1. Introduction
Hydrological forecasting is crucial for managing water resources effectively, helping to mitigate the risks associated with water-related hazards like floods and droughts (Brunner et al. 2021; Tang et al. 2016; Shyrokaya et al. 2025; Van Loon et al. 2024). The challenge in producing accurate forecasts stems from the complex interactions within hydrometeorological processes, which occur across a wide range of temporal and spatial scales. This complexity is further compounded by the dynamic impacts of climatic, environmental, and societal changes, such as altered precipitation patterns and extreme events, and human developments (including, e.g., water infrastructure and land use change), adding layers of uncertainty to the forecasting process (Di Baldassarre et al. 2019). Understanding and describing the response of a river system to different meteorological conditions and environmental factors has been a highly challenging problem in hydrology due to the nonlinear processes, particularly when the system response is impacted by human actions (Blöschl et al. 2019).
Advancements in monitoring and forecasting methods are essential to address these challenges, requiring the integration of unconventional sources (crowdsourced data and Earth observations, e.g., Dasgupta et al. 2021a; Nardi et al. 2022; Dasgupta et al. 2022; Nijzink et al. 2018); a diverse array of cutting-edge technologies such as artificial intelligence (AI), machine learning (ML), sensor networks, and advanced computational models (Geer 2021; Kraft et al. 2022; Slater et al. 2023; Xu and Liang 2021; Zhong et al. 2023); sophisticated data assimilation methods (Musuuza et al. 2023; Randrianasolo et al. 2014; Dasgupta et al. 2021b; Karimiziarani and Moradkhani 2023); and high-resolution, reliable climate, and numerical weather prediction models (Hoch et al. 2023; Schär et al. 2020; Lucas-Picher et al. 2021). These technological and scientific advancements promise enhanced predictive capabilities across different time horizons and ungauged or poorly gauged areas (Nearing et al. 2024). However, the transition of these advancements into practical, operational forecasting tools may be impeded by research-to-operations (R2O) challenges related to the effectiveness, efficiency, and integration of the systems into real-world applications.
A key challenge in operationalizing research-centric forecasting systems is scalability. While such systems often perform well under specific conditions, they may not always scale effectively across different hydroclimatic regimes(Giuliani et al. 2020; Pechlivanidis et al. 2020). An additional R2O challenge is the availability and quality of data, with high-quality, (near) real-time datasets primarily available and integrated within research-based settings. Moreover, technical limitations, i.e., computational infrastructure including the required maintenance and support for research-centric software limited due to the project-oriented focus typical of most research projects, remain as important challenges and constraints in operational settings. On top of these limitations, there are several organizational issues that compound these challenges. However, we acknowledge that some well-established forecasting agencies, particularly those operating at local and regional scales, have the capacity to integrate high-quality and (near) real-time datasets into operational settings. This is often due to their well-developed human capacity and direct access, control, or ownership of such data, which facilitates their use in real-time forecasting and decision-making.
Bridging the R2O gaps requires a transdisciplinary approach based on collaborative efforts between researchers, operational forecast practitioners, users, and policymakers, ensuring that new methods, tools, and systems are relevant, applicable, and tangible to real-world water management challenges (Lavers et al. 2020; White et al. 2022). Operational forecasting systems face significant hurdles in incorporating hydrological research due to the discrepancy between the granularity of scientific models and the scalability needed for operational contexts, which requires the ability to perform effectively and efficiently across different scales and scopes of operations. The rapid pace of technological advancements often outstrips the capacity of existing operational frameworks to adapt, resulting in a lag in the adoption of state-of-the-art methods. For example, the ECMWF’s advanced AI-based forecasting system, while globally available, transitioned from a research product to operational status only recently (February 2025). Successfully integrating new findings and tools into established operational practices not only demands technical adjustments and an open model integration approach but also shifts in organizational and cultural norms, as well as comprehensive capacity development, training programs, and institutional support to facilitate a smooth transition for managing institutions and their dependent organizations.
The concept of cocreation in establishing more user-tailored operational hydrological forecasting systems marks a shift toward a more integrative and participatory approach between service providers and stakeholders. By involving a diverse array of stakeholders—including researchers, practitioners, and users—right from the outset, forecasting systems are designed to be more aligned with actual operational needs and constraints (Cantone et al. 2023; Crochemore et al. 2024; Kidd et al. 2017; Vincent et al. 2018). This collaborative framework ensures the scientific robustness of these systems and enhances their real-world applicability, leading to tools that are both relevant and user-friendly. Cocreated hydrological forecasting systems offer numerous benefits, such as increased relevance to stakeholders through customization on real-world applications, enhanced usability, and greater acceptance among users. Building on this, these systems facilitate a smoother transition of research findings into operational practices, supporting and, ideally, improving the decision-making processes in water resource management (Lienert et al. 2022; Samaniego et al. 2019). Additionally, the collaborative nature of cocreation promotes the rapid integration of operational needs into research project designs. It facilitates the incorporation of the sociocultural and political contexts of users, and their (traditional) local knowledge, enhancing innovation uptake and enabling the direct application of scientific research into operational forecasting. Cocreation addresses complex hydrological challenges more effectively and enhances the resilience and responsiveness of forecasting to both environmental changes and societal needs (Vanelli et al. 2022).
Despite important scientific and technological developments in hydrological forecasting, globally our society is still challenged by the increased frequency and intensity of extreme weather events and other natural hazards (Banholzer et al. 2014; Coronese et al. 2019). Extreme events pose significant risks to lives, livelihoods, and economic stability, particularly in vulnerable regions. In an effort to mitigate these risks, the United Nations (UN) launched the Early Warnings for All (EW4All) initiative in 2022 (https://www.un.org/en/climatechange/early-warnings-for-all), ensuring that every person on the planet has access to timely and effective multihazard early warnings by 2027, enhancing global resilience and the ability to respond to natural hazards promptly (Basher 2006; Baudoin et al. 2016; United Nations 2022). To ensure the initiative’s success by promoting proactive responses to disasters, reducing the potential loss of life and property, and increasing resilience in communities worldwide, EW4All aims to enhance disaster risk knowledge, improve monitoring and warning services, ensure effective communication of warnings, and boost preparedness and response capabilities at all levels.
HEPEX (Schaake et al. 2007) is an international initiative and a Community of Practice (CoP) dedicated to enhancing the science and application of hydrological ensemble predictions to improve decision-making under uncertainty, particularly under extreme conditions. By bringing together researchers, practitioners, and stakeholders across weather-, climate-, water- and environment-related sectors, HEPEX has been instrumental in bridging the R2O gap, fostering cocreation within water services and underpinning EW4All. Through a variety of initiatives, including workshops, collaborative projects, and the dissemination of guidance materials, HEPEX promotes the participatory development of hydrological ensemble prediction systems tailored to user needs (Dasgupta et al. 2023). This collaborative effort not only facilitates the integration of cutting-edge scientific methods and ensemble forecasting techniques into operational practices but also emphasizes the significance of uncertainty quantification and communication in decision support tools. Consequently, HEPEX’s contributions have significantly propelled the field of hydrological ensemble forecasting forward, ensuring that it is better equipped to meet the evolving requirements of water resource and emergency response management in the face of extreme weather and climate variability.
