This paper developed a probabilistic framework for system level reliability and risk assessment that coupled hydraulic loading with structural response and explicitly modelled cascading interactions and statistical dependence between components. The contribution is a system-level reliability and risk modelling methodology that integrates dynamic cascading interactions, non-stationary design-life reliability accumulation, and system-level optimisation within a unified Monte Carlo architecture. Dynamic Monte Carlo simulation was used to evaluate individual, joint, conditional, and system-scale probabilities of failure across varying flood magnitudes and design lives. Model verification confirmed that discretisation and sampling errors were small relative to parameter-driven variability. Results showed that long-term system reliability arose from the combined influence of flood frequency, exposure duration, and the strength of interaction between interdependent structures. Frequent loading accelerates the accumulation of failure probability through repeated events, whereas rare events contribute more slowly but dominate extreme outcomes, indicating that cumulative reliability cannot be inferred by the linear extrapolation of annual probabilities. In an examined diversion-levee-basin configuration, strong structural coupling amplified vulnerability by contracting joint stability margins and increasing conditional failure probabilities. The system-level optimisation of structural parameters over the examined design life reduced cumulative system failure probability from 0.305 to 0.153, whereas single-component optimisation redistributed risk within the system without reducing total system risk. The framework advances beyond static risk analysis by integrating time-dependent reliability, cascading dependencies, and design-life optimisation for system-scale mitigation.

Urban flooding, particularly in arid and semi-arid regions, has become a growing concern in densely populated metropolitan areas. Roof rainwater harvesting is a form of low impact development that offers an effective solution for mitigating urban flooding, particularly in densely populated metropolitan areas. However, few scientific studies have attempted to model this dynamic, allowing us to increase water collection and reduce the risk of floods, especially in arid and semiarid areas. This study examined the impact of Roof rainwater harvesting on flood severity using the Borselli index of connectivity, which quantifies the degree to runoff connectivity between upstream and downstream areas, reflecting the potential for runoff to contribute to flooding. As a case study, we selected Hamadan, one of the world's oldest cities. The methodology involved quantifying HC under two scenarios: with and without Roof rainwater harvesting implementation. The results demonstrated a substantial reduction in hydrological connectivity when Roof rainwater harvesting was implemented, highlighting the valuable effectiveness of Roof rainwater harvesting. In contrast, without Roof rainwater harvesting, most of the areas, particularly in the middle and northern regions, had a high degree of hydrological connectivity. Overall, with the incorporation of RRWH, there was a significant shift toward increased connectivity. The areas characterized by very high and high degrees of hydrological connectivity in the absence of Roof rainwater harvesting decreased to 11.4 % and 20.4 %, respectively. Conversely, the moderate, low, and very low hydrological connectivity categories experienced increases of 27.9, 28.4, and 11.7 %, respectively. We conclude that this new evaluation of hydrological connectivity can serve as a valuable tool for assessing hydrological connectivity in urban areas and reveal to land managers where Roof rainwater harvesting is not effectively applied and where possible improvements must be made. However, the success of Roof rainwater harvesting in urban areas depends on technical, social, and regulatory support, as well as public acceptance for non-potable uses.

It has become common to compare crop model results in multi-model simulation experiments. In general, one observes a large variability in such studies, which reduces the confidence one can have in such models. It is important to understand the causes of this variability as a first step toward reducing it. For a given data set, the variability in a multi-model study can arise from uncertainty in model structure or in parameter values for a given structure. Previous studies have made assumptions about the origin of parameter uncertainty, and then quantified its contribution, generally finding that parameter uncertainty is less important than structure uncertainty. However, those studies do not take account of the full parameter variability in multi-model studies. Here we propose estimating parameter uncertainty based on open-call multi-model ensembles where the same structure is used by more than one modeling group. The variability in such a case is due to the full variability of parameters among modeling groups. Then structure and parameter contributions can be estimated using random effects analysis of variance. Based on three multi-model studies for simulating wheat phenology, it is found that the contribution of parameter uncertainty to total uncertainty is, on average, more than twice as large as the uncertainty from structure. A second estimate, based on a comparison of two different calibration approaches for multiple models leads to a very similar result. We conclude that improvement of crop models requires as much attention to parameters as to model structure.

