Future Soil Moisture Research Articles

Optimizing Irrigation Efficiency using Deep Reinforcement Learning in the Field

Agricultural irrigation is a significant contributor to freshwater consumption. However, the current irrigation systems used in the field are not efficient. They rely mainly on soil moisture sensors and the experience of growers but do not account for future soil moisture loss. Predicting soil moisture loss is challenging because it is influenced by numerous factors, including soil texture, weather conditions, and plant characteristics. This article proposes a solution to improve irrigation efficiency, which is called DRLIC (deep reinforcement learning for irrigation control). DRLIC is a sophisticated irrigation system that uses deep reinforcement learning (DRL) to optimize its performance. The system employs a neural network, known as the DRL control agent, which learns an optimal control policy that considers both the current soil moisture measurement and the future soil moisture loss. We introduce an irrigation reward function that enables our control agent to learn from previous experiences. However, there may be instances in which the output of our DRL control agent is unsafe, such as irrigating too much or too little. To avoid damaging the health of the plants, we implement a safety mechanism that employs a soil moisture predictor to estimate the performance of each action. If the predicted outcome is deemed unsafe, we perform a relatively conservative action instead. To demonstrate the real-world application of our approach, we develop an irrigation system that comprises sprinklers, sensing and control nodes, and a wireless network. We evaluate the performance of DRLIC by deploying it in a testbed consisting of six almond trees. During a 15-day in-field experiment, we compare the water consumption of DRLIC with a widely used irrigation scheme. Our results indicate that DRLIC outperforms the traditional irrigation method by achieving water savings of up to 9.52%.

Modeling of Soil Moisture Data by ARMA Time Series

The use of known data to predict future environmental parameters plays a crucial role in agriculture. In this paper, we propose a novel time series prediction method that combines the Auto-Regressive Moving Average Model (ARMA) and Gradient Boost Regression Tree(GBR) to forecast future soil moisture values. Firstly, the optimal number of decomposition modes for AMRA is determined by using Auto-correlation Function (ACF) and Partial Auto-correlation Function (PACF) plots. Secondly, according to statistics of XIlin Gol grassland offered by the Huawei Cup Mathematical Modeling Contest in 2022, the data including soil evaporation, precipitation, and soil moisture in the past ten years, are used as input parameters of ARMA to predict the precipitation and soil evaporation from 2022 to 2023. Then, the superiority of GBR was verified by comparing algorithms such as Support Vector Regression (SVR) and Random Forest(RF). Finally, GBR was used to realize the prediction for different soil moisture values from 2022 to 2023.

A new long short-term memory based approach for soil moisture prediction

Water scarcity is becoming more severe around the world as a result of suboptimal irrigation practices. Effective irrigation scheduling necessitates an estimation of future soil moisture content. This study presents deep learning models such as CNN-LSTM, a hybrid Deep Learning model that predicts future soil moisture using climate and soil information, including past soil moisture content. The study also investigates the appropriate number of observations and data sampling rate required to predict the next day’s soil moisture value. In terms of MSE, MAE, RMSE, and R 2 , the hybrid CNN-LSTM model is compared to standalone LSTM and Bi-LSTM models. The LSTM model achieved an MSE of 0.2471, MAE of 0.1978, RMSE of 0.4971, and R 2 of 0.9714. The LSTM model outperformed the Bi-LSTM model, which had an MSE of 0.3036, MAE of 0.3248, RMSE of 0.5510, and R 2 of 0.9614. With an MSE of 0.1348, MAE of 0.1868, RMSE of 0.3672, and R 2 of 0.9838, the hybrid CNN-LSTM model outperformed the LSTM. Our findings suggest that deep learning models, particularly the Convolutional LSTM, hold great potential for predicting soil moisture accurately. The Convolutional LSTM model’s superior performance can be attributed to its ability to capture spatial dependencies in soil moisture data. Furthermore, the results show that for better prediction, sub-hourly data samples from the previous three days should be considered.

