borne system as the first of its kind, installed within the locomotive of a cargo train. Data recorded from September 2021 to July 2022 along an ∼9 km railway segment were analyzed, as repeatedly used by the train, supported by local SWC measurements (soil samples and dielectric methods), car‐borne and stationary CRNS. The results revealed consistent spatial SWC patterns and temporary variation along the track at a daily resolution. The observed variability was mostly related to surface features, seasonal dynamics and different responses of the railway segments to wetting and drying periods, while some variations were related to measurement uncertainties. The achieved medium scale of SWC mapping could support large scale hydrological modeling and detection of environmental risks, such as droughts and wildfires. Hence, rail‐borne CRNS has the chance to become a central tool of continuous SWC monitoring for larger scales (≤10‐km), with the additional benefit of providing root‐zone soil moisture, potentially even in sub‐daily resolution.

Deep SOC stock dynamics under contrasting management systems: Is the EPIC model ready for carbon farming implementation?

Process-based models have been recognized as cost-effective tools to assess carbon farming mechanisms through quantifying the C fluxes in the agroecosystems. A result-based approach is suggested however the wide variability of agricultural environments makes further model implementation necessary to limit the uncertainty of the results, especially on deep soil organic carbon (SOC) stock estimation and stratification and in agroecosystems characterized by a shallow water table. In this study, a comprehensive soil and crop dataset collected over a seven-year period from different pedoclimatic conditions across the Veneto Region (NE Italy) was used for EPIC model calibration and validation of SOC stock dynamics. Experimental data included yields from several crops (corn, winter and spring wheat, rapeseed, and soybean), continuous monitoring of soil water content, and SOC stocks (1872 total samples within the 0–50 cm soil profile) under conventional, cover crop and conservation agriculture systems. Modelling was performed by testing two N-C sub-models (CENTURY and PHOENIX), which differentiated in terms of mineralization/immobilization rates.Results showed that the procedure was able to obtain parameters valuable for most of the management system and pedoclimatic condition, reproducing well the tested variables. The EPIC model acceptably captured soil water dynamics (Nash-Sutcliffe coefficient – NSE – was up to 0.26), especially in the topsoil. Furthermore, simulation of weed—crop competition in conservation agriculture strongly contributed to properly explain the variability in crop production among the contrasting agricultural systems (R2 ranged from 0.51 to 0.71). Likewise, EPIC skillfully simulated SOC stocks within the 0–50-cm profile regardless of the sub-model used (NSE was up to 0.64). Moreover, the model acceptably captured the profile SOC stratification among the different management practices. This study highlights the EPIC robustness for predicting SOC stocks and assessing with high accuracy carbon farming results.

Evaluation of the soil water content of two managed ecosystems using cosmic‐ray neutron sensing on China's Loess Plateau

AbstractCosmic‐ray neutron sensing (CRNS) is a promising method for the continuous monitoring of soil water content (θv) at the hectometre scale. However, few studies have validated its applicability within the complex terrains of various ecosystems in semiarid loess regions. In this study, CRNS‐based θv was measured on China's Loess Plateau under two managed ecosystems (Caragana korshinskii K. shrubland [KOS] and Medicago sativa L. grassland [SAG]). The accuracy of CRNS measurement was assessed with the traditional oven‐drying method, while its sensitivity to θv variation during the study period was compared with typical point‐scale monitoring sensors (EC‐5, Decagon Devices, Inc.). CRNS was found to be strongly correlated with the oven‐drying based θv under both ecosystems, with coefficients of determination (R2) of 0.99 and 0.93, and root mean square errors of 0.006 and 0.007 cm3/cm3, for KOS and SAG, respectively. The mean monitoring radius of the CRNS was approximately 204 and 231 m for KOS and SAG, with a mean measurement depth of approximately 39 and 45 cm, respectively. Moreover, the θv determined by the CRNS and EC‐5 sensors exhibited similar trends during the observation period, excluding that for April. However, the CRNS‐based θv had a higher spatial representation for both managed ecosystems compared with that for EC‐5. No significant difference was detected in the response range of θv to precipitation between the CRNS and EC‐5 sensors at a depth of 0–10 cm, whereas the response time of CRNS for KOS was much shorter than that of EC‐5 sensors (p < 0.01). Overall, the CRNS measured hectometre‐scale θv had acceptable accuracy and sensitively traced the precipitation response under different ecosystems with high θv heterogeneity in the semiarid loess regions.Highlights CRNS accurately estimates hectometre‐scale θv in ecosystems with complex terrains. The CRNS monitoring radius is ~218 m with ~42 cm measurement depth in the loess region. The θv in the managed shrubland is higher than that of the managed grassland. CRNS has a higher sensitivity in response to precipitation than EC‐5 sensors.

