Search

Extreme hydrological events have had several economic, ecologic, and social impacts on many regions around the world. Although the impacts of floods and droughts are equally important and severe, they are typically treated as separate phenomena due to their hydrological differences. However, in order to address and mitigate their impacts, it is important to analyse and model the interactions between the spatial and temporal characteristics of the events, not only separately but also in a joint framework. Moreover, understanding and simulating the effects of land-use land-cover changes on extreme events dynamics is crucial to improve the development of land use policies and risk management plans. The Central America Dry Corridor (CADC) is one of the regions in the world with the highest vulnerability to floods and droughts due to its marked precipitation seasonality and climate variability. Land use change is also an important variable in this mainly rural area, in which forest cover has declined rapidly during the last decades, modifying basin runoff and affecting extreme events generation. Therefore, this study proposes a methodology to analyse and represent in a joint framework the spatio-temporal characteristics of CADC’s floods and droughts, and identify their relationship with land-use land-cover change patterns. To achieve this, a hybrid modelling framework that integrates Machine Learning (ML) techniques with a spatially distributed hydrological model is presented. It is expected that the integration of ML techniques increases hydrological model capabilities to accurately simulate the effects of land-use land-cover change on floods and droughts propagation. It is also expected that the hybrid model can be used as a tool to assess the effectiveness of different risk management measures and land use policies in floods and droughts mitigation.

Droughts are one of the most dangerous natural hazards that are affecting societies, with an economic impact amounting to over 9 billion euros per year in Europe. Drought events usually originate from a precipitation deficit, which can then cause water shortages, agricultural losses, and environmental degradation. Despite the numerous efforts and recent advances in predicting weather and extreme weather events, accurately forecasting rainfall remains a challenge, especially at sub-seasonal lead-times. In this case, the reference period is short enough for the atmosphere to retain a memory of its initial conditions, but also long enough for oceanic variability to affect atmospheric circulation. However, the relative contribution of climate teleconnections and local atmospheric conditions to the genesis of total precipitation at sub-seasonal scale remains unclear. In this work, we aim to address this gap by advancing the Climate State Intelligence (CSI) framework to examine the impact of both teleconnection patterns and local atmospheric conditions on monthly total precipitation. We then use the information gained to forecast total precipitation with a one-month lead time, and we test three different Machine Learning (ML) models: (i) Extreme Learning Machine (ELM); (ii) Fully Connected Neural Network; (iii) Convolutional Neural Network (CNN). We finally assess the skill of our ML-based precipitation forecasts in predicting the Standardized Precipitation Index (SPI), using the ECMWF Extended Range forecasts as a benchmark. Our framework is developed within the CLImate INTelligence (CLINT) project and applied in the Rhine Delta area, in the Netherlands. Initial findings indicate that combining global and local climate contexts into ML-based models significantly improves state-of-the-art drought forecast accuracy, thus representing a promising option to timely prompt anticipatory drought management measures.

Water managers need accurate rainfall forecasts for a wide spectrum of applications, ranging from water resources evaluation and allocation, to flood and drought predictions. In the past years, several frameworks based on Artificial Intelligence have been developed to improve the traditional Numerical Weather Prediction (NWP) forecasts, thanks to their ability of learning from past data, unravelling hidden relationships among variables and handle large amounts of inputs. Among these approaches, Long Short-Term Memory (LSTM) models emerged for their ability to predict sequence data, and have been successfully used for rainfall and flow forecasting, mainly with short lead-times. In this study, we explore three different multi-variate LSTM-based models, i.e. vanilla LSTM, stacked LSTM and bidirectional LSTM, to forecast daily precipitation for the upcoming 30 days in the area of Rhine Delta, the Netherlands. We use both local atmospheric and global climate variables from the ERA-5 reanalysis dataset to predict rainfall, and we introduce a fuzzy index for the models to account for seasonality effects. The framework is developed within the H2020 project CLImate INTelligence (CLINT), and its outcomes have the potential to improve forecasting precipitation deficit in the study area.

Single extreme hydrometeorological events have been highly studied around the world, however, concerns related to the spatiotemporal variations have extended the studies to look for more insight into space and time dimensions. In this context, of increasing importance are the relations of extreme events properties over multiple spatial and temporal scales. Nevertheless, the study of these relations has not been widely developed. The interaction between events, like floods and droughts with different spatiotemporal characteristics, so far, has still yet to be further studied. Some studies show that there are complex relations linking between both extremes, since occurrences of both are observed in single catchment areas around the world. Furthermore, when analyzing time and space scales for concurrent or successive events, the complexity increase. Recent advances in the spatiotemporal analysis of droughts and floods include tracking approaches, data-driven probabilistic models and machine learning applications. Likewise, new studies have highlighted the usefulness of data mining techniques in extracting knowledge, identifying patterns and detecting anomalies from climate databases.Therefore, the main objective of this research is to characterize and identify spatial and temporal patterns related to extreme hydrometeorological events generation and propagation using data mining techniques. The selected case study is the Magdalena River basin in Colombia. This basin produces most of Colombia’s Gross Domestic Production (GDP), which is highly dependent on the water resource. Because of this, extreme hydrological events such as floods or droughts have a large impact all over the basin.ERA5-Land information (precipitation, temperature, surface pressure and wind U and V components) from 1980-2020 with a resolution of 0.1°x0.1° at multiple time scales (hourly and monthly) was collected for this study. This data was used to identify and characterize extreme hydrometeorological events for multiple time steps and indices thresholds. Temporal, spatial, climatic and geometrical properties of each extreme event region were calculated and stored in a hydrometeorological database. Unsupervised machine learning clustering algorithms (k-means, hierarchical clustering, DBSCAN and spectral) were applied on the database to cluster elements with similar property values. At last, a data mining association rules method (APRIORI) was applied to identify clear patterns between cluster elements of extreme hydrometeorological events. As a main result of this study, is expected an improved understanding of the extreme hydrometeorological events patterns and their associated hydro-climatic processes in the region. This knowledge can help to obtain more accurate and less uncertain estimations of extreme hydrological events, as these are major challenges of many water resources problems, such as monitoring and forecasting.