Hybrid CNN-SVM Classifier Approaches to Process Semi-Structured Data in Sugarcane Yield Forecasting Production

Information communication technology (ICT) breakthroughs have boosted global social and economic progress. Most rural Indians rely on agriculture for income. The growing population requires modern agricultural practices. ICT is crucial for educating farmers on how to be environmentally friendly. It helps them create more food by solving a variety of challenges. India’s sugarcane crop is popular and lucrative. Long-term crops that require water do not need specific soil. They need water; the ground should always have adequate water due to the link between cane growth and evaporation. This research focuses on forecasting soil moisture and classifying sugarcane output; sugarcane has so many applications that it must be categorized. This research examines these claims: The first phase model predicts soil moisture using two-level ensemble classifiers. Secondly, to boost performance, the proposed ensemble model integrates the Gaussian probabilistic method (GPM), the convolutional neural network (CNN), and support vector machines (SVM). The suggested approach aims to correctly anticipate future soil moisture measurements affecting crop growth and cultivation. The proposed model is 89.53% more accurate than conventional neural network classifiers. The recommended models’ outcomes will assist farmers and agricultural authorities in boosting production.

Research on soil moisture prediction based on VAR-ARIMA model

This study uses precipitation, soil moisture and evapotranspiration data as independent variables to predict future soil moisture in the Xilingol grassland of Inner Mongolia, China, while keeping the grazing strategy unchanged. Firstly, the data with anomalous values in the obtained dataset were differenced, and the soil moisture data for different depths were split to build a multi-group ARIMA prediction model. The reliability of the model was determined to obtain the soil moisture in grassland at different depths from 2022 to 2023. Through this study, we have added VAR to the ARIMA model to improve the limitations of predicting future soil moisture based on a single variable and to improve the accuracy of soil moisture prediction, thus providing some theoretical support for the ecological restoration and sustainable development of grasslands.

Development of a Soil Moisture Prediction Model Based on Recurrent Neural Network Long Short-Term Memory (RNN-LSTM) in Soybean Cultivation.

Due to climate change, soil moisture may increase, and outflows could become more frequent, which will have a considerable impact on crop growth. Crops are affected by soil moisture; thus, soil moisture prediction is necessary for irrigating at an appropriate time according to weather changes. Therefore, the aim of this study is to develop a future soil moisture (SM) prediction model to determine whether to conduct irrigation according to changes in soil moisture due to weather conditions. Sensors were used to measure soil moisture and soil temperature at a depth of 10 cm, 20 cm, and 30 cm from the topsoil. The combination of optimal variables was investigated using soil moisture and soil temperature at depths between 10 cm and 30 cm and weather data as input variables. The recurrent neural network long short-term memory (RNN-LSTM) models for predicting SM was developed using time series data. The loss and the coefficient of determination (R2) values were used as indicators for evaluating the model performance and two verification datasets were used to test various conditions. The best model performance for 10 cm depth was an R2 of 0.999, a loss of 0.022, and a validation loss of 0.105, and the best results for 20 cm and 30 cm depths were an R2 of 0.999, a loss of 0.016, and a validation loss of 0.098 and an R2 of 0.956, a loss of 0.057, and a validation loss of 2.883, respectively. The RNN-LSTM model was used to confirm the SM predictability in soybean arable land and could be applied to supply the appropriate moisture needed for crop growth. The results of this study show that a soil moisture prediction model based on time-series weather data can help determine the appropriate amount of irrigation required for crop cultivation.

Study on the effect of different grazing strategies on soil moisture and vegetation biomass

A reasonable grazing policy of "returning grazing to grass" is the key to boosting the regional economy, preventing desertification of grasslands and safeguarding people's livelihoods. In this paper, the Woodward model is used to establish the quantitative relationship between grazing and plant growth from the perspective of mechanistic analysis. To analyse the effects of different grazing strategies on the physical properties of Xilinguole grassland, this paper establishes a water balance equation for the soil-vegetation-atmosphere system and obtains a mathematical model for the effects of the physical properties of soil and vegetation biomass in Xilinguole grassland. Finally, on the basis of keeping the current grazing strategy unchanged, a model is developed to predict the future soil moisture.