To develop agricultural risk management strategies, the early identification of water deficits during the growing cycle is critical. This research proposes a deep learning hybrid approach for multi-step soil moisture forecasting in the Bundaberg region in Queensland, Australia, with predictions made for 1-day, 14-day, and 30-day, intervals. The model integrates Geospatial Interactive Online Visualization and Analysis Infrastructure (Giovanni) satellite data with ground observations. Due to the periodicity, transience, and trends in soil moisture of the top layer, time series datasets were complex. Hence, the Maximum Overlap Discrete Wavelet Transform (moDWT) method was adopted for data decomposition to identify the best correlated wavelet and scaling coefficients of the predictor variables with the target top layer moisture. The proposed 3-phase hybrid moDWT-Lasso-LSTM model used the Least Absolute Shrinkage and Selection Operator (Lasso) method for feature selection. Optimal hyperparameters were identified using the Hyperopt algorithm with deep learning LSTM method. This proposed model’s performances were compared with benchmarked machine learning (ML) models. In total, nine models were developed, including three standalone models (e.g., LSTM), three integrated feature selection models (e.g., Lasso-LSTM), and three hybrid models incorporating wavelet decomposition and feature selection (e.g., moDWT-Lasso-LSTM). Compared to alternative models, the hybrid deep moDWT-Lasso-LSTM produced the superior predictive model across statistical performance metrics. For example, at 1-day forecast, The moDWT-Lasso-LSTM model exhibits the highest accuracy with the highest R2≈0.92469 and the lowest RMSE ≈0.97808, MAE ≈0.76623, and SMAPE ≈4.39700%, outperforming other models. The moDWT-Lasso-DNN model follows closely, while the Lasso-ANN and Lasso-DNN models show lower accuracy with higher RMSE and MAE values. The ANN and DNN models have the lowest performance, with higher error metrics and lower R2 values compared to the deep learning models incorporating moDWT and Lasso techniques. This research emphasizes the utility of the advanced complementary ML model, such as the developed moDWT-Lasso-LSTM 3-phase hybrid model, as a robust data-driven tool for early forecasting of soil moisture.

Floods are a common natural disaster whose severity in terms of duration, water resource volume, peak, and accumulated rainfall-based damage is likely to differ significantly for different geographical regions. In this paper, we first propose a novel hourly flood index (SWRI24−hr−S) derived from normalising the existing 24-hourly water resources index (WRI24−hr−S) in the literature to monitor flood risk on an hourly scale. The proposed SWRI24−hr−S is adopted to identify a flood situation and derive its characteristics, such as the duration (D), volume (V), and peak (Q). The comprehensive result analysis establishes the practical utility of SWRI24−hr−S in identifying flood situations at seven study sites in Fiji between 2014 and 2018 and deriving their characteristics (i.e., D, V, and Q). Secondly, this study develops a vine copula-probabilistic risk analysis system that models the joint distribution of flood characteristics (i.e., D, V, and Q) to extract their joint exceedance probability for the seven study sites in Fiji, enabling probabilistic flood risk assessment. The vine copula approach, particularly suited to Fiji’s study sites, introduces a novel probabilistic framework for flood risk assessment. The results show moderate differences in the spatial patterns of joint exceedance probability of flood characteristics in different combination scenarios generated by the proposed vine copula approach. In the worst-case scenario, the probability of any flood event occurring where the flood volume, peak, and duration are likely to exceed the 95th-quantile value (representing an extreme flood event) is found to be less than 5% for all study sites. The proposed hourly flood index and the vine copula approach can be feasible and cost-effective tools for flood risk monitoring and assessment. The methodologies proposed in this study can be applied to other data-scarce regions where only rainfall data are available, offering crucial information for flood risk monitoring and assessment and for the development of effective mitigation strategies.

In large-scale arable cropping systems, understanding within-field yield variations and yield-limiting factors are crucial for optimizing resource investments and financial returns, while avoiding adverse environmental effects. Sensing technologies can collect various crop and soil information, but there is a need to assess whether they reveal within-field yield constraints. Spatial data regarding grain yields, proximal soil sensing data, and topographical and soil properties were collected from 26 maize (Zea mays L.) growing fields in the U.S. Mid-Atlantic. Apparent soil electrical conductivity (ECa) collected by an on-the-go sensor (Veris) was an effective method for estimating subsoil textural variation and water holding capacity in the Coastal Plain region, which was also the best predictor of spatial yield pattern when combined with surface pH and topographic wetness index in a Random Forest (RF) model. In the Piedmont Plateau region, proximal soil sensors showed a lower correlation to measured soil properties, while topographical properties (aspect and slope) were important estimators of spatial yield patterns in an RF model. In locations where the RF model failed to predict yield variation, soil compaction appeared to be limiting crop yields. In conclusion, the application of RF models using ECa sensors and topographical properties was effective in revealing within-field yield constraints, especially in the Coastal Plain region. On the Piedmont Plateau, the calibration of proximal sensor information needs to be improved with a particular focus on soil compaction.