Understanding the effects of climate change on surface-subsurface hydrology is critical for improving water resources management in a basin. In such cases, the use of hydrological models to quantify and assess water resources is a common practice. With the increasing population and human interventions, the land use changes drastically. The land cover plays a vital role in hydrology as it defines the properties of land surface in the models. So far, majority-of the studies accessing the future climate change consequences on hydrology take into account only the meteorological variables under different climatic projections, neglecting the future land use changes assuming it as static. However, that is not the case, because majority of the earth's surface has altered as a result of human activities, and these changes are represented in models via land use maps.The study presented herein, aims to assess the surface-subsurface response of the catchment under combined effect of meteorological variables and land use future projection. The analysis is performed on the Aa of Weerijs catchment which is a meso-scale transboundary watershed between Belgium and the Netherlands. The future projections of the meteorological variable were obtained from the Royal Netherlands Meteorological Institute (KNMI-14) website for the Netherlands and the same trends were implemented for the Belgium part of the catchment. For the land use, the European Space Agency (ESA) Climate Change Initiative (CCI) Land Cover (LC) maps of the study area for the year 1992 to 2021 were downloaded and linearly projected for the year 2050. The developed projected map was also compared with projected land use map of year 2050 by LUISA (Land Use-based Integrated Sustainability Assessment) modelling platform.To investigate the hydrological regime of the area, the fully distributed physically based hydrological model coupled with a hydrodynamic model using MIKE-SHE and MIKE-11 modeling tools was developed. The base model was set up for the year 2009 to 2016. In addition to discharge, the groundwater heads are used to evaluate the model performance.After setting up the base model, firstly, we analyzed the surface and subsurface response of the catchment considering that the land use in the area is the same as it was in 1992. Secondly, we analyzed the catchment response for the year 2050 by considering the meteorological variables as well as land use future projection.The study provides unique estimates of future climate change and associated hydrological implications. The findings of the study will be valuable to plan and suggest significant modifications in the current strategies for water management in the area. Moreover, it can contribute to the efficient integration of spatial planning with water management.

Machine learning and specifically deep learning has been applied in solving numerous hydrology related problems in the past. Furthermore, extensive research has been done on the evaluation and comparison of performances of different Machine learning techniques applied in solving hydrology related problems. In this research, the possible reasons behind these performance variations are being assessed. The performance of recently introduced deep learning techniques for rainfall-runoff modelling are being evaluated by looking in to the possible modelling set-up and training procedures. Therefore, model set-up and training procedures such as: normalization techniques, input variable selection (feature selection), sampling techniques, model complexity, optimization techniques and random initialization of weights are being examined closely in order to improve the performances of different deep learning techniques for rainfall-runoff modelling. As a result, this study is trying to answer whether these factors have significant effect on the model accuracy.The experiments are being conducted on different deep learning models such as: LSTMs, GRUs and MLPs as well as non-deep learning models such as: XGBoost, Random Forest, Linear Regression and Naïve models. Deep learning frameworks including TensorFlow and Keras are being implemented on Python. For better generalization, study areas from three different climatic zones namely: Bagmati catchment in Nepal, Yuna catchment in Dominican Republic and Magdalena catchment in Colombia are chosen to implement this experimental research. Additionally, in situ meteorological and stream flow data are being used for the rainfall-runoff modelling.The preliminary model results show that model performances in case of Bagmati catchment are higher as compared to the other catchments. The LSTMs and MLPs are performing good with NSE values of 0.71 and 0.72 respectively. Most importantly, the linear regression model was outperforming the other models with NSE up to 0.75 in case of considering 6 days lagged rainfall input. This implies the relationship between daily rainfall and runoff data from Bagmati catchment may not be as complex. On the contrary, the 3-hourly data from Yuna catchment shows results with lower values for the performance metrics. This may be an indication of more complex relationships within the Yuna catchment.This research provides key elements of the modelling process, especially in setting up and training deep learning models for rainfall-runoff modelling. The comparative analysis performed here, provides a basis of performance variations on different basins. This work contributes to the experiences in understanding machine learning requirements for different types of river basins.