An experimental setup to investigate the effect of mycorrhizal fungi inoculation on plant water uptake in unsaturated soils

The use of vegetation as a Nature Based Solution, NBS, is increasingly being used for climate adaptation and resilience for engineered and natural slopes. As such it is important to understand the biological processes associated with vegetation and how they could be beneficially implemented. Due to the large repetition required when experimenting with biological components it is impossible to carry out experiments that are highly instrumented. This paper looks at mycorrhizal fungi as a method to improve plant water uptake and presents an experimental setup to allow for a low level of instrumentation in triplicate, with a few being well instrumented. Through benchmarking against TDR-probes, it was found that the method of monitoring soil water content changes in a mini-lysimeter with a loadcell provides a reliable and convenient approach to carry out experiments in three replicates at least.

High-Sensitivity and Wide-Range Antenna Sensor Based on EBG and SiO2 for Soil Water Content Monitoring

There are many problems with agricultural technology around the world, such as extensive irrigation which causes the waste of water. To significantly improve the efficiency of irrigation water, an antenna sensor, with high sensitivity and wide range, is proposed to measure the soil water content wirelessly. When the soil water content changes, the resonant frequency of the antenna will shift. By reading the resonant frequency of the antenna sensor, the soil water content can be measured and transmitted wirelessly. The antenna sensor has a wider detection range of soil water content, ranging from 10% to 44%, and a novel metamaterial structure (electromagnetic bandgap, EBG) is designed and its function in improving sensitivity is verified. Furthermore, by sputtering 60 nm thick SiO <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> film on the surface of the antenna patch, its sensitivity has also been greatly improved. When the SiO <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> film is sputtered on the antenna sensor, its sensitivity is 7.2 MHz/% and 7.6 MHz/% when the soil water content is 10%~29% and 30%~44%, respectively. In summary, this work contributes to methods and instruments for wireless measurement of the soil water content.

The Correlation between Proximal and Remote Sensing Methods for Monitoring Soil Water Content in Agricultural Applications

Water shortages have increasingly become a global issue due to the acceleration of climate change. The consumption of freshwater can be reduced to a minimum using water irrigation techniques that are based on conservative methods. For example, one of these is precision irrigation, or PI, which uses advanced digital technology to regulate the amount of water used. The aim is to use the least amount of water necessary for a given purpose. This approach keeps consumption to a minimum while the amount remains effective for its purpose. It is also important to note that the variability which occurs in soil and crops will create different types of conditions. These different conditions will need to be studied so as to determine the correct and adequate dynamics for a water management approach that is efficient. In this study, three investigation methods were developed and compared. The first evaluation was performed on outputs from the geoelectric reading of Automatic Resistivity Profiling (ARP). A second evaluation was performed in real time via a sensor network placed in the soil for the duration of two growing seasons of two different crops. The last evaluation was carried out by using maps of spectral indices obtained by the Sentinel 2 satellites. The correlations between the three methods were evaluated to verify if satellite information may have significant potential in the use of water management in varying conditions. From the results obtained, some correlations have been found from the observations of the three systems under study. This has given a positive input towards using satellite maps which are integrated with simplified proximal sensor networks. The outcome of this technique can improve the efficiency of how to manage water distribution on cultivated land.

Artificial intelligence-based fiber optic sensing for soil moisture measurement with different cover conditions

Actively heated fiber Bragg grating (AH-FBG) can perform quasi-distributed monitoring of soil water content. However, the analysis method needs to be improved to minimize measuring errors. In this study, the artificial intelligence method is proposed and a model test was used to verify its feasibility and to explore the influence of cover conditions. Three cover layers were considered, including bare soil (S), grass (G), and biochar mixed soil (B). The water content measurements based on maximum temperature increase have higher accuracy for G, followed by S, and the worst is B. Fluctuation in the heat power and the longitudinal heat transfer are the main sources of errors. Artificial neural network (ANN) models can effectively improve monitoring accuracy. Cover conditions have a significant influence on the measurements by affecting initial ground temperatures and water content gradients. For field monitoring, the cover layer should be considered when analyzing AH-FBG measurements.