Here, we identify and present the key priorities for (co)creating R2O hydrological forecast systems that add value to decision-making and emergency management. We provide insights by addressing the following questions to add scientific value and guide developments in hydrological forecasting systems:
- (i)What science is needed to deliver systems that add value across spatial scales and time horizons?
- (ii)What has the HEPEX CoP already achieved that has contributed to R2O? and
- (iii)How can HEPEX contribute to global initiatives such as the EW4All addressing hydrological extremes?
These questions have been discussed in parallel within five breakout groups at the HEPEX 2023 workshop (“Forecasting across spatial scales and time horizons,” which was held from 13 to 15 September 2023, Norrköping, Sweden), and the contributions have been collected, analyzed, and prioritized. About 70 hydrology experts from around the world participated on-site in the workshop (see Fig. A1 in the appendix) and shared their knowledge to address the above questions, while also recommending scientific articles and reports that justify their responses.
2. Priorities for enhancing hydrological forecasting systems across spatial and temporal scales
In the evolving landscape of hydrological forecasting, the integration of advanced scientific data, methods, and transdisciplinary collaboration is paramount in enhancing the efficacy and usability of forecasting systems and forecasting information. This narrative explores five key topics that, if given priority, have the potential to transform the field, building on foundational works and recent advancements in hydrology research (see Fig. 1).
Conceptual diagram with the five top priorities for (co)creating R2O hydrological forecasting systems that add value across spatial scales and time horizons.
Citation: Bulletin of the American Meteorological Society 106, 5; 10.1175/BAMS-D-24-0322.1
a. Data quality and integration.
Hydrological forecasting relies fundamentally on the quality and integration of hydrometeorological data. In particular, physical process–based hydrological models require observational data to initialize and constrain the models, and specify boundary conditions, as well as define parameterizations and effective parameter values. The process of hydrological modeling typically starts with data collection, a central phase where various types of hydrometeorological forcing data and other data to parameterize the model(s) are gathered. Following data collection, data curation is an essential stage during which data undergo rigorous quality control to ensure their reliability and usability as inputs to hydrological models; poor data quality typically results in inaccurate representations of reality by all types of models, ranging from conceptual to data-driven. This step is vital for maintaining the integrity of the data from its point of origin, involving the establishment of protocols to address data anomalies, such as missing data or inconsistencies (Bourgin et al. 2014; Paiva et al. 2013; Pappenberger et al. 2019). Moving beyond collection, quality assurance, and curation, effective model calibration depends significantly on the data, ensuring that models deliver reliable and actionable insights into the water cycle. Moreover, data assimilation, seen also as a dynamic calibration of parameters, represents a key phase where observational data are integrated to update model states, parameters, or initializations (Dasgupta et al. 2021a; Geer 2021; Musuuza et al. 2023). This integration is important for providing an accurate representation of current hydrological conditions and for making predictions about future states.
Insufficient input data, characterized by a lack of long-term records with high spatiotemporal resolution, pose a significant challenge in hydrological forecasting (Kidd et al. 2017). This challenge is exacerbated by the high dependence on accurate meteorological inputs to drive models effectively, while observational uncertainty becomes particularly problematic in highly heterogeneous environments. To effectively address data quality challenges, strategic design and continuous adaptation of the monitoring networks, augmented by advanced data processing technologies like machine learning and the incorporation of soft data from citizen science initiatives, are all likely to play important roles into meeting evolving stakeholder needs and to optimize costs. Maintaining existing observation networks is also crucial, along with enhancing remote sensing capabilities, and employing alternative sensing techniques alongside conventional in situ measurements. The incorporation of diverse conventional and nonconventional data sources, including high-resolution data from remote sensors, precise ground-based observations, as well as data from private weather stations and citizens sensors, low cost sensors, and commercial microwave links, will ideally result in comprehensive datasets capable of capturing the complexities of hydrological systems (Alfieri et al. 2022; Bates 2012; Liu et al. 2012; Randrianasolo et al. 2014). Moreover, advocating for open and Findable, Accessible, Interoperable, and Reusable (FAIR) data principles within and outside the hydrological discipline facilitates transparency and collaborative efforts, enhancing the development, refinement, and intercomparison of hydrological models and forecasting methods.
b. Advanced modeling, algorithms, and tools.
The challenges of modeling ungauged conditions and/or nonstationarity, where historical data are either not available or no longer reliable predictors of future conditions due to shifts in climate patterns, land use, and other factors, underscore the importance of developing flexible, adaptable modeling frameworks. Most hydrological models often struggle with these conditions, constrained by the limitations of past computational capabilities and simplistic assumptions that fail to account for human alterations, influencing water movement and storage in the environment. AI, ML, and advanced statistical methods have proven beneficial across the hydrological modeling and forecasting value chain, enhancing data quality control, generating meteorological forcing input, contributing to process understanding, and summarizing model results. The integration of these techniques and tools has shown potential to revolutionize the field by modeling in a computationally efficient manner the complex spatial and temporal dynamics of hydrological processes (Konapala et al. 2020; Slater et al. 2023; Yang et al. 2020), marking a significant leap forward in our ability to detect, predict, and manage different (extreme) hydrological conditions (Bellier et al. 2018). However, the effectiveness of these techniques and tools hinges on the availability of sufficient data for training; the lack of long observational records complicates the use of AI/ML and their operational deployment. Despite these challenges, the capacity of these advanced models to learn from data and improve over time offers a significant opportunity to enhance the accuracy of hydrological forecasts, making the shift toward data-driven approaches central for harnessing the vast amounts of data now available through modern sensing technologies.
The shift toward leveraging AI/ML in hydrology has the potential to enhance both predictive accuracy and the effective management of predictive systems and resources. Such a shift will require upskilling of hydrological forecasting teams with better integration of data science, ML, and hydrology. This interdisciplinary effort will enable continuously refined models that incorporate the latest scientific advancements and adapt to the evolving landscape of environmental and societal changes. By embracing data-driven methods, the field of hydrological forecasting is moving toward a more dynamic, responsive approach that can adjust to new information and emerging trends (Nearing et al. 2021; Slater et al. 2021, 2023). This proactive shift in model development and application is critical for advancing our understanding of hydrological systems and for making informed decisions in water resource management, disaster preparedness, and environmental conservation. To make this possible, the field should also focus on increasing the trust in knowledge and explainability of AI/ML methods to allow for a smooth R2O (Slater et al. 2024). Increasing trust can occur via benchmarking studies against process-based models with the aim of broader acceptance and enhanced uptake of explainable AI approaches. In this context, verification plays a magnified role in the development of AI/ML methods, as modelers’ traditional reliance on the representation of processes within a model is often not possible. This requires, among other things, stringent cross-validation practices and the standardized use of common test datasets to ensure transparency, reliability, and comparability (Rasp et al. 2020).
c. Ensemble forecasting.