Water scarcity down to earth surface in a Mediterranean climate: The extreme future of soil moisture in Portugal

Climate change constitutes a major threat for all the Mediterranean countries due to the combination of large precipitation reductions and temperature increases and the higher frequency of climate extremes, especially driving water scarcity and all the derived multi-sectoral impacts. Portugal, as most of the Mediterranean countries, already endures larger frequencies of droughts and deficits in soil moisture and water storage. In the current study, the future projections of soil moisture are examined using a multi-model EURO-CORDEX regional climate ensemble, in agreement with three future emission scenarios (RCP2.6, 4.5 and 8.5). The drivers of future soil moisture dynamics are also analysed and its effect on relative humidity and evaporation rates. As expected, the projections show a clear reduction of soil moisture through the entire annual cycle, in response to the large decrease in precipitation and temperature increase, via a massive growth of potential evapotranspiration. The overall total soil moisture decreases ranges from −5% for the RCP2.6 to −20% (-10%) for the RCP8.5 (RCP4.5), w.r.t. the present climate. In the historical period, soil moisture deficits rarely reach values 3x over the standard deviation, but projections reveal that for the RCP4.5 (RCP8.5) for the mid-century deficits up to 5x (6x) are projected to occur, and for the end-of-century even 7x for the RCP8.5. The annual cycle of soil moisture is in present and future climate determined by precipitation and potential evapotranspiration, and deficit is both enhanced and covers a wider monthly window in the future, especially for the RCP8.5. The surface humidity also decreases importantly, up to −4% and −8% in spring and summer in the end-of-the-century, in agreement with RCP4.5 and RCP8.5, respectively. Resulting from the projected changes in precipitation and potential evapotranspiration, the typical semi-arid climate, which in present climate is confined to a small south-eastern region of Portugal, is expected to cover almost 2/3 of the mainland in the case of RCP8.5. Finally, this study was developed in the framework of the National Roadmap for Adaptation XXI - Portuguese Territorial Climate Change Vulnerability Assessment for XXI Century (RNA2100) project and aims at delivering a deeper and different featuring of terrestrial water for adaptation purposes in a Mediterranean country.

Climate change-induced peatland drying in Southeast Asia

When organic peat soils are sufficiently dry, they become flammable. In Southeast Asian peatlands, widespread deforestation and associated drainage create dry conditions that, when coupled with El Niño-driven drought, result in catastrophic fire events that release large amounts of carbon and deadly smoke to the atmosphere. While the effects of anthropogenic degradation on peat moisture and fire risk have been extensively demonstrated, climate change impacts to peat flammability are poorly understood. These impacts are likely to be mediated primarily through changes in soil moisture. Here, we used neural networks (trained on data from the NASA Soil Moisture Active Passive satellite) to model soil moisture as a function of climate, degradation, and location. The neural networks were forced with regional climate model projections for 1985–2005 and 2040–2060 climate under RCP8.5 forcing to predict changes in soil moisture. We find that reduced precipitation and increased evaporative demand will lead to median soil moisture decreases about half as strong as those observed during recent El Niño droughts in 2015 and 2019. Based on previous studies, such reductions may be expected to accelerate peat carbon emissions. Our results also suggest that soil moisture in degraded areas with less tree cover may be more sensitive to climate change than in other land use types, motivating urgent peatland restoration. Climate change may play an important role in future soil moisture regimes and by extension, future peat fire in Southeast Asian peatlands.

Climate Change Increases the Severity and Duration of Soil Water Stress in the Temperate Forest of Eastern North America

Under climate change, drought conditions are projected to intensify and soil water stress is identified as one of the primary drivers of the decline of forests. While there is strong evidence of such megadisturbance in semi-arid regions, large uncertainties remain in North American temperate forests and fine-scale assessments of future soil water stress are needed to guide adaptation decisions. The objectives of this study were to (i) assess the impact of climate change on the severity and duration of soil water stress in a temperate forest of eastern North America and (ii) identify environmental factors driving the spatial variability of soil water stress levels. We modeled current and future soil moisture at a 1 km resolution with the Canadian Land Surface Scheme (CLASS). Despite a slight increase in precipitation during the growing season, the severity (95th percentile of absolute soil water potential) and duration (number of days where absolute soil water potential is greater than or equal to 9,000 hPa) of soil water stress were projected to increase on average by 1,680 hPa and 6.7 days in 80 years under RCP8.5, which correspond to a 33 and 158% increase compared to current levels. The largest increase in severity was projected to occur in areas currently experiencing short periods of soil water stress, while the largest increase in duration is rather likely to occur in areas already experiencing prolonged periods of soil water stress. Soil depth and, to a lesser extent, soil texture, were identified as the main controls of the spatial variability of projected changes in the severity and duration of soil water stress. Overall, these results highlight the need to disentangle impacts associated with an increase in the severity vs. in the duration of soil water stress to guide the management of temperate forests under climate change.