Classification using spatial data is foundational for hydrological modelling, particularly for ungauged areas. However, models developed from classified land use drivers deliver inconsistent water quality results for the same land uses and hinder decision-making guided by those models. This paper explores whether the temporal variation of water quality drivers, such as season and flow, influence inconsistency in the classification, and whether variability is captured in spatial datasets that include original vegetation to represent the variability of biotic responses in areas mapped with the same land use. An Artificial Neural Network Pattern Recognition (ANN-PR) method is used to match catchments by Dissolved Inorganic Nitrogen (DIN) patterns in water quality datasets partitioned into Wet vs Dry Seasons and Increasing vs Retreating flows. Explainable artificial intelligence approaches are then used to classify catchments via spatial feature datasets for each catchment. Catchments matched for sharing patterns in both spatial data and DIN datasets were corroborated and the benefit of partitioning the observed DIN dataset evaluated using Kruskal Wallis method. The highest corroboration rates for spatial data classification with DIN classification were achieved with seasonal partitioning of water quality datasets and significant independence (p < 0.001 to 0.026) from non-partitioned datasets was achieved. This study demonstrated that DIN patterns fall into three categories suited to classification under differing temporal scales with corresponding vegetation types as the indicators. Categories 1 and 3 included dominance of woodlands in their datasets and catchments suited to classify together change depending on temporal scale of the data. Category 2 catchments were dominated by vineforest and classified catchments did not change under different temporal scales. This demonstrates that including original vegetation as a proxy for differences in DIN patterns will help guide future classification where only spatially mapped data is available for ungauged catchments and will better inform data needs for water modelling.

The aim of this paper is to assess the extent to which the Sad-Kalan watershed in Iran participates in floods and rank the Sad-Kalan sub-watersheds in terms of flooding potential by utilizing multi-criteria decision-making approaches. We employed the entropy of a drainage network, stream power index (SPI), slope, topographic control index (TCI), and compactness coefficient (Cc) in this investigation. After forming a decision matrix with 25 possibilities (sub-watersheds) and 5 evaluation indices, we used four MCDM approaches, including the analytic hierarchy process (AHP), best–worst method (BWM), interval rough numbers AHP (IRNAHP), picture fuzzy with AHP (PF-AHP), and picture fuzzy with linear assignment model (PF-LAM, hereafter PICALAM) algorithms, to rank the sub-watersheds. The study results demonstrated that PICALAM exhibited superior performance compared to the other methods due to its consideration of both local and global weights for each criterion. Additionally, among the methods used (AHP, BWM, and IRNAHP) that showed similar performances in ranking the sub-watersheds, the BWM method proved to be more time-efficient in the ranking process.

Transfer of processed data and parameters to ungauged catchments from the most similar gauged counterpart is a common technique in water quality modelling. But catchment similarities for Dissolved Inorganic Nitrogen (DIN) are ill posed, which affects the predictive capability of models reliant on such methods for simulating DIN. Spatial data proxies to classify catchments for most similar DIN responses are a demonstrated solution, yet their applicability to ungauged catchments is unexplored. We adopted a neural network pattern recognition model (ANN-PR) and explainable artificial intelligence approach (SHAP-XAI) to match all ungauged catchments that flow to the Great Barrier Reef to gauged ones based on proxy spatial data. Catchment match suitability was verified using a neural network water quality (ANN-WQ) simulator trained on gauged catchment datasets, tested by simulating DIN for matched catchments in unsupervised learning scenarios. We show that discriminating training data to DIN regime benefits ANN-WQ simulation performance in unsupervised scenarios ( p< 0.05). This phenomenon demonstrates that proxy spatial data is a useful tool to classify catchments with similar DIN regimes. Catchments lacking similarity with gauged ones are identified as priority monitoring areas to gain observed data for all DIN regimes in catchments that flow to the Great Barrier Reef, Australia.