Hydrological forecasting is of global importance, especially with the spotted increasing trend of flood-related disasters as seen in the last two decades. The causative rainfall events of these extreme events are primarily analysed in a one-dimensional method. However, through an object-based approach, more data on these rainfall fields can be generated and studied to link them to the hydrological response observed. Through an object-based methodology ST-CORA, features from rate of change of rain intensity in space and time can be extracted by simple visual inspection. Every side of an object provides time variations that can be used as images that contain features not easy to extract. In general, rainfall events in previous studies have used aggregated information, like the duration, area, volume, maximum intensity, and the centroid. In this work, more information is captured that describes the spatial and temporal properties of the event. The main objective of this research is to use these 3D objects and their features with a deep learning model to produce a 15-day hydrological probabilistic forecast for flood prediction.A calibrated version of a large-scale hydrological model (MGB) is used to study an Amazon subbasin. The model is forced with the 50-member perturbed forecast from the TIGGE dataset for the period 2006 to 2014 (from ECMWF). The purpose of using the large-scale model is to better capture the spatio-temporal characteristics over a wider area in an effort to reduce the uncertainty in the analysis. For data-driven models, there is a need for sufficiently large databases, in this case for both the causative rainfall events and the observed hydrological responses. As such, the first two steps relate to the data generation. The first database is developed from the daily streamflow which is generated from the calibrated hydrological model at specific locations of interest with the known higher performance metrics. Second, the ST-CORA methodology is applied to extract the features from the rainfall events in order to develop a database of the rainfall objects. Third, an analysis on the statistics of the features of the objects to understand the rainfall which occurs within the study area. The final part of the research involves the effective use of these features and objects with a deep learning model. From the average annual rainfall from 2001 to 2020, three distinct precipitation patterns are observed. For the streamflow, the subbasin shows a relatively fast response which is captured within a 15-day window.A convolutional LSTM deep learning model is developed to handle 3D rainfall objects as sequences of images representing space time sequences. The outcome of this research contributes to the end-to-end deep learning model which receives the forecasted rainfall as objects and generates a corresponding hydrograph at the area of interest for which it has been trained. A potential contribution of this Conv-LSTM network is that it may provide an efficient and automated approach for streamflow forecasting in basins where there is known complexity and non-linearity, which is especially useful for early warning systems.

Projections indicate that agricultural and hydrological droughts' frequency, severity, and duration are expected to increase globally in the twenty-one century. A better understanding of droughts drivers is key to creating preparedness and resilience to projected events. Typically, droughts are caused by lower precipitation and/or higher evaporation than normal in a region. The region's characteristics and anthropogenic influences may enhance or alleviate the drought events. Evaluating the multiple factors influencing droughts is complex and requires innovative approaches. To address this complexity, this study applies a multivariate approach to evaluate the relationship between ten hydroclimatic characteristics and the severity of agricultural and hydrological droughts. A process-based model (Soil Water Assessment Tool) is used for hydrological modeling. The model outputs (soil moisture and streamflow) are used to calculate the indicators for the drought's analysis: Soil Moisture Deficit Index for agricultural droughts and the Standardized Streamflow Index for hydrological droughts. Then, the Multivariate decision tree approach is applied to evaluate the relevance and relationship between the hydroclimatic characteristics and the agricultural and hydrological drought severity at each subbasin. The approach is applied in the Cesar River basin (Colombia, South America), an area of ecological interest declared RAMSAR site.Study outcomes indicate that evapotranspiration, precipitation, and percolation are the primary drivers of agricultural droughts. Other hydroclimatic parameters such as the curve number, water yield, solid yield, and slope play a relevant role in the subbasin's exposure to agricultural droughts. Subbasins with precipitation lower than 1318 mm, evapotranspiration higher than 1191 mm, percolation higher than 648 mm, and soil yield higher than 101 mm experienced more severe agricultural drought conditions during the period of analysis Regarding hydrological droughts; findings show that evapotranspiration and water yield are principal drivers. Results indicate that precipitation, percolation, and surface runoff also influence the severity of hydrological droughts. Most severe drought conditions during the evaluation period are observed in subbasins with evapotranspiration higher than 826 mm, water yield higher than 9 mm, and precipitation higher than 1398 mm. The outcomes of our analysis indicate that seven out of ten hydroclimatic characteristics evaluated influence the severity of agricultural and hydrological droughts. In addition, the results demonstrate that capturing the non-linear relationships between drivers of droughts and severity allows examining the hydroclimatic characteristics that influence droughts in a region.

The European Centre for Medium-Range Weather Forecasts (ECMWF) provides subseasonal to seasonal (S2S) precipitation forecasts; S2S forecasts extend from two weeks to two months ahead; however, the accuracy of S2S precipitation forecasting is still underdeveloped, and a lot of research and competitions have been proposed to study how machine learning (ML) can be used to improve forecast performance. This research explores the use of machine learning techniques to improve the ECMWF S2S precipitation forecast, here following the AI competition guidelines proposed by the S2S project and the World Meteorological Organisation (WMO). A baseline analysis of the ECMWF S2S precipitation hindcasts (2000–2019) targeting three categories (above normal, near normal and below normal) was performed using the ranked probability skill score (RPSS) and the receiver operating characteristic curve (ROC). A regional analysis of a time series was done to group similar (correlated) hydrometeorological time series variables. Three regions were finally selected based on their spatial and temporal correlations. The methodology first replicated the performance of the ECMWF forecast data available and used it as a reference for the experiments (baseline analysis). Two approaches were followed to build categorical classification correction models: (1) using ML and (2) using a committee model. The aim of both was to correct the categorical classifications (above normal, near normal and below normal) of the ECMWF S2S precipitation forecast. In the first approach, the ensemble mean was used as the input, and five ML techniques were trained and compared: k-nearest neighbours (k-NN), logistic regression (LR), artificial neural network multilayer perceptron (ANN-MLP), random forest (RF) and long–short-term memory (LSTM). Here, we have proposed a gridded spatial and temporal correlation analysis (autocorrelation, cross-correlation and semivariogram) for the input variable selection, allowing us to explore neighbours’ time series and their lags as inputs. These results provided the final data sets that were used for the training and validation of the machine learning models. The total precipitation (tp), two-metre temperature (t2m) and time series with a resolution of 1.5 by 1.5 degrees were the main variables used, and these two variables were provided as the global ECMWF S2S real-time forecasts, ECMWF S2S reforecasts/hindcasts and observation data from the National Oceanic and Atmospheric Administration (Climate Prediction Centre, CPC). The forecasting skills of the ML models were compared against a reference model (ECMWF S2S precipitation hindcasts and climatology) using RPSS, and the results from the first approach showed that LR and MLP were the best ML models in terms of RPSS values. In addition, a positive RPSS value with respect to climatology was obtained using MLP. It is important to highlight that LSTM models performed quite similarly to MLP yet had slightly lower scores overall. In the second approach, the committee model (CM) was used, in which, instead of using one ECMWF hindcast (ensemble mean), the problem is divided into many ANN-MLP models (train each ensemble member independently) that are later combined in a smart ensemble model (trained with LR). The cross-validation and testing of the CMs showed positive RPSS values regarding climatology, which can be interpreted as improved ECMWF on the three climatological regions. In conclusion, ML models have very low – if any – improvement, but by using a CM, the RPSS values are all better than the reference forecast. This study was done only on random samples over three global regions; a more comprehensive study should be performed to explore the whole range of possibilities.