Soil Moisture Conditions Determine Land‐Atmosphere Coupling and Drought Risk in the Northeastern United States

AbstractStrengthened land‐atmosphere coupling in the northeastern United States (NEUS), accompanied by a positive soil moisture‐rainfall feedback, may lead to more drought. Coupling between the land and atmosphere emerges when low soil moisture values limit surface latent heat flux, or evapotranspiration, so that a majority of absorbed solar radiation is emitted from the surface as sensible heat. In this study, the Weather Research and Forecasting model was run with four prescribed soil moisture levels across 7 years to elucidate the strength of land‐atmosphere coupling under potential, future soil moisture states in the NEUS. Under drier conditions, land‐atmosphere coupling strengthens, and a positive soil moisture‐precipitation feedback develops in all years despite differences in the amount of moisture advected into the study domain. As snowpack decreases and evaporative demand increases, soil conditions may become drier in future summers over the NEUS, resulting in the more frequent development of drought.

Evaluation of Soil Water Content Using SWAT for Southern Saskatchewan, Canada

Soil water content (SWC) is one of the most important hydrologic variables; it plays a decisive role in the control of various land surface processes. We simulated SWC using a Soil and Water Assessment Tool (SWAT) model in southern Saskatchewan. SWC was calibrated using measured data and Soil Moisture Active Passive (SMAP) Level-4 for the surface (0–5 cm) SWC for hydrological response units (HRU) at daily and monthly (warm season) intervals for the years 2015 to 2020. We used the SUFI-2 technique in SWAT-CUP, and observed daily instrumented streamflow records, for calibration (1995 to 2004) and validation (2005–2010). The results reveal that the SWAT model performs well with a monthly PBIAS < 10% and Nash–Sutcliffe efficiency (NS) and R2 ≥ 0.8 for calibration and validation. The correlation coefficient between ground measurement with SMAP and SWAT products are 0.698 and 0.633, respectively. Moreover, SMAP data of surface SWC coincides well with measurements in terms of both amount and trend compared with the SWAT product. The highest r value occurred in July when the mean r value in SWAT and SMAP were 0.87 to 0.84, and then in June for r value of 0.75. In contrast, the lowest values were in April and May (0.07 and 0.04, respectively) at the beginning of the growing season in southern Saskatchewan. Furthermore, calibration in the SWAT model is based on a batch form whereby parameters are adjusted to corresponding input by modifying simulations with observations. SWAT underestimates the abrupt increase in streamflow during the snowmelt months (April and May). This study achieved the objective of developing a SWAT model that simulates SWC in a prairie watershed, and, therefore, can be used in a subsequent phase of research to estimate future soil moisture conditions under projected climate changes.

Field Data Forecasting Using LSTM and Bi-LSTM Approaches

Water, an essential resource for crop production, is becoming increasingly scarce, while cropland continues to expand due to the world’s population growth. Proper irrigation scheduling has been shown to help farmers improve crop yield and quality, resulting in more sustainable water consumption. Soil Moisture (SM), which indicates the amount of water in the soil, is one of the most important crop irrigation parameters. In terms of water usage optimization and crop yield, estimating future soil moisture (forecasting) is an essentially valuable task for crop irrigation. As a result, farmers can base crop irrigation decisions on this parameter. Sensors can be used to estimate this value in real time, which may assist farmers in deciding whether or not to irrigate. The soil moisture value provided by the sensors, on the other hand, is instantaneous and cannot be used to directly compute irrigation parameters such as the best timing or the required water quantity to irrigate. The soil moisture value can, in fact, vary greatly depending on factors such as humidity, weather, and time. Using machine learning methods, these parameters can be used to predict soil moisture levels in the near future. This paper proposes a new Long-Short Term Memory (LSTM)-based model to forecast soil moisture values in the future based on parameters collected from various sensors as a potential solution. To train and validate this model, a real-world dataset containing a set of parameters related to weather forecasting, soil moisture, and other related parameters was collected using smart sensors installed in a greenhouse in Chiang Mai province, Thailand. Preliminary results show that our LSTM-based model performs well in predicting soil moisture with a 0.72% RMSE error and a 0.52% cross-validation error (LSTM), and our Bi-LSTM model with a 0.76% RMSE error and a 0.57% cross-validation error. In the future, we aim to test and validate this model on other similar datasets.