Ensembles have marked a significant advancement in hydrometeorological forecasting by directly addressing the uncertainties associated with predicting future impactful events (Cloke and Pappenberger 2009; Thiboult et al. 2017; Valdez et al. 2022). The inherent uncertainty of hydrological forecasts is driven by the stochastic nature of the atmosphere (Bauer et al. 2015), uncertainties in estimating hydrological responses to precipitation and evaporation, insufficient real-time data, and the impacts of a rapidly changing climate (Bauer et al. 2015; Hirabayashi et al. 2013). Ensemble methods leverage multiple realizations (e.g., scenarios of initial conditions or stochastic perturbations) to generate a range of possible outcomes, which, unlike single-value forecasts, offer a probabilistic view of future conditions (Das et al. 2022; Troin et al. 2021; Wu et al. 2020). Ensembles provide decision-makers with explicit forecast uncertainties, enabling them to assess risks and make informed decisions based on probabilities rather than single-value deterministic forecasts. The potential of ensemble forecasting lies in its ability to reflect the unpredictability of hydrometeorological systems and capturing a wider range of potential scenarios that may occur under different weather, climate, and environmental conditions (Bellier et al. 2018; Boucher et al. 2012; Thiboult et al. 2016).
The ongoing development and refinement of forecasting models through specialized testbeds—controlled experiments where new models, methods, and developments are tested and validated—further enhance the effectiveness of ensemble forecasting. Testbeds provide a level playing field for evaluating and tailoring forecasting techniques, ensuring that researchers can accurately quantify and communicate models’ relative strengths and weaknesses (Crochemore et al. 2020; Girons Lopez et al. 2021; Jean et al. 2023; Girons Lopez et al. 2025). For instance, among the various efforts to enhance forecast accuracy, postprocessing testbeds have been instrumental in refining hydrometeorological predictions by systematically correcting biases and quantifying uncertainties arising from model variability (Li et al. 2017). By continuously testing and adjusting these models against observed data, their ability to predict complex hydrological phenomena can be improved. This iterative process helps in fine-tuning the models so that they not only provide a range of possible outcomes but also improve in accuracy and reliability over time. The ultimate goal of these endeavors is to refine how ensemble forecasts can inform risk management practices, making them indispensable tools for decision-makers facing the challenges of managing water resources and related risks (floods, droughts) in an increasingly variable climate. The collaborative efforts in research and development within this field are essential for advancing our capabilities to forecast and effectively address hydrological uncertainties.
d. User-centric cocreation.
The incorporation of user-centric cocreation principles significantly enhances the utility and effectiveness of the forecasts; the user here is any stakeholder or decision-maker who actively engages with and relies on hydrological forecasting systems to make informed decisions. By actively involving users in the codesign, codevelopment, and coevaluation phases, developers can gain a deep understanding of the users’ specific needs, preferences, and constraints. This understanding is crucial to help embed science-based hydrological forecasting systems within the decision space of users, acknowledging the multiple factors and knowledge that are relevant to decision-making (Goddard 2016; Lemos et al. 2012). From the user perspective, cocreation enables a better understanding of the data, methods, and uncertainties underlying forecasting systems and contributes to building trust in the provided information. As a result, this approach leads to the creation of interfaces, visualizations, and products that are not only intuitive but also tailored to support the decision-making processes of the users (e.g., Jean et al. 2023; Zappa et al. 2014). Additionally, through participation in cocreation, local knowledge held by users is recognized and considered alongside the scientific knowledge embedded in these systems, contributing to credibility, salience, and legitimacy; three dimensions that contribute to uptake and use (Taylor and de Loë 2012). Such systems are central for providing clear and accessible information, enabling users to make informed decisions based on the data and information provided. Furthermore, the success of user-centric approaches in enhancing user satisfaction and the operational efficiency of forecasting systems has recently been demonstrated. These improvements are particularly valuable in critical situations, where timely and accurate information can improve the management of water resources and mitigate the impacts of extreme weather events (Cantone et al. 2023; Lienert et al. 2022; Matthews et al. 2023).
As interdisciplinary and intersectoral collaboration grows, the integration of social scientists into the hydrological forecasting process adds a key element to the modeling chain to bridge the gap between technical forecasters and users. The involvement of social scientists ensures that the forecasts are not only scientifically robust but also socially relevant, actionable, and tangible. For instance, social scientists contribute by integrating vulnerability in risk assessments, translating physical variables into value-oriented metrics, or interpreting how forecast data might be perceived and used by different communities, thus aiding in the customization of forecasts to meet diverse societal needs (Crochemore et al. 2024; Ramos et al. 2013). This collaboration can enhance the relevance and usability of hydrological data, improving community preparedness and response strategies, including early actions as part of end-to-end early warning systems (EWSs). This synergy between technical accuracy and social relevance in hydrological forecasting helps ensure that the forecasts serve their intended purpose, supporting effective decision-making and ultimately contributing to the resilience and safety of communities affected by water-related challenges.
e. Education and capacity development.
Education and capacity development play a pivotal role in ensuring that forecasting systems and tools are being appropriately used, and that forecast information is not only disseminated but also effectively understood and used by the targeted audience. Delivering forecast information in a manner that is accessible, timely, and comprehensible to stakeholders is of high importance (Crochemore et al. 2016; Dasgupta et al. 2023). To achieve this, developing communication strategies that are tailored to the local context and can be aided by cocreation is critical. Communication strategies should resonate with the specific experiences and expectations of the local and regional communities and first responders, thereby enhancing the likelihood of an effective response to hydrological warnings and advisories (Zappa et al. 2013). Tailored strategies help bridge the gap between scientific data and practical applications, ensuring that forecasts serve their ultimate purpose of safeguarding lives, properties, and livelihoods against water-related hazards.
Educating users, first responders, and decision-makers about the quality and characteristics of hydrological forecasts and inherent uncertainties is crucial for cultivating a knowledgeable community capable of effectively addressing challenges posed by water-related risks. Training stakeholders on accurately interpreting forecasts, understanding their limitations, and grasping the implications of various scenarios is vital for building resilience, as it empowers users to make informed decisions based on a nuanced understanding of forecast implications and associated uncertainties (Crochemore et al. 2021; Van Loon et al. 2020). Serious games, inspired by scenarios of natural hazards such as flood control, drought adaptation, and water management, have been developed to facilitate decision-making (Arnal et al. 2016; Terti et al. 2019). While these games provide simplified representations of reality and do not aim to replicate the full complexity of operational environments, they serve as effective support materials in teaching and training activities for diverse target groups. Additionally, the implementation of robust feedback mechanisms from users is indispensable for the continual refinement and evolution of forecasting systems. User feedback loops ensure that the systems evolve to remain relevant and user-centered, adapting over time to meet changing needs and incorporating new insights gained from user experiences. The ongoing dialogue between forecast developers and users enhances the reliability of the forecasting systems and reinforces the community’s capacity to effectively use these tools in planning and response efforts.