The typical causes of droughts are lower precipitation and/or higher than normal evaporation in a region. The region’s characteristics and anthropogenic interventions may enhance or alleviate these events. Evaluating the multiple factors that influence droughts is complex and requires innovative approaches. To address this complexity, this study employs a combination of modelling and machine learning tools to assess the relationship between hydroclimatic characteristics and the severity of agricultural and hydrological droughts. The Soil Water Assessment Tool is used for hydrological modelling. Model outputs, soil moisture and streamflow are used to calculate the drought indicators for the subsequent drought analysis. Other simulated hydroclimatic parameters are treated as hydroclimatic drivers of droughts. A machine learning technique, the multivariate regression tree approach, is then applied to identify the hydroclimatic characteristics that govern agricultural and hydrological drought severity. The case study is the Cesar River basin (Colombia). Our research indicates that multiple parameters influence the Cesar River basin’s exposure to agricultural and hydrological droughts. Accordingly, the basin can be divided into three distinct areas. First is the upper part of the river valley. Due to precipitation shortfalls and high potential evapotranspiration, this region is very susceptible to agricultural and hydrological droughts. The second area is the middle part of the river valley. This area is likewise very susceptible to agricultural and hydrological droughts; however, severe drought conditions are brought on by inadequate rainfall partitioning and an unbalanced water cycle that favours water loss through percolation and evapotranspiration. Third, the Zapatosa marsh and the Serrania del Perijá foothills present moderate exposure to agricultural and hydrological droughts. Mild drought conditions appear to be related to the capacity of the subbasins to retain water, which lowers evapotranspiration losses and promotes percolation. Our results show that the presented methodology, in combining hydrologic modelling and machine learning techniques, provides valuable information about an interplay between the hydroclimatic factors that influence drought severity in the Cesar River basin.

Water resources managers need to make decisions in a constantly changing environment because the data relating to water resources is uncertain and imprecise. The Robust Optimization and Probabilistic Analysis of Robustness (ROPAR) algorithm is a well-suited tool for dealing with uncertainty. Still, the failure to consider multiple uncertainties and multi-objective robustness hinder the application of the ROPAR algorithm to practical problems. This paper proposes a robust optimization and robustness probabilistic analysis method that considers numerous uncertainties and multi-objective robustness for robust water resources allocation under uncertainty. The Copula function is introduced for analyzing the probabilities of different scenarios. The robustness with respect to the two objective functions is analyzed separately, and the Pareto frontier of robustness is generated. The relationship between the robustness with respect to the two objective functions is used to evaluate water resources management strategies. Use of the method is illustrated on a case study of water resources allocation in the Huaihe River Basin. The results demonstrate that the method opens a possibility for water managers to make more informed uncertainty-aware decisions.

Successful modelling of the groundwater level variations in hydrogeological systems in complex formations considerably depends on spatial and temporal data availability and knowledge of the boundary conditions. Geostatistics plays an important role in model-related data analysis and preparation, but has specific limitations when the aquifer system is inhomogeneous. This study combines geostatistics with machine learning approaches to solve problems in complex aquifer systems. Herein, the emphasis is given to cases where the available dataset is large and randomly distributed in the different aquifer types of the hydrogeological system. Self-Organizing Maps can be applied to identify locally similar input data, to substitute the usually uncertain correlation length of the variogram model that estimates the correlated neighborhood, and then by means of Transgaussian Kriging to estimate the bias corrected spatial distribution of groundwater level. The proposed methodology was tested on a large dataset of groundwater level data in a complex hydrogeological area. The obtained results have shown a significant improvement compared to the ones obtained by classical geostatistical approaches.

Climate change has increased the possibility of more severe and prolonged droughts worldwide, which requires innovative methods to predict their impacts on different sectors such as agriculture. Crop growth models calculate yield and variables related to plant development and are used for crop yield estimation, a useful variable for monitoring drought impacts. Although used for prediction, these crop models are not explicit forecasting models; they are limited to the physical assumptions reflected in their conceptual model. In addition, the input data availability, the spatial and temporal aggregation, and different sources of uncertainty make the crop yield prediction challenging. Given these limitations, machine learning (ML) models are often utilised following a multivariable forecasting approach, but their use with the spatial characteristics of droughts as input data is limited. This research explored the spatial extent of drought as input data for building an approach for predicting seasonal crop yield. This ML approach is made up of two components. The first includes polynomial regression (PR) models, and the second considers artificial neural network (ANN) models. This approach aimed to evaluate both types of ML models (PR and ANN) and integrate them into one operational tool. The logic is as follows: ANN models determine the most accurate predictions, but in practice, issues regarding data retrieval and processing can make the use of equations, i.e. PR, preferable. The proposed approach provides these PR equations with early and preliminary input to perform such calculations. The estimates can be further improved when the ANN models are run with the final input data. The results indicated that the empirical equations (PR) produced good predictions when using drought area as the input. ANN provides better estimates, in general. Research results show that the spatiotemporal changes of drought area and its temporal aggregation provide an important pre-processing alternative to implement ML models for drought impact prediction.