Importance of Parameter and Climate Data Uncertainty for Future Changes in Boreal Hydrology

AbstractSoil moisture and evapotranspiration (ET) are important components of boreal forest hydrology that affect ecological processes and land‐atmosphere feedbacks. Future trends in soil moisture in particular are uncertain. Therefore, accurate modeling of these dynamics and understanding of concomitant sources of uncertainty are critical. Here, we conduct a global sensitivity analysis, Monte Carlo parameterization, and analysis of parameter uncertainty and its contribution to future soil moisture and ET uncertainty using a physically based ecohydrologic model in multiple boreal forest types. Soil and plant hydraulic parameters and LAI have the largest effects on simulated summer soil moisture at two contrasting sites. In future scenario simulations, the selection of parameters and global climate model (GCM) choice between two GCMs influence projected changes in soil moisture and ET about as much as the projected effects of climate change in the less sensitive GCM with a late‐century, high‐emissions scenario, though the relative effects of parameters, GCM, and climate vary among hydrologic variables and study sites. Saturated volumetric water content and sensitivity of stomatal conductance to vapor pressure deficit have the most statistically significant effects on change in ET and soil moisture, though there is considerable variability between sites and GCMs. The results of this study provide estimates of: (a) parameter importance and statistical significance for soil moisture modeling, (b) parameter values for physically based soil‐vegetation‐atmosphere transfer models in multiple boreal forest types, and (c) the contributions of uncertainty in these parameters to soil moisture and ET uncertainty in future climates.

LSTM integrated with Boruta-random forest optimiser for soil moisture estimation under RCP4.5 and RCP8.5 global warming scenarios

Future soil moisture (SM) estimation is a practically useful task for eco-hydrologists, agriculturists, and stakeholders in environment health monitoring to generate comprehensive understanding of hydro-physical and soil dynamic system. This paper demonstrates the capability of a hybridised long short-term memory (LSTM) predictive framework to emulate SM under global warming scenarios. The proposed model is developed by integrating Boruta-random forest (BRF) feature selection and capturing significant antecedent memory of SM behaviour were applied to estimate the future SM using Coupled Model Intercomparison Phase-5 (CMIP5) repository. The BRF is adapted to extract pertinent features in hydro-meteorological variables intrinsically related to SM, and therefore, is used to construct a hybridised deep learning (i.e., BRF-LSTM) model. To establish the viability of deep learning model for SM estimation until 2100, five stations closely matched to the global climate model grid are selected in Australia's Murray Darling Basin. The performance skill of BRF-LSTM model is compared against standalone models (i.e., LSTM, SVR, and MARS). The results showed that the hybrid deep learning model (i.e., BRF-LSTM) with a feature selection capability could significantly outperform the standalone models for both warming simulations. The proposed hybrid model also demonstrated superiority in SM estimation with over 95% of all predictive errors lying below 0.02 mm, and low relative root means square error (≈ 1.06% for RCP4.5 and ≈ 1.888% for RCP8.5) to outperform all the benchmark models. This study demonstrates the capability of LSTM algorithm coupled with BRF feature selection to simulate future soil moisture under climate change, and so, can be successfully implemented in hydrology, agriculture, soil use management and environmental management.