A database of soil physical properties for the Kansas Mesonet

AbstractThis study developed a comprehensive database of soil physical properties to expand the potential applications of soil moisture observations across the Kansas Mesonet environmental monitoring network. The database comprises 14 site‐ and depth‐specific soil hydraulic properties for 40 stations monitoring soil water content. The measured soil hydraulic properties include sand, clay, and silt contents; bulk density; particle density; total porosity; effective saturation; saturated hydraulic conductivity; and water retention at six matric potentials. The soil database spans eight soil textural classes, with silty clay loam, silt loam, and silty clay soils dominating the fine‐textured soils, and sandy loam and sandy clay loam soils comprising the only coarse‐textured soils. All the measured soil hydraulic properties showed low coefficient of variation (CV ≤10%) within soil textural classes, except for the saturated hydraulic conductivity, which varied from a median of 0.520 ± 308 cm d–1 in the clay soils to 47.1 ± 119 cm d–1 in the sandy loam soils. The uncertainty of the Kansas Mesonet soil water content observations was reduced by at least 29% using laboratory sensor calibration equations instead of the factory default sensor equation. Our soil physical property database offers new prospects to use historical and near real‐time in situ soil water content observations across the Kansas Mesonet for diverse agricultural and hydrological applications, including drought monitoring, wildfire preparedness, and estimating potential drainage and groundwater recharge rates.

Long-term soil water content dynamics under different land uses in a small agricultural catchment

Abstract Longer term monitoring of soil water content at a catchment scale is a key to understanding its dynamics, which can assist stakeholders in decision making processes, such as land use change or irrigation programs. Soil water monitoring in agriculturally dominated catchments can help in developing soil water retention measurements, for assessment of land use change, or adaptation of specific land management systems to climate change. The present study was carried out in the Pannonian region (Upper-Balaton, Hungary) on Cambisols and Calcisols between 2015 and 2021. Soil water content (SWC) dynamics were investigated under different land use types (vineyard, grassland, and forest) at three depths (15, 40, and 70 cm). The meteorological data show a continuous decrease in cumulative precipitation over time during the study with an average of 26% decrease observed between 2016 and 2020, while average air temperatures were similar for all the studied years. Corresponding to the lower precipitation amounts, a clear decrease in the average SWC was observed at all the land use sites, with 13.4%, 37.7%, and 29.3% lower average SWC for the grassland, forest, and vineyard sites, respectively, from 2016 to 2020 (measured at the 15 cm depth of the soil). Significant differences in SWC were observed between the annual and seasonal numbers within a given land use (p &lt; 0.05). The lowest average SWC was observed at the grassland (11.7%) and the highest at the vineyard (28.3%). The data showed an increasing average soil temperature, with an average 6.3% higher value in 2020 compared to 2016. The grassland showed the highest (11.3 °C) and the forest soil the lowest (9.7 °C) average soil temperatures during the monitoring period. The grassland had the highest number of days with the SWC below the wilting point, while the forest had the highest number of days with the SWC optimal for the plants.

Deep soil water deficit causes Populus simonii Carr degradation in the three north shelterbelt region of China

Large-scale afforestation exacerbates soil water scarcity in a semiarid region of northern China, causing the formation of a dried soil layer (DSL) and ecosystem degradation, and overmature Populus simonii Carr (>30 years) degradation in the Three North Shelterbelt Program (TNSP) area has recently been observed. Determining the relationship between degraded Populus simonii Carr and soil water deficit is essential for afforestation in semiarid regions. We monitored soil water content (SWC) to a 160 cm depth in an area with five degradation degrees (healthy, light degradation, moderate degradation, severe degradation, and dead) to investigate soil water dynamics and DSL characteristics from May 2018 to December 2021. The results showed that: compared with that of healthy Populus simonii Carr (94.7 ± 26.2 mm), the shallow layer (0–80 cm) soil water storage of severely degraded (114.1 ± 29.9 mm) and dead Populus simonii Carr (128.0 ± 35.7 mm) was restored by 19.4 mm and 33.3 mm, respectively. However, the deep-layer (80–160 cm) SWC was difficult to recover and a DSL gradually developed, the DSL − SWC was only 9.3–11.95 %. The deep-layer relative extractable soil water (REW) of healthy Populus simonii Carr was significantly higher than that of moderately degraded, severely degraded, and dead Populus simonii Carr (P < 0.05). The deep-layer SWC decreased with increasing degradation due to coarser soils and increased BD. A greater bulk density (BD) and higher sand proportion reduced the soil water holding capacity. Poor site conditions reduced the resistance of shelterbelts, causing the degradation and mortality of Populus simonii Carr during drought. Therefore, our results improve the understanding of the soil water dynamics and degradation mechanism of Populus simonii Carr, providing a theoretical basis for managing and maintaining the TNSP.