Appropriate water resource allocation schemes are essential for the coordinated and stable development of the basin. Identifying the risks existing in a basin and proposing a robust water resource allocation scheme are of great significance for water resource management in a basin. In this study, the Coupled Robust Optimization and Robust Probabilistic Analysis (CROPAR) algorithm is proposed based on the Robust Optimization and Robust Probabilistic Analysis (ROPAR) algorithm, taking into account the multiple uncertainties of water resources allocation in a basin. First, this study calculates the multi-objective optimal allocation of water resources under certainty. In this study, a single Pareto front is obtained by minimizing the water shortage rate and minimizing the typical pollutant emissions as two objective functions. Then, this study analyzes the frequency and uncertainty of inflow based on historical record data. This study assumes that the basin inflows vary within a certain interval, while the basin has multiple inflows. In this study, the joint probability distribution function of the inflows was constructed with the Copula function, and nine scenarios were generated. Then, the ROPAR algorithm was applied to these nine cases. A total of 9,000 Pareto fronts were calculated through 1,000 Monte Carlo samples for each scenario. Finally, a probabilistic analysis is performed for each scenario to reach a robust optimal solution for a specific scenario according to the robustness criterion. The results show that the CROPAR algorithm can adequately tackle the uncertainty of water allocation in the basin. It helps to make a wide range of risk-based decisions.

Reducing the threat of severe spatiotemporal events like floods and droughts is raising concerns for water resource development and management. The severity of drought and floods increases more due to human interventions. Recent studies have focused on finding long-term solutions that mimic nature's process, while posing no environmental risks and targeting sustainability in traditional approaches. The terminology given in Europe for this natural solution is Nature-Based Solutions (NBS). Some examples of NBS are afforestation or reforestation, storage areas, vegetation buffers, and riparian forest. The main principle of NBS is that they slow down the rate of runoff by boosting interception, infiltration, or storage for flood water, hence mitigating the risk downstream. However, there is still not enough experimental nor theoretical experience on how they could be implemented to optimize their use. The way to represent NBS and the scale of implementation in models and real life is for now a process based on approximated propositions of the empirical knowledge of experts in the field. Although some experience have shown important contributions, this is not enough for an optimal implementation and a complete understanding of all possible outcomes. This is the problem expected to be addressed in this research. The main goal is to construct machine learning models to explore their use as an alternative (surrogate) that will aid in performing multiple scenario analyses of NBS, and quantifying their impacts. This approach will consider spatial and temporal data and create a link between several environmental variables and human actions without explicitly knowing the physical behavior of the system, yet clustering(grouping) behaviors or processes responses to structural properties of the hydrological model representation. The case study area for this research is the Bagmati River Basin of Nepal, covering a catchment area of 2822 sq. km, and the flow is dominated by spring and monsoon rainfall. Soil and Water Assessment Tool (SWAT) is used and the baseline scenario (without the implementation of NBS) is modeled. Different scenarios of afforestation, ponds, and conservation tillage will be intervened in the SWAT model and the changes made by those interventions will be replicated in Artificial Neural Network-Multi Layer Perceptron (ANN-MLP). Several unforeseen scenarios will also be tested in machine learning. Thus, the Spatio-temporal analysis will be done regarding the impact of NBS on the flows, and the machine learning model’s ability to replicate such complex systems will be evaluated. The outcome presented here shows the construction of a SWAT model and the preliminary results of machine learning models capable of promptly predicting changes in flow induced by the adoption of various Nature-Based Solutions. It is anticipated to be a simple, effective, and time-saving way for studying the effectiveness of various Nature-Based Solutions for flood and drought mitigation. Thus, this study contributes to the experiences in interpreting and linking complex hydrological problems in machine learning systems. Keywords: SWAT, Machine learning, Nature-Based Solutions, Hydrological extremes, Spatio-temporal analysis

Preventive Drought Management Measures (PDMM) aim to reduce the chance of droughts and minimise their negative consequences in the short and long term. A wide range of interventions can be considered PDMM, including Nature-Based Solutions, grey infrastructure, land use management, and soil conservation practices, among others. This study intends to apply an optimisation procedure to find optimal combinations and allocations of PDMM that contribute to minimising the agricultural and hydrological drought's severity at a basin scale. To achieve this goal, we coupled the multi-objective genetic algorithm (NSGA-II) with the semi-distributed hydrologic model, Soil and Water Assessment Tool (SWAT). The PDMM evaluated in this study are rainwater harvesting ponds, parallel terraces, forest conservation, grade stabilisation structures and floodplains restoration. Preliminary results indicate that optimal combinations and allocations of PDMM reduce the drought's severity in downstream subbasins. The analysis was developed in the La Vieja basin (West-central Colombia).