3. Essential scientific directions for enhancing hydrological forecasting systems
The advancement of hydrological forecasting systems relies on a multifaceted scientific approach that encompasses several critical research domains. By exploring core methods and integrating interdisciplinary insights, here, we explore the scientific directions essential for evolving current hydrological forecasting capabilities to meet forecasting (systems) challenges. Each scientific direction discussed below represents a strategic area of focus, contributing to the top priorities for refining forecasting technologies and methods in hydrology; these are conceptually presented in Fig. 2.
Scientific disciplines and fields identified for their contributions to the scientific priorities of (co)-creating R2O hydrological forecasting systems.
Citation: Bulletin of the American Meteorological Society 106, 5; 10.1175/BAMS-D-24-0322.1
a. Standardized evaluation frameworks.
The development of standardized and comprehensible evaluation frameworks is essential for quantifying the value and inclusivity of hydrological forecasting systems and allowing their intercomparability. These frameworks are designed to measure accuracy, reliability, use/utility, and societal impact, ensuring that the forecasting systems are not only technically proficient but also broadly accessible and equitable (Alfieri et al. 2014; Arnal et al. 2018; Das et al. 2022; Kim et al. 2019; Pappenberger et al. 2015). By adopting a multidimensional approach, we can assess how well these systems perform across different environments and communities, comparing them against a benchmark to identify areas where improvements are necessary (Du et al. 2023; Pechlivanidis et al. 2020). Such an approach involves rigorous testing under diverse conditions and temporal scales to ensure that forecasts are robust and reliable across the geographic or societal contexts in which they are applied (testbeds in which testing happens). We note here that diagnostic evaluation frameworks should assess both meteorological inputs and hydrological outputs across regions and seasons, using diverse verification metrics that capture key forecast attributes (Huang and Zhao 2022). Moreover, the focus on equity within these frameworks is crucial for addressing the diverse needs of global communities. This includes developing user-driven metrics that evaluate the effectiveness and usability of forecasts in serving populations vulnerable to hydrological disasters and water managers of the water–food–energy nexus (Boucher et al. 2012; Crochemore et al. 2024; Giuliani et al. 2020). The effective implementation of standardized evaluation frameworks is heavily reliant on the availability of open and shareable data, which underscores the urgency for advancing monitoring and observational techniques. Given that many ground observation systems are currently deteriorating, enhancing these systems—coupled with fostering improved data sharing practices—is crucial. These improvements not only lead to superior data products, such as more accurate precipitation inputs, but also promote the intercomparability of forecasting systems, a fundamental requirement for the robust evaluation of their effectiveness across space and time (Demirel et al. 2018; Tauro et al. 2018). By integrating perspectives from social sciences and economics, the goal is to bridge the gap between quantifying forecast quality and valuing forecast utility, and to create a standardized yet flexible evaluation tool that can adapt to new scientific insights and societal changes, thereby supporting the continuous improvement of forecasting technologies (Cassagnole et al. 2021; Matte et al. 2017; Thiboult et al. 2017).
b. Advanced computational science.
Incorporating advanced computational sciences, utilizing, for example, AI/ML methods, into hydrological forecasting has the potential to transform the prediction and management of water-related risks. These methods and tools allow for the development of models or computationally efficient model emulators that can process vast amounts of data and ensembles, learn from them, and make predictions with comparable or even greater accuracy than current state-of-the-art systems. For instance, AI/ML algorithms can detect subtle patterns in historical data that traditional models might overlook, leading to a better exploitation of our observations, and more accurate predictions of extreme events and their impacts. Moreover, an evaluation of the ability of large-sample ML models to predict extreme events revealed good predictive accuracy compared with conceptual and process-based approaches (Frame et al. 2022). However, we note that the availability of adequate data for AI/ML applications is highly region-dependent (Bône et al. 2023; Vitanza et al. 2023). The challenge lies in integrating data-informed models with traditional hydrological understanding to ensure that forecasts remain grounded in physical reality and testing their approach on unseen datasets (spatiotemporal validation). This integration is embodied in so-called “hybrid” models that combine the strengths of process-/physical-based and data-informed approaches combine the detailed mechanistic understanding of hydrological processes provided by traditional models with the predictive power and computational efficiency of AI/ML algorithms (Kraft et al. 2022; Liu et al. 2024). Machine learning–based data assimilation techniques is another possible avenue to integrate data-driven techniques into hydrological forecasting systems that remain centered on a physics-based or conceptual modeling approach (Boucher et al. 2020). Novel AI-enhanced postprocessing frameworks have the potential to tailor forecasts of global hydrological models to local conditions in gauged and ungauged areas. Another scientific direction involves AI-powered large language models that are designed to sift through vast information, including written warnings and forecast data, which may enable expert responses, enrich research, and communicate complex climate information to local users in an understandable format. By developing and refining these hybrid models, scientists can provide more reliable and actionable forecasts, tailored to the specific context and needs of users in various regions. Special focus should also be given to the transferability of data-driven modeling approaches to climate extremes and other regions with (yet) insufficient training data.
c. Science communication.
Effective science communication is key to ensuring that hydrological forecasts can be translated into meaningful and tangible actions (Demeritt et al. 2010; Ramos et al. 2010). The communication effort involves delivering data or conveying the scientific basis of forecasts, as well as contextualizing the information to be relevant to local communities. Codeveloping communication methods like narratives and storylines, which offer structured and relatable descriptions, enhances the clarity and comprehension of uncertainties associated with probabilistic forecasts (e.g., identifying sources of uncertainty and understanding how it propagates). This approach makes the information more accessible and interpretable for both first responders and nonexperts (Chan et al. 2024; Jean et al. 2023, 2024; Rangecroft et al. 2018). Narratives can illustrate potential scenarios in a way that resonates with local experiences, enhancing public understanding and engagement. Additionally, integrating social science methods is critical for tailoring communication strategies to diverse audiences. These methods involve assessing and better understanding societal impacts and behavioral responses to forecasts, thereby informing communication strategies, including visualization, and dissemination means, to effectively motivate community preparedness and response (Botzen et al. 2009; Di Baldassarre et al. 2019). Benchmarking different communication techniques across various scenarios allows identifying the most effective methods for engaging with different communities, ensuring that forecasts lead to informed decision-making and contribute to reducing the risks associated with hydrological extremes.
d. Impact analysis and inclusive forecasting.