In the past few decades, the Oceania region has experienced a general trend of rising temperature, increased rainfall variability, declining rainfall in parts of Australia and New Zealand, and increased rainfall, particularly extreme events in other parts of the Oceanian region. These changes pose challenges to agriculture, forcing agricultural management to adapt. Systems modelling has played a key role in understanding how climate variability and change impact on agricultural systems. This chapter summarises the climate change in the recent past and future projections for the Oceanian region, briefly reviews how agriculture has responded to past climate changes, and how systems modelling has helped to quantify impact of future climate change, evaluate adaptation strategies and influence policy. Case studies in Australia, New Zealand and Papua New Guinea are presented and future research needs for modelling are discussed. Web links for further information of relevant research and applications are provided at the end.

Incorporating cover crops into the rotation is a practice applied across many parts of the globe to enhance soil biological activities. In dryland farming, where crop production is highly dependent on rainfall and soil water storage, cover cropping can affect soil water, yet its effects on soil hydrological and biological health require further investigation. The objective of this study was to evaluate the effect of different timing of summer sorghum cover crop termination on soil water, total and labile organic carbon, arbuscular mycorrhizal fungi and their mediating effects on wheat yield. Through on-farm trial, soil characteristics along with wheat biomass, yield and grain quality were monitored. In comparison with the control (fallow), the early terminated cover crop was the most effective at retaining greater soil water at wheat sowing by 1~4% in 0-45cm soil profile. An increase in water use efficiency, yield and grain protein by 10%, 12% and 5% was observed under early termination. Under late terminated summer cover crop, there was 7% soil water depletion at wheat planting which resulted in 61% decline in yield. However, late-terminated cover crop achieved the greatest gain in soil total and particulate organic carbon by 17% and 72% and arbuscular mycorrhizal fungal Group A and B concentration by 356% and 251%. Summer cover crop incorporation resulted in a rapid gain in labile organic carbon, which constituted hotspots for arbuscular mycorrhizal fungi growth, conversely, fungal activities increased labile organic carbon availability. The combined effect of increased soil water at sowing and over the growing season, organic carbon, and microbial activities contributed to greater yield. The findings suggest that summer cover cropping with timely termination can have implications in managing soil water at sowing time and enhancing soil water storage during the season, soil carbon, and facilitating microbial activities while enhancing productivity in the dryland cropping system.

In hydrological modelling, classification of catchments is a fundamental task for overcoming deficits in observational datasets. Most attention on this issue has focussed on identifying the catchments with similar hydrological responses for streamflow. Yet, effective methods for catchment classification are currently lacking in respect to Dissolved Inorganic Nitrogen (DIN), a water quality constituent that, at increasing concentrations, is threatening nutrient sensitive environments. Pattern recognition, using standard Artificial Neural Network algorithm is applied, as a novel approach to classify datasets that are considered to be suitable proxies for biological and anthropogenic drivers of observed DIN releases. Eleven gauged Great Barrier Reef (GBR) catchments within Queensland Australia are classified using spatial datasets extracted from ecosystem (e.g. original ecosystem responses to biogeographic, land zone, land form, and soil type attributes) and land use maps. To evaluate the performance of the examined spatial datasets as a proxy for deductive classification, the classification process is repeated inductively, using observed DIN and streamflow data from gauging stations. The ANN-PR method is seen to generate the same classification score format for the differing dataset types, and this facilitates a direct comparison for model output for observed data corroborations. The Kruskal-Wallis test for independence, at p > 0.05, identifies the deductive classification approach as a predictor for classification using DIN observations, which lacks an independence from each other at a p value of 0.01 and 0.02. This study concludes that an ANN-PR method can integrate the ecosystem and land use mapping data to deductively classify the GBR catchments into four regions that also have similar patterns of DIN concentrations. Due to the uniform availability of the mapping data, the findings provide a sound basis for further investigations into the transposing of knowledge from gauged catchments to ungauged areas.

