Abstract. Neural networks have been shown to be extremely effective rainfall-runoff models, where the river discharge is predicted from meteorological inputs. However, question remains: what these models learned? Is it possible extract information about learned relationships that map inputs outputs, and do mappings represent known hydrological concepts? Small-scale experiments demonstrated internal states of long short-term memory (LSTMs), a particular neural network architecture predisposed...

Abstract We present an approach that uses a deep learning model, in particular, MultiLayer Perceptron, for estimating the missing values of variable multivariate time series data. focus on filling long continuous gap (e.g., multiple months daily observations) rather than individual randomly observations. Our proposed algorithm automated method determining optimal MLP model architecture, thus allowing prediction performance given series. tested our by gaps various lengths (three to three...

Hydro-economic modeling (HEM) addresses research and policy questions from socioeconomic biophysical perspectives under a broad range of water-related topics. Applications HEM include economic evaluations existing new water projects, alternative management actions or policies, risk assessments hydro-climatic uncertainty (e.g., climate change), the costs benefits mitigation and/or adaptation to such events. This paper reviews applications in five different categories: (1) change impacts...

Abstract Wastewater treatment plays a crucial role in removing pollutants. Water conservation and reuse of wastewater help to reduce freshwater use alleviate water stress. However, the extent which conservation, treatment, can contribute stress mitigation is not clear. This study aims investigate impact on both quantity quality China. The investigation based dataset mapping pollutant flows across 32 sectors 31 provinces 2017 7411 plants containing information quality. findings show that...

With the growing use of machine learning (ML) techniques in hydrological applications, there is a need to analyze robustness, performance, and reliability predictions made with these ML models. In this paper we accuracy variability groundwater level obtained from Multilayer Perceptron (MLP) model optimized hyperparameters for different amounts types available training data. The MLP trained on point observations features like levels, temperature, precipitation, river flow various...

To develop appropriate climate change adaptation plans, evidence of the effectiveness measures is required. At a regional scale, however, this information usually lacking. The region Seewinkel in Austria was taken as case study because its extensive agricultural industry and unique ecosystem saline lakes. goal to provide stakeholders with support their process. Adaptation discussed by local were analyzed determine efficacy. A system dynamics (SD) based model developed serve tool for water...

Groundwater resources play an important role for irrigation, particularly in arid and semi-arid regions, where groundwater depletion poses a critical threat to agricultural production associated local livelihoods. However, the relationship between use, farming, poverty, with regards informal mechanisms of management, remains poorly understood. Here, we assess this by developing behavioural model user groups, empirically grounded politically fragile context Tunisia. The integrates biophysical...

Abstract. Accurate hydrologic modeling is vital to characterizing how the terrestrial water cycle responds climate change. Pure deep learning (DL) models have been shown outperform process-based ones while remaining difficult interpret. More recently, differentiable physics-informed machine with a physical backbone can systematically integrate equations and DL, predicting untrained variables processes high performance. However, it unclear if such are competitive for global-scale applications...

Abstract. Accurate hydrological modeling is vital to characterizing how the terrestrial water cycle responds climate change. Pure deep learning (DL) models have shown outperform process-based ones while remaining difficult interpret. More recently, differentiable, physics-informed machine with a physical backbone can systematically integrate equations and DL, predicting untrained variables processes high performance. However, it was unclear if such are competitive for global-scale...

Abstract Revenues from hydropower generation often depend on the operator's ability to provide firm power in presence of uncertain inflows. The primary options available for optimizing revenue are negotiation a contract before operations begin and adjustment reservoir release during operations. Contract strategy optimization closely coupled most appropriately analyzed with stochastic real‐time control methods. Here we use an ensemble‐based approach that provides convenient way construct...

Abstract. Neural networks have been shown to be extremely effective rainfall-runoff models, where the river discharge is predicted from meteorological inputs. However, question remains, what these models learned? Is it possible extract information about learned relationships that map inputs outputs? And do mappings represent known hydrological concepts? Small-scale experiments demonstrated internal states of Long Short-Term Memory Networks (LSTMs), a particular neural network architecture...

Drought hazards have intensified in many world regions during the recent century, exposing multiple environmental and socio-economic systems to increased risks. Nevertheless, estimating drought risk is still challenging due complex links between their potentially disastrous impacts. The recently published JRC European atlas, an outcome of Observatory for Resilience Adaptation (EDORA) project, has utilized a data-driven approach, linking drought’s hazard, vulnerability, exposure...

The Middle East and North Africa (MENA) region is struggling with a continuous decline in water availability, attributed to climate change variability, exacerbating the existing scarcity. At same time, factors such as population growth, urbanization, economic development mismanagement further stress scarce resources. This study aims assess impact of socioeconomic changes on availability use resources related environmental conditions MENA at high spatial temporal resolutions, order provide...

Climate change, land cover water use, and flow regulation are driving river streamflow changes globally, it is crucial to understand the varying contributions of these drivers prevent mitigate harmful impacts caused by alteration. However, previous, scenario-based approaches on this notably uncertain may miss interdependencies between different drivers. Here, overcome shortcomings, we use a large sample observed data globally quantify regime align those against trends in precipitation,...

Streamflow – a key component of the water cycle is experiencing drastic alteration due to human actions. While existing studies have widely assessed global extent and degree this change, understanding its drivers has been limited because previous global-scale approaches largely relied on modelled hypothetical scenarios. Here, we overcome these limitations by providing systematic association analysis streamflow change drivers. We use observed data in 5,163 catchments globally combine them...

Abstract Streamflow—a key component of the water cycle—is experiencing drastic alteration due to human actions. The global extent and degree this change have been widely assessed, but understanding its drivers remains limited because previous global-scale approaches largely relied on modelled hypothetical scenarios. Here, we advance by providing an observation-based association analysis streamflow drivers. We use observed data in 3,293 catchments globally combine them with precipitation,...

Abstract. Water scarcity is one of the most critical global environmental challenges. Addressing this challenge requires implementing economically-profitable and environmentally-sustainable water management interventions across scales globally. This study presents development version ECHO hydro-economic model (ECHO-Global 1.0), for assessing economic performance options. covers 282 subbasins worldwide, includes a detailed representation irrigated agriculture its management, incorporates...

Abstract Unsustainable depletion of aquifer storage and diversion groundwater from downstream users continue to be serious problems, even though their adverse effects are widely recognized. Groundwater involves interactions between economically motivated pumping decisions physical constraints. Here, we investigate these by using optimal control techniques describe the economic agents who share an aquifer. Our approach relies on a multi‐scale description unconfined flow, applied computational...

We present an approach that uses a deep learning model, in particular, MultiLayer Perceptron (MLP), for estimating the missing values of variable multivariate time series data. focus on filling long continuous gap (e.g., multiple months daily observations) rather than individual randomly observations. Our proposed algorithm automated method determining optimal MLP model architecture, thus allowing prediction performance given series. tested our by gaps various lengths (three to three years)...