Our understanding of bedform processes and their associated effect on bedform roughness is limited, and accounts for large uncertainties in hydraulic roughness computation. It is a standard practice in hydraulic modelling to consider hydraulic roughness as a roughness coefficient and to calibrate the model to this coefficient. Such an approach is empirical and does not well capture the physical processes involved in hydraulic roughness dynamics. When bedforms are present, they can account for a significant portion of hydraulic roughness. Consequently, when bedform transitions occur, an abrupt and significant disruption in the hydraulic roughness regime occurs; affecting our water management applications, such as navigability, flood risk management, sediment transport, etc. Bedform transitions are rarely captured, either in laboratory or in real-scale river channels. As such, our understanding of such transition behaviour is further constrained.In this research, we modelled a CFD physics-based bedform model for the Chiyoda channel, Japan based on previous study of Yamaguchi et al. (2019). The model configuration and results of that study had been validated. The CFD model was initialized at flat-bed condition and run till a dynamic equilibrium in dune regime was obtained. In our research, we captured the bedform in this simulation in each time step, effectively obtaining a timeseries of bed evolution from flatbed regime to dune regime.It is hoped, the use of physics-based CFD models can simulate the physical processes that invoke bedform transitions. As these have not been easily observed in the field or in lab, the simulations can provide an important insight into these complex processes. This is particularly important in the context of changing hydraulic regimes under the changing climate scenario – possibly making past calibrations of river systems incompatible in the future. An alternative to physics-based (CFD) model is the use of a data-driven model (using machine learning techniques). The use of surrogate machine learning models that capture the behaviour of these physics-based (CFD) models, provides an advantage in terms of computational cost and computational time.We also developed a proof-of-concept artificial neural network ML models to predict dune height and mean flow depth respectively based on the CFD model results as input. Several models were built using various combinations of input variables: the lagged values of dune height and mean flow depth, mean flow depth or dune height (alternatively), as well as the present and lagged values of spectral power from Fast Fourier Transform spectral analysis. The lagged values of the predicted variable were the most important input parameters compared to other variables. The use of spectral power as predictive variable did not much improve the results, owing to a strong cross-correlation of the parameter with dune height and mean flow depth.Alternative predictive variables such as stream discharge, Froude number, etc may be considered in future studies to ensure better prediction ability. Validation of these ML and physics-based CFD model results remain a challenge as bedform transition timeseries dataset is not much available. Future outlook of the research in this direction is discussed.

Reservoir operation causes spatiotemporal variations in outflow, which influence the dynamics of downstream aquatic communities. However, empirical evidence of community responses to flow regime (FR) and water quality (WQ) remains limited for dam-regulated rivers. This study focused on identifying the influences of both FR and WQ on metacommunity dynamics downstream of the reservoir. First, the metacommunity dynamics model (MDM) was used to simulate aquatic community dynamics under changing FR and WQ. Then, the flow-ecology relationship was established to identify community response to reservoir outflow. Third, the novel ecological indicators were proposed to evaluate the resilience and resistance of multi-population systems. Finally, the reservoir operating rule curves were optimized by considering tradeoffs between socioeconomic and ecological objectives. The coevolution processes of multi-population systems (fish, phytoplankton, zooplankton, zoobenthos, and macrophytes) were simulated by MDM for each local community. The population densities of stable states showed continuous downward trends with increasing alteration degree of FR and WQ for multi-population systems, and aquatic community systems could be destroyed when alteration reached its acceptable maximum. The greater the alteration degree of FR and WQ, the longer the recovery time from an unstable to a stable state, and the weaker resistance for each population system. The resilience and resistance of downstream multi-population systems can be enhanced by optimizing reservoir outflow. The optimization results illustrated that all the performances of the multiple objectives of water supply, hydropower generation, and ecological benefits were improved by no less than 2.5% compared with the conventional operation. This study provided an approach to identify dual effects of FR and WQ on aquatic community systems, which is helpful in guiding ecological restoration for river ecosystems.

There is a current concern for the health of river ecosystems due to their vulnerability and increasing deterioration from human pressures, as well as the interest in achieving freshwater environmental sustainability in the well-known climate change challenges.Analysis of international monitoring frameworks of river health have highlighted the need to increase data availability and frequency as well as reduce data uncertainty. With this, new aggregation, standardization, and classification methods are required as the development of technologies have grown and reached citizens at different social and cultural levels, their participation have increased in the recent years, showing important time-cost advantages. However, still there are no clear protocols to implement as assessment using mobile phone tools and platforms. This study aims to develop a dynamic framework for smart river health ecosystem monitoring, employing citizen science and remote sensing. This concept uses hydro-morphological and biological river indicators, combined with machine learning algorithms to analyze spatiotemporal data. The smart framework for assessment presented here aims to be provide to 1) Characterized natural and non-natural changes of river ecosystem health; 2) Improve river monitoring methods linking local observation and remote sensing data; 3) Develop databases and data visualization of river condition components; 4) Enable citizens to become a large sensor network to contribute to river health monitoring; and 5) Determine and georeferenced the causes of river health changes to support nature-based solutions for river ecosystem management.

Preventive Drought Management Measures (PDMMs) aim to reduce the chance of droughts and minimize drought-associated damages. Selecting PDMMs is not a trivial task, and it can be asserted that actual contributions to drought alleviation still need to be adequately researched. This study evaluates the effects of three potential PDMMs, namely, rainwater harvesting ponds, forest conservation, and check dams, on agricultural and hydrological drought severity. The Soil Water Assessment Tool is used for hydrological modeling and representing PDMMs. The threshold level method is applied to analyze droughts and evaluate the impact of PDMMs on drought severity. Findings show that rainwater harvesting ponds applied on agricultural land reduce the severity of agricultural droughts and hydrological droughts, particularly during the first months of the drought events observed in the rainy season. Results also reveal that forest conservation contributes to reducing the severity of hydrological droughts by up to 90%. Finally, check dams and ponds in upstream subbasins considerably reduce agricultural and hydrological drought severity in the areas where the structures are applied; however, they exacerbate drought severity downstream. The analysis was developed in the Torola River Basin (El Salvador) for the period spanning 2004 to 2018.