Optimal irrigation management for avocado (cv. 'Hass') trees by monitoring soil water content and plant water status

Irrigation scheduling based on soil water content (Ɵw) sensors requires that Ɵw be maintained within a range (management lines) that is optimal for plant growth. The lower limit or “breaking point” is determined following the soil water content dynamics on the transition of a rapid rate of depletion to a slower, under similar reference evapotranspiration. Although this criterion is practical, its implementation should be validated with plant water status measurement that contemplate weather condition, such as stem water potential “non-stressed” baseline (Ψₓ as a function of vapor-pressure deficit (VPD) in Ɵw conditions that do not limit yield). A study was conducted on a mature cv. ‘Hass’ avocado orchard in Central Chile during two seasons. There were 5 irrigation treatments: T1, Control; T2 and T3 with 29% less and 25% more of what was applied in T1, respectively; T4 and T5 same as Control until first and second fruit drop abscission, respectively, and then with 29% less. T1 trees were irrigated using a continuous frequency domain reflectometry (FDR) probe to maintain the root zone between field capacity and the breaking point. There was biweekly monitoring of the Ɵw prior to irrigation, Ψₓ and VPD. The Ψₓ decline proportional to the intensity and the timing of water restriction; however, no treatment affected the crop load in either season. T2 did not show significant detrimental in fruit size, production and maturation, despite that frequently reached water content levels at the limit of the breaking point, and showed lower levels of stem water potential than Control, being the treatment with the highest water productivity. The results confirm that breaking point is an effective criterion to establish irrigation management. Additionally, when comparing the baseline for our non-stressed trees with a baseline from full irrigation treatments obtained from the literature, 30% water savings were achieved.

Estimation of variability in soil water content in a forested critical-zone experimental catchment in Eastern China

Knowledge of soil water content (SWC) dynamics within soil profiles is crucial for the effective management of water and soil resources. This study aims to clarify the temporal variability and stability of SWC in a forested critical-zone experimental catchment, and further to improve the understanding of the temporal and spatial distribution of soil water in a typical hilly catchment in eastern China. The selected Nandadish (NDD) catchment covering 0.79 ha was instrumented with 34 SWC monitoring sites using Frequency Domain Reflectometry. The consecutive high-resolution monitoring data of soil water at different depths of the sites were collected from January 2017 to December 2019. The results showed that the SWC of the shallow layer (0–30 cm) had the strongest variability over time during the three hydrologic years. The interannual variability of SWC showed the opposite regularity with that of the seasonal variability. Specifically, the spatial variability of SWC in the dry years was greater than that in wet years; whilst the temporal stability of SWC in dry seasons was greater than that in rainy seasons. Precipitation and temperature were the two dominant factors influencing the temporal variation of SWC. Precipitation controlled the interannual variation of SWC, while temperature controlled the seasonal variation of SWC. Additionally, soil water had high temporal stability throughout the observation period in NDD catchment, and the most representative point was located at a relatively flat and central place, which can be used to simulate the variability of SWC under different rainfall conditions in the study area. The temporal stability of SWC patterns was controlled by topography, geographic location, throughfall, and the groundwater level in the study area, which was characterized by sloping terrain and forested cover. This research provides scientific bases for the optimum design of ground sampling, and the temporal and spatial prediction for soil moisture in a typical eastern hilly area with forest land uses.

Development of a Current Injection-Type Impedance Measurement System for Monitoring Soil Water Content and Ion Concentration.

This study was conducted with the aim of developing a circuit system that enables the measurement of the moisture content and ion concentration with a simple circuit configuration. Our previous studies have shown that soil can be represented by an equivalent circuit of a parallel circuit of resistors and capacitors. We designed a circuit that can convert the voltage transient characteristics of the soil when a current is applied to it into a square wave and output frequency information and developed an algorithm to analyze the two types of square waves and calculate R and C. Normal operation was confirmed in the range of 10 kΩ–1 MΩ for the designed circuit, and the calculation algorithm matched within a maximum error of 5%, thus confirming the validity of the program. These successfully confirmed the changes in the water content and ionic concentration. The soil moisture content measurement succeeded in measuring a maximum error of about 10%, except at one point, and the soil ion concentration measurement succeeded in measuring a maximum error of 6.6%. A new, simple, noise-resistant moisture content and ion concentration measurement circuit system with square wave output has been realized.