Linking hydrological forecasts to societal impacts demonstrates their practical value. Impact analysis examines how forecasts influence real-world outcomes across economic, environmental, and social dimensions (Cassagnole et al. 2021; Cloke et al. 2017). It is essential to develop methods that clearly demonstrate the real-world benefits of accurate hydrological forecasting, such as reduced flood damage or enhanced transboundary water resource cooperation (De Angeli et al. 2024; Merz et al. 2020; Shyrokaya et al. 2024a). Evaluating the value of impact forecasts presents a notable challenge, as they are dependent on complex and variable human decision-making in response to hydrological forecasts, complicating the consistency and reproducibility needed for traditional forecast evaluation methods (Boucher et al. 2012; Giuliani et al. 2020; Thiboult et al. 2017). This requires a deep understanding of the various ways in which different communities interact with and rely on hydrological systems. Moreover, the intersection of forecasting science and gender studies, including demographic research, ensures that forecasts address the needs of all community segments (Bailie et al. 2022; Dube and Mhembwe 2019; Lechowska 2022; Mustafa et al. 2015). Understanding the differential impacts of hydrological events on various socioeconomic groups is crucial and highlights the importance of inclusive communication and resource allocation to overcome water-related (even occasionally violent) conflicts, for example, ensuring that emergency response plans consider the specific needs of pregnant women, the elderly, or those with disabilities. Additionally, it is vital to communicate in all the languages spoken by the community members, ensuring that critical information is accessible to everyone, including those who may not speak the dominant language. This integration promotes equitable long-term resilience toward water-related disasters by considering the specific vulnerabilities and needs of diverse populations, helping NGOs and disaster managers develop tailored multilevel plans for each stage of the disaster management cycle.
e. Local implementation and policy science.
To ensure the effectiveness of hydrological forecasts, it is necessary to develop frameworks that support their integration into the decision-making processes of local and regional governance, policymaking bodies, emergency services, and water management agencies (Bakhtiari et al. 2024; Cantone et al. 2023; Samaniego et al. 2019). By aligning (scientific) predictions with policy needs, forecasts can directly inform strategies for disaster preparedness, resource allocation, and environmental protection. Additionally, demonstrating the economic value of these forecasting initiatives, compared to existing benchmark systems, is crucial for securing ongoing support and funding (Pappenberger et al. 2015). We still need to develop methods to quantify the economic or societal benefits of improved impact-based forecasting, such as cost savings or damages/casualties avoided from disaster mitigation and enhanced water resource management (Thiboult et al. 2017; Shyrokaya et al. 2024b). The development of such methods is, however, hampered by the lack of accessible databases regarding damages. Damage data are usually confidential for private properties, while it is also tricky to determine what should ideally be included in a damage model: tangible assets such as public infrastructure and temporary shelters as well as a vast array of intangible damages (stress, health issues, etc.) that are difficult to quantify. Therefore, providing compelling evidence to policymakers and funders about the importance of investing in advanced impact-based forecasting technologies is essential (Arnal et al. 2016; Bruno Soares et al. 2018; Girons Lopez et al. 2021; Teague et al. 2021). By establishing logical links between R2O forecasting and economic and societal outcomes, the effort will not only support policy formulation but also ensure continuous investments in forecasting.
f. Behavioral science and crisis management.
Integrating insights from behavioral science and crisis management into hydrological forecasting is crucial for enhancing individual and collective responses to forecasts and warnings. Behavioral science, including behavioral economics and cognitive psychology, examines how biases, heuristics, imagination, and emotional responses influence decision-making under stress and uncertainty (Almeida and Curado 2019; Dionne et al. 2018). Crisis management examines psychological and cognitive processes in high-pressure situations with incomplete or multiple sources of information. Cognitive biases significantly affect decisions in complex situations, when uncertainty is high, when events and information are probabilistic, and when people are under stress (Merz et al. 2015). Sociohydrology complements these insights by examining the dynamic interplay between society and hydrological systems, emphasizing the need for forecasting models that reflect both human influence and natural processes (Di Baldassarre et al. 2019; Vanelli et al. 2022). This interdisciplinary approach is championed by the International Association of Hydrological Sciences (IAHS), which advocates for integrating human and ecological aspects in water management to enhance forecasting accuracy and utility (Ceola et al. 2016). These disciplines are crucial for hydrological forecasting because both individual and collective (organizational) decision-makers and forecasters are often subject to cognitive biases, such as availability bias or wishful thinking, which can distort the interpretation of probabilistic information, e.g., representativeness heuristic (AlKhars et al. 2019). For instance, the tendency to rely on readily available memories can lead to underestimating the severity of unlikely but possible flood forecasts (Ommer et al. 2024). Additionally, the emotional impact of past crises can either enhance vigilance or hinder the ability to imagine new threats (Bastardi et al. 2011). By developing tools and frameworks that consider the human factors influencing judgments and choices, forecasting systems can be tailored to better meet the diverse needs and decision-making processes at the individual, collective, research, operational, and public levels, thereby improving overall decision-making.
4. Advancing R2O in hydrological forecasting—the contribution of HEPEX
The HEPEX CoP (HEPEX 2024) has played a pivotal role in transforming hydrological forecasting into a dynamic field that seamlessly blends research with operational applications. Various strategic innovations and collaborative efforts have broadened the scope and impact of forecasting, enabling more effective management of water resources and disaster response strategies (Schaake et al. 2007). Through its dedicated initiatives, HEPEX has not only advanced the science behind hydrological forecasting but also enhanced the practical utility and accessibility of these forecasts to a global audience (see Fig. 3).
HEPEX key impacts on hydrological forecasting toward bridging the capacity, needs and progress of research, services, and decision-making. The empty hexagons symbolize other important roles than those in advancing the R2O continuum in hydrological forecasting.
Citation: Bulletin of the American Meteorological Society 106, 5; 10.1175/BAMS-D-24-0322.1
a. Advancement in ensemble forecasting.
HEPEX has been pivotal in transforming ensemble forecasting from a novel concept to a standard practice within the hydrological forecasting community (Das et al. 2022; Thielen et al. 2009). Ongoing commitment to improving ensemble forecasting reflects HEPEX’s core mission to bridge the gap between cutting-edge research and practical application, ensuring that the latest scientific advancements are accessible and beneficial to a broad range of users. Initially perceived as a complex and unconventional approach, ensemble forecasting has become widely recognized for its ability to enhance forecast reliability and address uncertainty. This approach offers decision-makers a probabilistic view that better captures the inherent uncertainties (i.e., initial conditions, model structure and parameters, meteorological forcing) of hydrological events.