To assess the capability of globally available satellite or remote sensed data products (GASRSDP) for distributed hydrological modelling and expedite their widespread uptake require a comprehensive knowledge-base related to their efficiency, temporal and spatial specifications and extents. Moreover, it is important to assess their performance in setting up hydrological models, their use as forcing data of models or for calibration, validation or evaluation of the model itself, along with an assessment of their limitation.Hydrological models are the key tools for sustainable water management decision-making process. In order to capture the spatio-temporal variation in hydrological fluxes, the input data and representation of physical parameters in hydrological models plays an important role in their credibility. Models require rich amounts of data, which is mostly not readily available in data scare regions. The remotely sensed or satellite derived globally available data products are a vast and rich source of data with continuous addition to daily inventory. This data is widely in use for setting up hydrological models, their calibration, validation, evaluation and improvement.Each day new data products are being released by different agencies. The scientific community is continuously mentioning the use of these data products in scientific articles. Present article does a systematic literature review of the articles over the last 5 years (2016 to 2021) in order to analyse the use of remote sensed / satellite globally available data products for detailed distributed hydrological modelling so that the progress in this context can be ascertain and future directions can be established. The review process was started by sourcing 179 articles from Scopus and 206 articles from Web of Science. After excluding the common and out of scope articles, the full analysis has been performed on about 100 articles. We conclude that the use of GASRSDP for the hydrological modelling of macro-scaled catchments has extensively explored and their performance is being evaluated by many authors while their worth for setting up physically based distributed hydrological model for catchment at meso-scale still need exploration, evaluation and assessments.

Extreme peak runoff forecasting is still a challenge in hydrology. In fact, the use of traditional physically-based models is limited by the lack of sufficient data and the complexity of the inner hydrological processes. Here, we employ a Machine Learning technique, the Random Forest (RF) together with a combination of Feature Engineering (FE) strategies for adding physical knowledge to RF models and improving their forecasting performances. The FE strategies include precipitation-event classification according to hydrometeorological criteria and separation of flows into baseflow and directflow. We used ∼ 3.5 years of hourly precipitation information retrieved from two near-real-time satellite precipitation databases (PERSIANN-CCS and IMERG-ER), and runoff data at the outlet of a 3391-km2 basin located in the tropical Andes of Ecuador. The developed models obtained Nash-Sutcliffe efficiencies varying from 0.86 to 0.59 for lead times between 1 and 6 h. The best performances were obtained for peak runoffs triggered by short-extension precipitation events (2) where infiltration- or saturation-excess runoff responses are well learned by the RF models. Conversely, the forecasting difficulty is associated with extensive precipitation events. For such conditions, a deeper characterization of the biophysical characteristics of the basin is encouraged for capturing the dynamic of directflow across multiple runoff responses. All in all, the potential to employ near-real-time satellite precipitation and the use of FE strategies for improving RF forecasting provides hydrologists with new tools for real-time runoff forecasting in remote or complex regions.

Crop yield is one of the variables used to assess the impact of droughts on agriculture. Crop growth models calculate yield and variables related to plant development and become more suitable for crop yield estimation. However, these models are limited in that specific data are needed for computation. Given this limitation, machine learning (ML) models are often widely utilised instead, but their use with the spatial characteristics of droughts as input data is limited. This research explored the spatial extent of drought (area) as input data for building an approach to predict seasonal crop yield. This ML approach is made up of two components. The first includes polynomial regression (PR) models, and the second considers artificial neural network (ANN) models. In this approach, the purpose was to evaluate both types of ML models (PR and ANN) and integrate them into one operational tool. The logic is as follows: ANN models determine the most accurate predictions, but in practice, issues regarding data retrieval and processing can make the use of equations, i.e. PR, preferable. The proposed approach provides these PR equations to perform such calculations with early and preliminary input. The estimates can be further improved when the ANN models are run with the final input data. The results indicated that the empirical equations (PR) produced good predictions when using drought area as the input. ANN provides better estimates, in general. This research will improve drought monitoring systems for assessing drought effects. Since it is currently possible to calculate drought areas within these systems, the direct application of the prediction of drought effects is possible to integrate by following approaches such as the one presented or similar.

Extreme runoff modeling is hindered by the lack of sufficient and relevant ground information and the low reliability of physically-based models. The authors propose to combine precipitation Remote...

In its three-dimensional (3-D) characterization, drought is approached as an event whose spatial extent changes over time. Each drought event has an onset and end time, a location, a magnitude, and...

Due to the complicated nature of environmental processes, consideration of uncertainty is an important part of environmental modelling. In this paper, a new variant of the machine learning-based me...

