Jahangir Alam

To improve preparedness and minimise losses, this paper presents the development of artificial intelligence (AI)-based forecasting models using a large-scale climate index for Victoria (Australia), which is known to be one of the most fire-prone areas in the country. Using an Artificial Neural Network (ANN) approach, this study investigates the nonlinear relationships between SOI and the Forest Fire Danger Index (FFDI) to develop a robust predictive model. Levenberg–Marquardt optimisation through the backpropagation method was employed to train the ANN models. Based on local climate data, FFDI values were calculated for eight locations within southeast Australia, and SOI values of earlier months were correlated with the FFDI values of the later months. A total of 55 years (1965–2019) of monthly SOI and FFDI values were used to train, validate, and test the developed ANN models. The findings show that the developed models can predict future FFDI values, having correlation coefficients ranging 0.71~0.96, 0.70~0.95, and 0.75~0.93 for 1-month, 2-month, and 3-month lagged periods, respectively. As is obvious, one-month-ahead predictions were more accurate than two/three-month-ahead predictions. In general, the stations located in the eastern parts are attributed to higher prediction accuracy than stations located in the western regions, possibly due to their closer proximity to the location from where SOI originates (i.e., southern Pacific). These variations between the stations located in the eastern and western parts may partly exhibit the applicability of FFDI to different vegetation types. However, the outcomes hold potential for informing stakeholders, improving resource allocation for fire preparedness, and mitigating the devastating impacts of bushfires on communities and ecosystems.

Planning, managing and optimising surface water quality is a complex and multifaceted process, influenced by the effects of both climate uncertainties and anthropogenic activities. Developing an innovative and robust decision support framework (DSF) is essential for effective and efficient water quality management, so it can provide essential information on water quality and assist policy makers and water resource managers to identify potential causes of water quality deterioration. This framework is crucial for implementing actions such as infrastructure development, legislative compliance and environmental initiatives. Recent advancements in computational domains have created opportunities for employing artificial intelligence (AI), advanced statistics and mathematical methods for use in improved water quality management. This study proposed a comprehensive conceptual DSF to minimise the adverse effects of extreme weather events and climate change on water quality. The framework utilises machine learning (ML), deep learning (DL), geographical information system (GIS) and advanced statistical and mathematical techniques for water quality management. The foundation of this framework is the outcomes from our three studies, where we examined the application of ML and DL models for predicting water quality index (WQI) in reservoirs, utilising statistical and mathematical methods to find the seasonal trend of rainfall and water quality, exploring the potential connection between streamflow, rainfall and water quality, and employing GIS to show the spatial and temporal variability of hydrological parameters and WQI. Three potable water supply reservoirs in the Toowoomba region of Australia were taken as the study area for practical implementation of the proposed DSF. This framework can serve as a comprehensive mechanism to identify distinct seasonal characteristics and understand correlations between rainfall, streamflow and water quality. This will enable policy makers and water resource managers to enhance their decision making processes by selecting the management priorities to safeguard water quality in the face of future climate variability, including prolonged droughts and flooding.

An understanding of the trend and relationship between rainfall patterns and water quality dynamics can provide valuable guidelines for the effective management of water resources. The aim of this study was to reveal the synchronous trends in rainfall and water quality and to explore the potential connection between seasonal variation in rainfall volume and the water quality index. This study scrutinised the seasonal temporal trends of rainfall and water quality parameters of three water supply reservoirs in the Toowoomba region of Australia by applying the modified Mann–Kendall (MMK) test and innovative trend analysis (ITA) methods from data collected over 22 years (2002–2022). The models showed a significant increasing trend of rainfall in two rainfall stations during autumn season. The water quality parameters, such as PO₄³⁻, exhibited a significant decreasing trend in all seasons in three reservoirs. On the other hand, the water quality index (WQI) showed a decreasing trend in the Cooby and Cressbrook reservoirs, excepting the Perseverance reservoir, which exhibited an increasing trend. In addition to the detection of trends, this study investigated the potential correlation between seasonal variation of rainfall volume and the water quality index using the wavelet transform coherence (WTC) method. The data of twelve rainfall stations were brought into this analysis. The WTC analysis displayed an apparent correlation between the water quality index and rainfall pattern for 70% of the rainfall stations across 8–16 periods. The highest coherency was noticed in 8–16 periods from 2002–2022, as observed at both the Cooby Creek rainfall station and in the WQI of the Cooby reservoir. This evaluation revealed the intertwined dynamics of rainfall patterns and water quality, providing a deeper understanding of their interdependence and implications, which might be useful for environmental and hydrological management practices.