Forecasting river water levels or streamflow water levels (SWL) is vital to optimising the practical and sustainable use of available water resources. We propose a new deep learning hybrid model for SWL forecasting using convolutional neural networks (CNN), bi-directional long-short term memory (BiLSTM), and ant colony optimisation (ACO) with a two-phase decomposition approach at the 7-day, 14-day, and 28-day forecast horizons. The newly developed CBILSTM method is coupled with complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) and variational mode decomposition (VMD) methods to extract the most significant features within predictor variables to build a hybrid CVMD-CBiLSTM model. We integrate three distinct datasets (satellite-derived, climate mode indices, and ground-based meteorological observations) to improve the forecasting capability of the CVMD-CBiLSTM model, applied at nineteen different gauging stations in the Australian Murray River system. This proposed model returns a significantly accurate performance with ~98% of all prediction errors within less than ±0.020 m and a low relative root mean square of ~0.08%, demonstrating its superiority over several benchmark models. The results show that using the new hybrid deep learning algorithm with ACO feature selection can significantly improve the accuracy of forecasted river water levels, and therefore, the method is attractive for adopting remote sensing data to the model ground-based river flow for strategic water savings planning initiatives and dealing with climate change-induced extreme events such as drought events.

Regenerative grazing, which generally involves some form of rotational grazing with strategic rest, is increasingly seen as a profitable management approach that will accelerate landscape recovery. However, there is limited quantitative evidence supporting the benefits of this approach in northern Australia. This space-for-time study collected vegetation and soil data from a range of properties in the Burdekin catchment in Queensland that have implemented regenerative grazing strategies for between 5 and 20 years. Data were also collected at adjacent control sites that did not undergo regenerative grazing, but where more traditional continuous set-stocking grazing approaches were applied. Coincident data were also collected from several sites where grazing had been excluded for ~30 years. Data suggested that improvements in vegetation, soil and land condition can be obtained from implementing regenerative grazing principles, although it is likely to take at least 3–5 years, and up to 15–20 years for statistically significant improvements to be measurable at a site, particularly for areas that are moving from a degraded baseline condition. Vegetation attributes such as plant biomass and basal area and litter incorporation all appeared to be better surrogates than percentage ground cover for representing improved landscape condition and soil health. Sites that maintained remotely sensed percentage ground cover at or above the minimally disturbed reference benchmark levels for >10 years, as well as having statistically higher biomass, basal area and litter, had significant increases in total nitrogen (TN) and soil organic carbon (SOC) relative to the local control site. Although there are indications that regenerative grazing can lead to improvements in land condition, this study does not enable us to conclude whether regenerative grazing will accelerate improvements compared with other best-practice grazing land management (GLM) approaches, and further research on the social and economic dimensions of regenerative grazing is needed.

Evaporation, as a core process within the global hydrological cycle, requires reliable methods to monitor its variation, for decision-making in agriculture, irrigation systems and dam operations, also in other areas of hydrology and water resource management. Accurate monitoring of pan evaporation (E-p) is one the most popular approaches to understand the evaporative process. This work aims to construct a hybrid Long Short-Term Memory (LSTM) predictive model that is coupled with Neighbourhood Component Analysis for feature selection to predict E-P in drought-prone regions in Queensland, Australia (Amberley, Gatton, Oakey, & Townsville). Utilizing the daily-scale dataset [31 August 2002 to 22 September 2020], the performance of the proposed deep learning (DL) hybrid model, denoted as NCA-LSTM, is compared with competitive benchmark models, i.e., standalone LSTM, other types of DL, single hidden layer neuronal architecture and decision tree-based method. The testing results reveal the lowest Relative Root Mean Square Error ( <= 20%), Absolute Percentage Bias ( <= 14.5%) and the highest Kling-Gupta Efficiency ( >= 87%) attained by the NCA-LSTM hybrid model (relative to benchmark models) tested for Amberley, Gatton, and Oakey sites. In respect to the predictive efficiency, the proposed NCA-LSTM hybrid model, improved with feature selection, outperforms all benchmark models, indicating its future utility in the prediction of daily E-p. In practical sense, the predictive model developed for E-p estimation provides an accurate estimation of evaporative water loss in hydrological cycle and therefore, can be implemented in areas of irrigation management, planning of irrigation-based agriculture, and mitigation of financial losses to agricultural and related sectors where, regular monitoring and forecasting of water resources are a vital part of sustainable livelihood and business.