Soil moisture retrievals using L-band radiometry from variable angular ground-based and airborne observations

Surface soil moisture is a vital variable in the process of energy exchange between the land and atmosphere. Monitoring the surface soil moisture at the local and global scales has become feasible due to the development of microwave remote sensing. With the new development of potential satellite missions, it is important to evaluate existing soil moisture retrieval algorithms, which will greatly contribute to the improvement of current soil moisture products and the development of new methodologies. This paper compared the performance of four well-known soil moisture retrieval algorithms with L-band radiometry at fixed incidence angles, including the single channel algorithm at horizontal polarization (SCA-H), the single channel algorithm at vertical polarization (SCA-V), the dual channel algorithm (DCA), and the land parameter retrieval model (LPRM). The experimental data used for evaluation was from the Soil Moisture Experiment in the Luan River (SMELR), which consists of variable angular ground-based and airborne observations covering a wide range of incidence angles (22.5°–55°) at L-band. The microwave radiative transfer models are set to be consistent to guarantee that the four different algorithms are comparable. The results showed that the retrieval accuracy of the SCA-H and SCA-V is significantly affected by the input vegetation optical depth, and the calibration of vegetation parameters should be considered in the implementation of the SCA-H and SCA-V. The DCA does not rely on the auxiliary vegetation information and can also achieve good performance for both the soil moisture and vegetation optical depth. However, its retrieval requires a penalty on parameter constraints since the input brightness temperatures at horizontal and vertical polarization are correlated. The LPRM has a poor performance at incidence angles less than 30°, as it analytically utilizes the polarization difference in the brightness temperature, which is quite small at lower incidence angles. The accuracy of four soil moisture algorithms achieve their best performances at intermediate incidence angles of 40° to 45°, and is slightly degraded when the incident angles increased to larger than 50°, which is contributed to the increasing vegetation effects and depolarization that leads to an information loss. These findings provide quantitative evidence to help understand the differences in various current soil moisture algorithms and further promote the development of new methodologies for future soil moisture missions.

Soil organic carbon more strongly related to soil moisture than soil temperature in temperate grasslands

AbstractStatistical models for soil organic carbon (SOC) differ substantially regarding the degree to which climate variables, such as precipitation and air temperature, influence SOC. These discrepancies contribute to uncertainties regarding the potential feedback effects of climate change on SOC. Soil‐climate variables, such as soil moisture and soil temperature, can strongly influence SOC gains and losses, but these variables have not been widely utilized in statistical SOC models. We hypothesized that SOC is more strongly related to soil‐climate variables than to climate variables and that using soil‐climate variables in statistical SOC models would improve SOC estimates. We tested these hypotheses using >10 years of daily climate and soil‐climate data from 67 temperate grassland sites across Oklahoma, USA, along with soil physical properties, and corresponding SOC data at 5‐, 25‐, and 60‐cm depths. Soil moisture was the single most influential predictor variable, as indicated by least absolute shrinkage and selection operator (LASSO) regression analysis. The effects of precipitation and air temperature were insignificant, apart from a weak negative effect of precipitation at 25 cm. Models including soil‐climate variables provided more accurate SOC estimates than models excluding soil‐climate variables and were statistically superior based on the Akaike information criterion. Future soil moisture conditions, not precipitation or air temperature, may be the key determinant of climate change–SOC feedback effects at these grassland sites and similar sites worldwide.

Climate-Driven Shifts in Soil Temperature and Moisture Regimes Suggest Opportunities to Enhance Assessments of Dryland Resilience and Resistance