Changes in water quality are closely linked to seasonal fluctuations in streamflow, and a thorough understanding of how these variations interact across different time scales is important for the efficient management of surface water bodies such as rivers, lakes, and reservoirs. The aim of this study is to explore the potential connection between streamflow, rainfall, and water quality and propose an optimised ensemble model for the prediction of a water quality index (WQI). This study modelled the changes in five water quality parameters such as ammonia nitrogen (NH₃-N), phosphate (PO₄³⁻), pH, turbidity, total dissolved solids (TDS), and their associated WQI caused by rainfall and streamflow. The analysis was conducted across three temporal scales, weekly, monthly, and seasonal, using a generalised additive model (GAM) in Toowoomba, Australia. TDS, turbidity, and WQI exhibited a significant nonlinear variation with the changes in streamflow in the weekly and monthly scales. Additionally, pH demonstrated a significant linear to weakly linear correlation with discharge across the three temporal scales. For the accurate prediction of WQI, this study proposed an ensemble model integrating an extreme gradient boosting (XGBoost) and Bayesian optimisation (BO) algorithm, using streamflow as an input across the same temporal scales. The results for the three temporal scales provided the best accuracy of monthly data, based on the accuracy metrics R² (0.91), MAE (0.20), and RMSE (0.42). The comparison between the test and predicted data indicated that the prediction model overestimated the WQI at some points. This study highlights the efficiency of integrating rainfall, streamflow, and water quality correlations for WQI prediction, which can provide valuable insights for guiding future water management strategies in similar catchment areas, especially amidst changing climatic conditions.

The severity and frequency of extremes are changing; thus, it is becoming necessary to evaluate the impacts of land cover changes and urbanisation along with climate change. A framework of the Generalised Extreme Value (GEV) method, Google Earth Engine (GEE), and land cover patterns’ classification including Random Forest (RF) and Support Vector Machine (SVM) can be useful for streamflow impact analysis. For this study, we developed a unique framework consisting of a hydrological model in line with the Process-informed Nonstationary Extreme Value Analysis (ProNEVA) GEV model and an ensemble of General Circulation Models (GCMs), mapping land cover patterns using classification methods within the GEE platform. We applied these methods in Southeast Queensland (SEQ) to analyse the maximum instantaneous floods in non-stationary catchment conditions, considering the physical system in terms of cause and effect. Independent variables (DEM, population, slope, roads, and distance from roads) and an integrated RF, SVM methodology were utilised as spatial maps to predict their influences on land cover changes for the near and far future. The results indicated that physical factors significantly influence the layout of landscapes. First, the values of projected evapotranspiration and rainfall were extracted from the multi-model ensemble to investigate the eight GCMs under two climate change scenarios (RCP4.5 and RCP8.5). The AWBM hydrological model was calibrated with daily streamflow and applied to generate historical runoff for 1990–2010. Runoff was projected under two scenarios for eight GCMs and by incorporating the percentage of each land cover into the hydrological model for two horizons (2020–2065 and 2066–2085). Following that, the ProNEVA model was used to calculate the frequency and magnitude of runoff extremes across the parameter space. The maximum peak flood differences under the RCP4.5 and RCP8.5 scenarios were 16.90% and 15.18%, respectively. The outcomes of this study suggested that neglecting the non-stationary assumption in flood frequency can lead to underestimating the amounts that can lead to more risks for the related hydraulic structures. This framework is adaptable to various geographical regions to estimate extreme conditions, offering valuable insights for infrastructure design, planning, risk assessment, and the sustainable management of future water resources in the context of long-term water management plans.