The widespread adoption (Pappenberger et al. 2019) of ensemble methods in routine practice reflects HEPEX’s advocacy and demonstration of their value through numerous workshops, publications, and collaborations. This movement has advanced the scientific basis of ensemble forecasting and helped its integration into operational practices, improving the overall robustness and confidence in hydrological predictions across all geographical domains (Girons Lopez et al. 2021; Harrigan et al. 2018). HEPEX’s role in advancing ensemble forecasting extends to the development of testbeds (https://hepex.org.au/test-beds-in-hepex-lessons-learned-and-way-forward) and real-time forecasting systems that utilize these methods (Arnal et al. 2020). These platforms allow for the practical application and continuous refinement of ensemble techniques, ensuring they are adapted to the evolving needs of the forecasting and the stakeholder community. Enhancements in forecast accuracy and utility have been reached by providing venues for rigorous testing and validation. Moreover, the testbeds are also useful in establishing trust in a new modeling approach, since the benchmarking exercise includes popularly used models that many users know already. As the forecasting community embraces the potential of data-driven AI models, the contributions of the HEPEX testbeds will be pivotal in establishing trust and credibility of the systems (e.g., Nearing et al. 2024).
b. Decision-making based on probabilistic forecasts.
The use of probabilistic forecasts to inform decision-making in water resource management and disaster response is at the heart of HEPEX activities. By providing a spectrum of scenarios, probabilistic forecasts mark a significant shift from traditional deterministic decision-making processes enabling decision-makers to consider various potential futures and plan accordingly, enhancing preparedness and response strategies (Crochemore et al. 2021; Lienert et al. 2022). HEPEX’s advocacy for this method has led to its broader acceptance and implementation across various sectors, improving the overall effectiveness of hydrological risk management (Bennett et al. 2017; Hao et al. 2018; Schaake et al. 2007; Wu et al. 2020). HEPEX’s efforts extend beyond promoting probabilistic forecasts to actively involving the user community in the development and refinement of these tools. Through workshops, training sessions, and collaborative projects, HEPEX engages with stakeholders to tailor forecasting methods to their specific needs and contexts. This bottom-up approach ensures that the forecasts are not only scientifically robust but also practically applicable, facilitating their integration into policymaking and operational planning.
c. Formation of a collaborative global community of practice.
HEPEX has successfully cultivated a vibrant and collaborative international community, bringing together experts from academia, government, and industry at various career stages to foster a collective approach to hydrological forecasting (HEPEX 2024). Moreover, this global network has been instrumental in pooling resources, sharing knowledge, and coordinating efforts to tackle universal challenges in hydrological ensemble forecasting. The strength of this community lies in its diversity, encompassing a wide range of perspectives and expertise that enriches the dialogue and innovation within the field. Regular conferences, workshops, and collaborative projects (https://hepex.org.au/hepex-workshops/) under the HEPEX umbrella have not only facilitated knowledge exchange but also fostered long-term partnerships that transcend geographical and disciplinary boundaries. The impact of this international community is evident in the accelerated dissemination of best practices and cutting-edge research findings across the globe. By acting as a central hub for hydrological forecasting expertise, HEPEX ensures that valuable insights and innovative techniques are shared widely, preventing the duplication of efforts and promoting the efficient use of scientific resources. Furthermore, the organization’s commitment to inclusivity and collaboration is fundamental to its mission, ensuring that advancements in hydrological forecasting are leveraged to benefit communities worldwide, enhanced by workshops that rotate globally to engage a diverse international audience.
d. Bridging the gap between research and practice.
HEPEX is dedicated to narrowing the gap between hydrological research and operational practice, hence not only enhancing the effectiveness of forecasting systems but also ensuring these remain adaptable and forward-looking, capable of meeting the challenges posed by a changing climate and evolving sociotechnological landscape. By fostering a dynamic environment where scientists and practitioners codevelop and test forecasting methods, HEPEX ensures that theoretical advancements are grounded in real-world applicability, enhancing the capacity of operational frameworks to adopt new findings and tools. This collaborative approach not only speeds up the translation of research into practice but also ensures that the developed solutions are robust, user-friendly, and tailored to the specific needs of the users and decision-makers. HEPEX’s testbeds (https://hepex.org.au/category/testbed/) provide practical platforms for experimenting with new theories and technologies in controlled, real-world settings, enabling continuous feedback and refinement, as well as trust-building in newer modeling systems. HEPEX’s serious games (https://hepex.org.au/resources/hepex-games/) provide hands-on and interactive ways of interacting with forecast users about forecast uncertainties, performance, or visualization in controlled and fictitious environments. Moreover, HEPEX actively promotes the adoption of best practices through its extensive network, facilitating the exchange of ideas and experiences that enhance the practical knowledge of both researchers and practitioners. By doing so, HEPEX plays a crucial role in creating a more integrated and responsive hydrological community that is better equipped to handle the complexities and variabilities of water-related phenomena.
e. Promotion of accessibility, transparency, and inclusiveness.
HEPEX is committed to enhancing accessibility, transparency, and inclusiveness in the field of hydrological forecasting. This enriches the scientific and operational aspects of forecasting and also strengthens community resilience by ensuring that everyone has the knowledge (open science) and tools (open source) necessary to respond to hydrological risks (Hall et al. 2022). This commitment is manifested in various initiatives aimed at generating data, creating materials, games, and tools available to a broader audience, including policymakers, community leaders, and the general public. The HEPEX Handbook (Duan et al. 2019), for instance, serves as a comprehensive resource that outlines methods, case studies, and best practices in a format that is accessible to students and nonexperts. This effort not only democratizes information but also empowers more stakeholders to engage with and utilize hydrological data effectively. HEPEX actively promotes inclusiveness by encouraging the participation of diverse groups in the forecasting process. This includes integrating perspectives from all ranges of the career stage (from early career, e.g., through EC-HEPEX, to senior scientists), underrepresented communities and ensuring that forecasting tools and information are sensitive to the cultural and social contexts of different regions. By fostering a more inclusive approach, HEPEX enhances the relevance and impact of hydrological forecasts, ensuring that they serve the needs of all community segments.
5. Toward supporting the UN EW4All initiative
The UN EW4All initiative is structured around four pillars designed to enhance global disaster resilience: pillar 1—disaster risk knowledge—focusing on improving understanding of disaster risks to inform early warnings [led by the United Nations Office for Disaster Risk Reduction (UNDRR)]; pillar 2—observations, monitoring, analysis, and forecasting—enhancing monitoring and data capabilities for accurate forecasting of hazards [led by the World Meteorological Organization (WMO)]; pillar 3—warning dissemination and communication—ensuring that warnings are timely and reach those at risk effectively [led by the International Telecommunication Union (ITU)]; and pillar 4—preparedness and response capabilities [led by the International Federation of Red Cross and Red Crescent Societies (IFRC)]—building capacities at all societal levels for emergency response and readiness. Through a combination of training, technology codevelopment and sharing, interdisciplinary collaboration, and enhanced global communication, HEPEX has contributed to building robust EWS designed to improve disaster preparedness and response capabilities worldwide. Here, we present how HEPEX’s approaches and broad network can support the development and implementation of each EW4All pillar (Fig. 4).