Application of image technology to simulate optimal frequency of automatic collection of volumetric soil water content data

Volumetric soil water content (VSWC) monitoring is an important aspect of environmental monitoring of farmland. Accurate and real-time determination of the VSWC is important for crop drought stress diagnosis and smart irrigation. Image technology is commonly used in agricultural information technology. Based on image technology, we simulated the optimal frequency of the sensors to automatically collect VSWC data, thereby solving the problems of data redundancy and data analysis difficulties in real-time monitoring. In this study, a cotton field under mulched drip irrigation in 2018 was utilised as the research subject, 5TE sensors were arranged on the soil profile of the cotton field using the "grid method", and Voxler and Surfer software was used to model the VSWC data and draw contour maps. Image processing technology (image greyscale and image similarity comparison) was employed to determine the image algorithm suitable for contour map pre-processing and the best time period for VSWC monitoring. These results indicated that the contrast-limited adaptive histogram equalisation (CLAHE) greyscale algorithm is a suitable pre-processing algorithm for processing contour maps using image processing technology, and the best 5TE sensor data monitoring time period is every 10 h. This conclusion provides a theoretical reference for VSWC monitoring and water management in production.

Monitoring soil water content and its salinity with high-precision and low-cost in-situ sensors

The article presents an in situ sensor to monitor soil moisture and water salinity by measuring soil permittivity and conductivity at a frequency 24 MHz via an admittance approach. It addresses difficulties encountered in other permittivity-based sensors. A self-balanced bridge provides on one measurement a digital output linear to capacitance and conductance of electrodes in soil with low offset, high phase resolution, and taking into account all parasitic impedances. Circuit design and procedures to achieve it are described, using electric circuit theory and component specifications. Probe geometry, two parallel cylinders, permits insertion in soil from surface, low disturbances and a high sample volume around electrodes. It provides an analytical relation to convert admittance to soil permittivity. Calibration of a sensor with air, alcohols and water fixes its sensitivity and residual phase and inductance for an uncertainty lower than 2% over a range up to 80 and 200 mS m–1. Identical sensors can use same parameters after adjusting in one simple operation bridge phases. Thermal drift and its consequences on soil variable measurement are investigated and physical origins identified. It permits to reduce the phase drift within ±150 ppm °C– 1 rad to keep accuracy.

Towards near real-time national-scale soil water content monitoring using data fusion as a downscaling alternative

Soil water is a critical component of the water balance to make water management decisions at multiple scales. While soil water can be sensed remotely, this is generally at coarse scales (> 12.5 km). In addition, soil moisture products developed at field scale resolutions (< 250 m) have been mostly limited to shallow observations (up to 10 cm depth) and are impacted by land use. The objective of this study was to create an accurate downscaled soil water content product at multiple depths and at a fine 90 m resolution by fusing modelled and remote sensing datasets via deep learning. Reference data was based on the OzFlux and Oznet networks. Covariates included the North America Space Agency (NASA) United States Department of Agriculture (USDA) Soil Moisture Active Passive (SMAP) remote sensing data assimilation model, Sentinel 1 from Copernicus, surface reflectance, land surface temperature and land cover from the Moderate Resolution Imaging Spectroradiometer (MODIS), and gridded soil properties. Two model approaches were used including a multilayer perceptron for the surface (0–10 cm), and recurrent neural networks for the surface/subsurface (0–30 cm and 30–60 cm) soil water content. The surface prediction performance resulted in a root mean square error (RMSE), mean absolute error (MAE) and Pearson’s correlation of 0.073, 0.057, and 0.74, degrading in depth to 0.07, 0.062, and 0.5. Overall, these 90 m resolution predictions improve on the performance of NASA-USDA SMAP (10 km resolution) and the Australian Landscape Water Balance (5 km resolution) simulated soil water contents. Land use/land cover (LULC) are important explaining factors for performance and SHapley Additive exPlanations (SHAP) indicate high importance of NASA-USDA SMAP for surface predictions, while soil properties and LULC increase in importance with depth.