The effective management of surface water bodies, such as rivers, lakes, and reservoirs, necessitates a comprehensive understanding of water quality status. Altered precipitation patterns due to climate change may significantly affect the water quality and influence treatment procedures. This study aims to identify the most suitable water quality prediction models for the assessment of the water quality status for three water supply reservoirs in Toowoomba, Australia. It employed four machine learning and two deep learning models for determining the Water Quality Index (WQI) based on five parameters sensitive to rainfall impact. Temporal WQI variations over a period of 22 years (2000–2022) are scrutinised across 4 seasons and 12 months. Through regression analysis, both machine learning and deep learning models anticipate WQI gauged by seven accuracy metrics. Notably, XGBoost and GRU yielded exceptional outcomes, showcasing an R² value of 0.99. Conversely, Bidirectional LSTM (BiLSTM) demonstrated moderate accuracy with results hovering at 88% to 90% for water quality prediction across all reservoirs. The Coefficient of Efficiency (CE) and Willmott Index (d) showed that the models capture patterns well, while MAE, MAPE and RMSE provided good performance metrics for the RFR, XGBoost and GRU models. These models have provided valuable knowledge that can be utilised to assess the adverse consequences of extreme climate events such as shifts in rainfall patterns. These insights can be used to improve strategies for managing water bodies more effectively.

Estimating peak flow for a catchment is commonly undertaken using the design event method; however, this method does not allow for the understanding of uncertainty in the result. This research first presents a simplified method of fragments approach to rainfall disaggregation that ignores the need to consider seasonality, offering a greater diversity in storm patterns within the resulting sub-daily rainfall. By simulating 20 iterations of the disaggregated sub-daily rainfall within a calibrated continuous simulation hydrologic model, we were able to produce multiple long series of streamflow at the outlet of the catchment. With these data, we investigated the use of both the annual maximum and peaks over threshold approaches to flood frequency analysis and found that for a 1-in-100-year annual exceedance probability peak flow, the peaks over threshold method (333 m³/s ± 50 m³/s) was significantly less uncertain than the annual maximum method (427 m³/s ± 100 m³/s). For the 1-in-100-year annual exceedance probability, the median peak flow from the peaks over threshold method (333 m³/s) produced an outcome comparable to the design event method peak flow (328 m³/s), indicating that this research offers an alternative approach to estimating peak flow, with the additional benefit of understanding the uncertainty in the estimation. Finally, this paper highlighted the impact that length and period of streamflow has on peak flow estimation and noted that previous assumptions around the minimum length of gauged streamflow required for flood frequency analysis may not be appropriate in particular catchments.