Identified HEPEX CoP fields of contribution to the four pillars of the UN EW4All initiative.
Citation: Bulletin of the American Meteorological Society 106, 5; 10.1175/BAMS-D-24-0322.1
a. Localized training and global collaboration.
Contributions to EW4All pillar 1—HEPEX’s emphasis on localized training and global collaboration is instrumental in enhancing the effectiveness of EWS as part of the EW4All initiative. By fostering a training framework that combines local presence with global insights, HEPEX ensures that learning and development opportunities are not only globally informed but also regionally adapted. Such training takes place at each workshop with invited stakeholders and dedicated sessions on local challenges and knowledge transfer. This approach allows for the incorporation of local expertise and traditional knowledge, which are critical in tailoring EWS to the specific environmental and cultural contexts of diverse regions. The collaborative training programs engage participants from around the world, creating a rich exchange of ideas and practices that enhance the capacity of local communities to manage disaster risks effectively.
By cocreating these training modules with participants from various backgrounds, HEPEX highlights and leverages existing resources while simultaneously developing new tools and methods. This cocreation process ensures that the training materials are continually updated and relevant, addressing the evolving challenges faced by different regions. Such initiatives not only empower local communities but also foster a sense of ownership and accountability toward the maintenance and improvement of EWS.
b. Promotion of open-source sharing and community engagement.
Contributions to EW4All pillars 2 and 3—HEPEX facilitates collective knowledge building across the hydrological and meteorological communities and promotes open-source sharing and community engagement by creating a platform where hydrological models, data, best practices, and case studies are openly shared. This open exchange not only democratizes access to advanced forecasting tools but also encourages a collaborative approach to EWS development, ensuring that these systems are robust, scalable, and adaptable to different environmental conditions.
The engagement facilitated by HEPEX extends beyond traditional academic and professional boundaries to include local practitioners, policymakers, and other stakeholders. This inclusive approach ensures that the development of EWS is a participatory process, reflecting the needs and inputs of diverse groups, including those often underrepresented in scientific and technical discussions. By simulating different scenarios during workshops and training sessions, HEPEX helps stakeholders better understand and prepare for a range of potential disasters, thereby enhancing the overall resilience of communities to adverse and unprecedented weather events and other hydrometeorological hazards.
c. Focus on interdisciplinary initiatives and communication.
Contributions to EW4All pillar 3—commitment to interdisciplinary initiatives and effective communication plays a crucial role in enhancing the comprehensiveness and equity of the EWS. By integrating knowledge from various disciplines and focusing on trans -sectoral collaboration, HEPEX ensures that EWSs are not only scientifically sound but also socially relevant and widely accessible. This approach addresses the need for EWSs that can communicate risk effectively across different cultural and social landscapes, particularly focusing on vulnerable groups, whose needs and voices might otherwise be overlooked.
Through initiatives that enhance dialogues between scientists, forecasters, users, and policymakers, HEPEX fosters a deeper understanding of how forecasts are perceived and acted upon by different communities. The focus on social sciences methods such as disaster economics, behavioral and social psychology, and gender studies enriches the technical aspects of forecasting with insights into human behavior and communication strategies, making EWSs more user-friendly and actionable. The result is a more equitable dissemination of critical information that empowers all segments of the population (including vulnerable ones) to take timely and appropriate actions in response to potential threats.
d. Expansion of global representation and online initiatives.
Contributions to EW4All Pillar 4—HEPEX’s efforts to expand global representation and enhance online initiatives significantly contribute to the accessibility and diversity of EWS. By actively seeking to grow its membership (currently >650 registered members), particularly from countries in the Global South, HEPEX can ensure that the perspectives and experiences of these regions are integrated into the ongoing developments and refinements of state-of-the-art EWS. This focus on diversity not only enriches the pool of knowledge and strategies but also ensures that the systems developed are scalable and applicable across different geographical and socioeconomic contexts. Moreover, the growth of online platforms and resources, such as free workshops, and the practice of plenary invited talks by early career individuals extend the reach of HEPEX’s training and development initiatives, allowing for greater participation and engagement from around the world. These initiatives are especially important in regions such as Africa, South America, and Asia, where access to traditional educational and professional development opportunities may be limited.
By additionally linking with international programs like UNESCO’s International Hydrological Program (IHP), the IAHS, and WMO’s hydrology-related initiatives within the World Weather Research Program (WWRP) and World Climate Research Programme (WCRP)—including projects like GEWEX, the S2S project, InPRHA, and GPEX—HEPEX facilitates the broader dissemination of knowledge and best practices. These collaborations significantly contribute to the global effort to boost preparedness and response capabilities as outlined in the EW4All initiative. Furthermore, HEPEX’s involvement with the broader hydrometeorological and climatological research community, evident through initiatives like the European Commission’s Destination Earth program and the WMO’s integration of hydrology with WWRP and WCRP projects, demonstrates the increasing convergence of disciplines (e.g., high-resolution weather and climate modeling) that enriches our understanding and effectiveness in managing hydrological extremes.
e. Technical support and networking.
Contributions to all four EW4All Pillars—drawing from the CoP extensive experience with users and warning chains, HEPEX offers expert advice and support that enhance the technical robustness and operational efficiency of EWSs, while promoting, through best practices, the adoption of state-of-the-art methods and tools. The HEPEX website serves as a vital hub for information exchange, offering updates, resources, and networking opportunities that keep the community informed and connected. This platform is instrumental in promoting the EW4All initiative, leveraging existing partnerships to enhance the reach and impact of EWS.
Additionally, HEPEX’s role in training the next generation of hydrological forecasters and facilitating connections between various stakeholders helps to foster a collaborative environment, where knowledge and resources are shared freely. This networking capability is essential for maintaining the relevance and effectiveness of EWS, as it allows for continuous feedback and adaptation based on the latest scientific research and practical experiences from around the globe. Through these efforts, HEPEX can support the technical development of EWSs and ensure that these systems are integrated seamlessly into broader disaster risk management and response strategies.
6. Conclusions
In the past two decades, HEPEX has led efforts to advance hydrological ensemble prediction and its integration into risk-based decision-making processes. HEPEX has promoted the use of state-of-the-art techniques, models, and data to refine forecasting methods and enhance user services in water-related sectors. Here, we outline the community’s strategic innovations for enhancing hydrological forecasting systems that are effective and applicable across various spatial and temporal scales. The messages concluded here are based on the collection, analysis, and prioritization of the responses from five breakout groups (about 70 international hydrology experts, consisting of both practitioners and researchers) participating at the HEPEX 2023 workshop (13–15 September 2023, Norrköping, Sweden).