Assessing landscape patterns in climate vulnerability, as well as resilience and resistance to drought, disturbance, and invasive species, requires appropriate metrics of relevant environmental conditions. In dryland systems of western North America, soil temperature andmoisture regimes have been widely utilized as an indicator of resilience to disturbance and resistance to invasive plant species by providing integrative indicators of long-term site aridity, which relates to ecosystem recovery potential and climatic suitability to invaders. However, the impact of climate change on these regimes, and the suitability of the indicator for estimating resistance and resilience in the context of climate change have not been assessed. Here we utilized a daily time-step, process-based, ecosystem water balance model to characterize current and future patterns in soil temperature and moisture conditions in dryland areas of western North America, and evaluate the impact of these changes on estimation of resilience and resistance. Soil temperature increases in the twenty-first century are substantial, relatively uniform geographically, and robust across climate models. Higher temperatures will expand the areas of mesic and thermic soil temperature regimes while decreasing the area of cryic and frigid temperature conditions. Projections for future precipitation are more variable both geographically and among climate models. Nevertheless, future soil moisture conditions are relatively consistent across climate models for much of the region. Projections of drier soils are expected in most of Arizona and New Mexico, as well as the central and southern U.S. Great Plains. By contrast, areas with projections of increasing soil moisture include northeastern Montana, southern Alberta and Saskatchewan, and many areas dominated by big sagebrush, particularly the Central and Northern Basin and Range and the Wyoming Basin ecoregions. In addition, many areas dominated by big sagebrush are expected to experience pronounced shifts toward cool season moisture, which will create more area with xeric moisture conditions and less area with ustic conditions. In addition to indicating widespread geographic shifts in the distribution of soil temperature andmoisture regimes, our results suggest opportunities for enhancing the integration of these conditions into a quantitative framework for assessing climate change impacts on dryland ecosystem resilience and resistance that is responsive to long-term projections.

Soil moisture forecasting for irrigation recommendation

This study integrates measured soil moisture sensor data, a remotely sensed crop vegetation index, and weather data to train models, in order to predict future soil moisture. The study was carried out on a cotton farm, with wireless soil moisture monitoring equipment deployed across five plots. Lasso, Decision Tree, Random Forest and Support Vector Machine modeling methods were trialled. Random Forest models gave consistently good results (mean 7-day prediction error from 8.0 to 16.9 kPA except in one plot with malfunctioning sensors). Linear regression with two of the most important predictor variables was not as accurate, but allowed extraction of an interpretable model. The system was implemented in Google Cloud Platform and a model was trained continuously through the season. An online irrigation dashboard was created showing previous and forecast soil moisture conditions, along with weather and normalized difference vegetation index (NDVI). This was used to guide operators in advance of irrigation water needs. The methodology developed in this study could be used as part of a closed-loop sensing and irrigation automation system.

A new exploration for accurately quantifying the effect of afforestation on soil moisture: A case study of artificial Robinia pseudoacacia in the Loess Plateau (China)

Soil moisture is crucial for the sustainable development of artificial forests in arid and semi-arid regions, and an accurate quantification of the influence of artificial vegetation on soil moisture has profound implications in the assessment of the ecological effects of artificial vegetation in arid and semi-arid regions. This study analyzed the variations in the soil moisture content (SMC) following the conversion of croplands into forests with Robinia pseudoacacia (RP) at different vegetation ages (5, 20, and 40 years) in the Loess Plateau and attempted to regard abandoned cropland (AL) with different restoration years (5, 20, and 40 years) as control samples to accurately quantify soil moisture depletion in the absence of RP at different ages. The results showed that the SMC of the 0–500-cm soil profile, which the zone that supports the whole growth processes for RP in the study area, was in a state of depletion, with an average water depletion of 40.75% and a maximum water depletion of 68.39%. RP mainly consumes shallow soil moisture (0–100 cm) during the early stage of afforestation, and with increases in forest age, RP mainly consumes soil moisture at increasingly deeper layers (100–500 cm). The largest soil water consumption in the presence of RP was observed at 150–180 cm. As RP forests grow, the water demand increases and shifts from the upper to deeper soil layers, and when the deeper soil water consumption reaches a certain level, the water consumption shifts upward. This phenomenon results in different degrees of drying between surface and deep soil layers over time. Abandoned cropland has a decreased influence on the SMC of the whole profile and appears to benefit water conservation at deeper soil layers in the mid-recovery period. Therefore, the vegetation succession in control plots should be considered to enhance the accuracy of the findings in future soil moisture research and afforestation practice. To maintain the sustainability of afforestation, succession should progress from less to more advanced plants (grasses-shrubs-forests), and the cultivation of shallow-rooted herbaceous plants (the main herb species in abandoned croplands) and natural restoration should be prioritized. Moreover, the ecological importance of natural restoration should be considered, the degree of blind artificial transformation should be reduced, and the strong adaptability and evolutionary ability of tree species to survive should receive proper attention.