Remote sensing and crowdsourcing data are new sensing methods that have the potential to improve significantly inundation modelling. That is especially true in data-scarce situations, for example when resources for acquiring sufficient traditional data are limited or when field conditions are not favourable. Crowdsourced water depths and velocities have been demonstrated to be useful for improving inundation models, ranging from the calibration of 1D models to data assimilation in 2D models. In this study, we aim to further evaluate how much the amount and type of crowdsourced data influence model calibration and validation, in comparison with data from traditional measurements. Further, we aim to assess the effects of combining both sources. For that, we developed a 2D inundation model of the Sontea-Fortuna area, a part of the Danube Delta in Romania. This is a wetland area, where data was collected during two 4-day field campaigns, using boat navigation together with the involved citizens. Citizens obtained thousands of images and videos that were converted into water depth and velocity data, while technicians collected ADCP data. We calibrated and validated the model using different combinations of data (e.g. all water depth data, half water depth and half water velocity). Results indicated that velocity data by themselves did not yield good calibration results, being better used in conjunction with water depths or by combining them into discharge. We also observed that calibration by crowdsourced water depths is comparable to the use of water depths from traditional measurements.

With more machine learning methods being involved in social and environmental research activities, we are addressing the role of available information for model training in model performance. We tested the abilities of several machine learning models for short-term hydrological forecasting by inferring linkages with all available predictors or only with those pre-selected by a hydrologist. The models used in this study were multivariate linear regression, the M5 model tree, multilayer perceptron (MLP) artificial neural network, and the long short-term memory (LSTM) model. We used two river catchments in contrasting runoff generation conditions to try to infer the ability of different model structures to automatically select the best predictor set from all those available in the dataset and compared models’ performance with that of a model operating on predictors prescribed by a hydrologist. Additionally, we tested how shuffling of the initial dataset improved model performance. We can conclude that in rainfall-driven catchments, the models performed generally better on a dataset prescribed by a hydrologist, while in mixed-snowmelt and baseflow-driven catchments, the automatic selection of predictors was preferable.

Many calibrated hydrological models are inconsistent with the behavioral functions of catchments and do not fully represent the catchments’ underlying processes despite their seemingly adequate performance, if measured by traditional statistical error metrics. Using such metrics for calibration is hindered if only short-term data are available. This study investigated the influence of varying lengths of streamflow observation records on model calibration and evaluated the usefulness of a signature-based calibration approach in conceptual rainfall-runoff model calibration. Scenarios of continuous short-period observations were used to emulate poorly gauged catchments. Two approaches were employed to calibrate the HBV model for the Brue catchment in the UK. The first approach used single-objective optimization to maximize Nash–Sutcliffe efficiency (NSE) as a goodness-of-fit measure. The second approach involved multiobjective optimization based on maximizing the scores of 11 signature indices, as well as maximizing NSE. In addition, a diagnostic model evaluation approach was used to evaluate both model performance and behavioral consistency. The results showed that the HBV model was successfully calibrated using short-term datasets with a lower limit of approximately four months of data (10% FRD model). One formulation of the multiobjective signature-based optimization approach yielded the highest performance and hydrological consistency among all parameterization algorithms. The diagnostic model evaluation enabled the selection of consistent models reflecting catchment behavior and allowed an accurate detection of deficiencies in other models. It can be argued that signature-based calibration can be employed for building adequate models even in data-poor situations.

The paper presents the development of a Flood Index (FI) based on the flood hydrograph characteristics, namely flood magnitude ratio, rising curve gradient and time to peak. These characteristic values are normalized to their respective values corresponding to a 100-year flood. The methodology developed to compute FI is applied to three case studies; Kentucky in USA, Oc-gok in the Republic of Korea and Haor in Bangladesh. The obtained results show advantages of the presented methodology over the existing ones. The computed FIs at different locations of the catchment corresponding to different exceedance probabilities provide the summarized understanding of the flooding characteristics of the catchment. The spatial and temporal variation of FI presents a snapshot of flood risk in a catchment and can be used in strategic decision making in flood risk management, for example, in spatial planning, flood zoning and flood event management. The developed methodology can be easily applied to poorly gauged catchments where it is difficult to build accurate flood forecasting models.

Abstract: Modern water resource management requires a more robust flood control operation of cascade reservoirs to cope with a more dynamic external environment, whose ultimate goal is to ensure the robust optimization for multiple purposes. To this end, a number of studies with the theme of flood control operation have developed various methods for robust optimization in the presence of uncertainties and in some cases, they may work well. However, these approaches usually incorporate uncertainty into the flood control objectives or constraints and consequently lack explicit robustness indicators that can assist the decision-makers to fully assess the impact of the uncertainty. In order to construct a mature framework of explicit robust optimization of flood control operation, this study uses the Robust Optimization and Probabilistic Analysis of Robustness (ROPAR) technique to identify the robust flood limited water levels of cascade reservoirs for satisfactory compromise hydropower production and flood control risk taking into account the streamflow variability during the flood season: (1) The Monte Carlo method is employed to sample the input set according to the historical streamflow records; (2) The effective non-dominated sorting genetic algorithm II algorithm (NSGA-II) generates a series of Pareto fronts for each hydrograph sample; (3) the ROPAR technique helps building the empirical distribution of the values of hydropower production corresponding to the chosen levels of flood control risk and carry out probabilistic analysis of the Pareto fronts; (4) the ROPAR technique identifies the final robust solutions according to certain criteria. A reservoirs cascade in the Yangtze River basin, China, is considered as a case study. The presented approach allows for studying propagation of uncertainty from the uncertain inflow to the candidate optimal solutions, and selecting the most robust solution, thus better informing decisions related to reservoir operation.Key words：multi-objective reservoir system, robust optimization, uncertainty, flood control operation, Yangtze River basinReference:Marquez-Calvo, O.O., Solomatine, D.P., 2019. Approach to robust multi-objective optimization and probabilistic analysis: the ROPAR algorithm. J Hydroinform, 21(3): 427-440. DOI:10.2166/hydro.2019.095