Long-term forecasting of any hydrologic phenomena is essential for strategic environmental planning, hydrologic and other forms of structural design, agriculture, and water resources management. Climate mode indices, utilising machine learning methods, are frequently considered as predictor variables in order to forecast several different hydrological variables. In this study, a feature selection algorithm based on two different deep learning models, i.e., long short-term memory and a gated recurrent unit, is applied to improve the forecasting capability of streamflow water levels at six gauging stations in the Murray Darling Basin of Australia. This paper therefore aggregates the significant antecedent lag memory of climate mode indices, rainfall, and the monthly factor based on the periodicity as the predictor variables to attain significantly accurate stream water level forecasts. This novel method identifies an improved relationship between the stream water level and climate mode indices through the aggregation of the significant lagged datasets capturing the historical features to predict the future streamflow water level. The boruta feature selection algorithm (BRF) was then applied in a two phase process before and after attaining the significant lagged inputs to screen the optimum predictor variables. The merits of the forecast models were evaluated through different performance evaluation criteria. The results show that the accumulated significant lagged inputs based on climate mode indices, along with the rainfall and periodicity factors are seen to provide improved forecasting of the SWL over the non-BRF deep learning approaches where no prior feature selection was applied. The hybrid LSTM method (i.e., BRF-LSTM model) achieved a unique advantage in terms of SWL forecasting, particularly attaining over 98% of the predictive errors lying within a band of +/-0.015 m with relatively low relative errors (RRMSE approximate to 1.30% and RMAE approximate to 0.882%), outperforming all of the benchmark models. It is also found that the periodicity factor has a potential influence on the accuracy of the forecast models for the four monitored study stations. This study concludes that the newly developed hybrid deep learning approaches, coupled with the BRF feature selection, provide improved forecasting performance. The hybrid approach developed in this paper can therefore be used to provide a strong provide predictive response algorithm for the hydrological variables that were influenced by the low-frequency variability of the climate model indices in respect to streamflow water level.

Accelerated soil erosion is an earth surface process that threats sustainable land management. Among different forms of soil erosion, rills, gullies, and landslides are the most dominant features, which have a considerable contribution in sediment yield at the landscape and catchment scale. Rill erosion may contribute as a significant erosion process in a landscape by the removal of topsoil layer while gully erosion removes the entire soil profile causing extreme land degradation. Therefore, to design effective sediment management strategies, there is a need to having a fundamental knowledge of rill and gully erosion development processes, measurements, and monitoring techniques, in particular, new technologies and approaches. This chapter briefly explores these topics with a greater focus on the definition, significance, and characteristics of rill and gully erosion as well as their measurement methods including both traditional (direct) methods and new (indirect) methods. Finally, the data mining technique is mentioned as a tool for analyzing and evaluating rill and gully erosion.

Dispersive spoil/soil management is a major environmental and economic challenge for active coal mines as well as sustainable mine closure across the globe. To explore and design a framework for managing dispersive spoil, considering the complexities as well as data availability, this paper has developed a Bayesian Belief Network (BBN)-a probabilistic predictive framework to support practical and cost-effective decisions for the management of dispersive spoil. This approach enabled incorporation of expert knowledge where data were insufficient for modelling purposes. The performance of the model was validated using field data from actively managed mine sites and found to be consistent in the prediction of soil erosion and ground cover. Agreement between predicted soil erosion probability and field observations was greater than 74%, and greater than 70% for ground cover protection. The model performance was further noticeably improved by calibration of Conditional Probability Tables (CPTs). This demonstrates the value of the BBN modelling approach, whereby the use of currently best-available data can provide a practical result, with the capacity for significant model improvement over time as more (targeted) data come to hand.

This paper aims to test different methods used for assessing sediment yield indices to identify hotspots and rank sediment yield hotspots. This process includes the assessment of the entropy weighting (EW), fractal dimension (FD), slope length gradient (SL), and sediment connectivity (SC) methods. The indices at different sub-catchment levels were applied in the Ilanlu catchment (Iran) and organized based on five sediment hazard classes. To assess the performance of sediment hazard mapping, a method of superimposition was used and assessed with the erosion potential model (EPM). The superimposition showed that 8, 10, 4 and 9 sub-catchments, based on the degree of susceptibility, obtained the highest results from FD, EW, SL and SC, respectively, compared with the EPM model. The results show that EW and SC methods can achieve greater performance than FD and SL in identifying sediment production hotspots.

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.