The frequency and severity of extremes, including extreme precipitation events, extreme evapotranspiration and extreme water storage deficit events, are changing. Thus, the necessity for developing a framework that estimates non-stationary conditions is urgent. The aim of this paper is to develop a framework using the geeSEBAL platform, Generalised Extreme Value (GEV) models and spatiotemporal analysis techniques that incorporate the physical system in terms of cause and effect. Firstly, the geeSEBAL platform has enabled the estimation of actual evapotranspiration (ET_a) with an unprecedented level of spatial-temporal resolution. Following this, the Non-stationary Extreme Value Analysis (NEVA) approach employs the Bayesian method using a Differential Evolution Markov Chain technique to calculate the frequency and magnitude of extreme values across the parameter space. Station and global climate datasets have been used to analyse the spatial and temporal variation of rainfall, reference evapotranspiration (ET_o), ET_a and water storage (WS) variables in the Lockyer Valley located in Southeast Queensland (SEQ), Australia. Frequency analysis of rainfall, ET_a, and water storage deficit for 14 stations were performed using a GEV distribution under stationary and non-stationary assumptions. Comparing the ET_a, ET_o and ERA5 rainfall with station data showed reasonable agreement as follows: Pearson correlation of 0.59–0.75 for ET_a, RMSE of 45.23–58.56 mm for ET_a, Pearson correlation of 0.96–0.97 for ET_o, RMSE of 73.13–87.73 mm for ET_o and Pearson correlation of 0.87–0.92 for rainfall and RMSE of 37.53–57.10 mm for rainfall. The lower and upper uncertainty bounds between stationary and non-stationary conditions for rainfall station data of Gatton varied from 550.98 mm (stationary) to 624.97 mm (non-stationary), and for ERA5 rainfall datasets, 441.30 mm (stationary) to 450.77 mm (non-stationary). The results demonstrate that global climate datasets underestimate the difference between stationary and non-stationary conditions by 9.47 mm compared to results of 73.99 mm derived from station data. Similarly, the results demonstrate less variation between stationary and non-stationary conditions in water storage, followed by a sharp variation in rainfall and moderate variation in evapotranspiration. The findings of this study indicate that neglecting the non-stationary condition in some hydrometeorological variables can lead to underestimating their amounts. This framework can be applied to any geographical area for estimating extreme conditions, providing valuable insights for infrastructure planning and design, risk assessment and disaster management.

Estimating peak flow for a catchment is commonly undertaken using the design event method, however this method does not allow for the understanding of uncertainty in the result. This research first presents a simplified method of fragments approach to rainfall disaggregation that ignores the need to consider seasonality, offering a greater diversity in storm patterns within the resulting sub-daily rainfall. By simulating 20 iterations of the disaggregated sub-daily rainfall within a calibrated continuous simulation hydrologic model, we were able to produce multiple long series of stream flow at the outlet of the catchment. With this data, we investigated the use of both the annual maximum and peaks over threshold approaches to flood frequency analysis and found that for a one in 100 year annual exceedance probability peak flow, the peaks over threshold method (333m3/s ±50m3/s) was significantly less uncertain than the annual maximum method (427m3/s ±100m3/s). For the one in 100 year annual exceedance probability, the median peak flow from the peaks over threshold method (333m3/s) produced an outcome comparable to the design event method peak flow (328m3/s), indicating that this research offers an alternative approach to estimating peak flow, with the additional benefit of understanding the uncertainty in the estimation. Finally, the paper highlighted the impact that length and period of streamflow has on peak flow estimation and noted that previous assumptions around the minimum length of gauged streamflow required for flood frequency analysis may not be appropriate in particular catchments.

Understanding the effects of global change and human activities on water supplies depends greatly on surface water dynamics. A comprehensive examination of the hydroclimatic variations at the transboundary level is essential for the development of any adaptation or mitigation plans to deal with the negative effects of climate change. This research paper examines the hydroclimatic factors that contribute to the desiccation of the Doosti Dam’s basin in the transboundary area using multisensor satellite data from the Google Earth Engine (GEE) platform. The Mann–Kendall and Sens slope estimator test was applied to the satellite datasets to analyse the spatial and temporal variation of the hydroclimate variables and their trend over the transboundary area for 18 years from 2004 to 2021 (as the dam began operating in 2005). Statistical analysis results showed decreasing trends in temperature and an increase in rainfall with respect to station-observed available data. Evapotranspiration and irrigated area development followed the increasing pattern and a slight decrease in snow cover. The results confirmed a large expansion of the irrigated area, especially during the winter growing season. The increase in irrigated cultivated areas during both winter and summer seasons is possibly the main reason for the diversion of water to meet the irrigation requirements of the developed agriculture areas. The approach followed in this study could be applied to any location around the globe to evaluate the hydrological conditions and spatiotemporal changes in response to climate change, trend analysis and human activities.