Overall, efforts for a robust and usable operational hydrological forecasting science underscore the importance of a collaborative, inter- and transdisciplinary approach to creating hydrological forecast systems that are accurate, relevant, and user-friendly across different spatial scales and time horizons. The review has highlighted several pivotal advancements and ongoing challenges in the field of hydrological forecasting, particularly in bridging the R2O gaps. The integration of advanced scientific methods, including AI and ML, alongside high-resolution weather and climate modeling and data assimilation techniques, has significantly propelled our predictive capabilities across various spatial scales and time horizons. These technological advancements have proven crucial in enhancing the accuracy and reliability of forecasts, thereby improving decision-making processes in water resource and emergency management under varying hydroclimatic conditions.
Moreover, we have underscored the critical role of data quality and integration in the development of robust hydrological models. Ensuring the integrity and usability of data through collection, sharing, curation, and assimilation processes is paramount. The integration of diverse data from conventional and nonconventional sources, including, for instance, Earth observations, private weather stations and ground-based measurements, enriches the data information content, allowing for a comprehensive representation of hydrological systems. This is crucial for maintaining the consistency and accuracy of hydrological forecasts, which are integral to managing the increasing frequency and severity of water-related hazards.
Our findings also emphasize the importance of user-centric focus in the cocreation of action-based forecasting systems. Engaging users in the design, development, and evaluation phases not only ensures that the systems are aligned with their needs and constraints but also enhances their usability and adoption. The integration of social scientists into the forecasting process further ensures that the services and early warning systems are not only technically robust but also socially relevant and actionable.
Education and capacity development emerge as crucial elements for the effective dissemination and utilization of forecast information. Developing communication strategies that resonate with the local context and user experiences significantly enhances the responsiveness to hydrological advisories and warnings. Moreover, continuous education conditioned toward forecast interpretation and the implications of various scenarios builds a well-informed community capable of making prompt and actionable decisions under uncertainty.
It is imperative to continue refining early warning systems through ongoing research and collaboration, supporting the UN EW4All targets. Addressing the scalability challenges and enhancing the operational frameworks to accommodate rapid technological advancements will be crucial, while expanding the scope to include more underrepresented regions and communities will ensure a more equitable distribution of the benefits of advanced hydrological forecasting. HEPEX has been significantly contributing to this initiative by enhancing global disaster resilience through robust training, collaborative projects, and open-source sharing, aligned with the EW4All pillars. These efforts improve disaster risk knowledge, enhance forecasting accuracy, ensure effective warning dissemination, and build capacities for emergency response. Collaboration with international partners such as UNESCO’s IHP, IAHS, and WMO’s hydrology-related initiatives extends HEPEX’s impact, promoting inclusivity and ensuring that advances in hydrological forecasting benefit all, particularly in vulnerable regions. Moreover, HEPEX’s interdisciplinary approach highlights the essential synergy between disciplines for managing hydrological extremes effectively. By promoting collaboration across various scientific fields, HEPEX not only enhances community resilience and safety against extremes but also drives the continual advancement of forecasting systems, ensuring they are adaptable to new scientific insights and responsive to evolving societal needs.
We finally note that while significant progress has been made in hydrological forecasting, considerable work remains. To further enhance hydrological forecasting systems, several key scientific directions were identified and proposed, ensuring that hydrological forecasting tools evolve to meet the dynamic needs of global water-related challenges:
The development of standardized evaluation frameworks is crucial for assessing the inclusivity and value of hydrological forecasting systems. These frameworks are essential for ensuring that the systems are not only technically sound but also broadly accessible and equitable. They rely on a multidimensional approach to performance assessment across various environments and require open and shareable data to enhance monitoring techniques and promote system intercomparability.
Incorporating advanced computational sciences, such as AI and ML, is transforming the prediction and management of water-related risks. These technologies enable the development of models that process vast amounts of data and uncover patterns that traditional models might miss, thus improving predictions of extreme events. However, integrating these data-driven models with traditional hydrological approaches necessitates careful validation to ensure their grounding in physical reality.
Effective science communication is vital for translating hydrological forecasts into actionable strategies. Developing communication methods like narratives and storylines clarifies the uncertainties of probabilistic forecasts, making them accessible to both experts and the general public. This approach not only enhances public engagement but also ensures that forecasts are effectively understood and utilized across different cultural and social contexts.
Linking forecasts to their societal impacts underscores their practical value. This necessitates clear methods to demonstrate the actual benefits to the society and users. The complex decision-making processes involved in this aspect of impact analysis highlight the importance of inclusive forecasting considering diverse community needs to ensure equitable resilience against water-related hazards.
Lastly, effective forecasting systems require frameworks that integrate scientific predictions with policy and decision-making processes at local and regional levels. This alignment is crucial for informed disaster preparedness and resource allocation, necessitating continuous engagement with policymakers to demonstrate the economic and societal benefits of advanced forecasting technologies.
Acknowledgments.
We would like to deeply thank all organizers and participants of the HEPEX 2023 workshop (13–15 September 2023, Norrköping, Sweden) on “Forecasting across spatial scales and time horizons”; they not only shared their excellent scientific and operational challenges, solutions, and insights but also set the ground with their engagement for the future path of the HEPEX initiative. Thanks also go to the Swedish Energy Research Centre, Energiforsk (https://energiforsk.se/), that sponsored the HEPEX 2023 workshop. We would further like to thank all members of the Task Team on Hydrology Research (https://community.wmo.int/en/task-team-hydrology-research) operating under the Research Board of the World Meteorological Organization for reviewing and providing comments on the first version of the manuscript. I. G. P., J. L., and M.-H. R. received financial support from the EU Horizon Europe project Mediterranean and pan-European forecast and Early Warning System against natural hazards (MedEWSa) under Grant Agreement 101121192. Y. D. received financial support from the EU Horizon 2020 project Climate Intelligence: Extreme events detection, attribution, and adaptation design using machine learning (CLINT) under Grant Agreement 101003876. J. B. received travel support to the workshop from the EU Horizon 2020 project CLINT under Grant Agreement 101003876 and the research partnership between CSIRO and Hydro Tasmania. G. D. B. received the financial support from the EU Horizon Europe project Transformational and Robust Adaptation to Water Scarcity and Climate Change under Deep Uncertainty (TRANSCEND) under Grant Agreement 101084110 and from the EU Horizon 2020 project Innovating Climate services through Integrating Scientific and local Knowledge (I-CISK) under Grant Agreement 101037293. M. W. received financial support from the EU Horizon 2020 project I-CISK under Grant Agreement 101037293. The authors of this work declare no conflict of interest. Our funders had no role in the choice of research project; design of the study; in the collection, analyses, or interpretation of responses; in the writing of the manuscript; or in the decision to publish the insights. I. G. P., M.-A. B., F. W., and J. B. designed the study and led the workshop. I. G. P. and Y. D. led the writing of the manuscript, with all coauthors contributing in the interpretation of the responses and commenting on the manuscript.
Data availability statement.
The insights discussed in this study were derived by systematically reviewing, synthesizing, and contextualizing the participants’ responses, whose expertise and results can be found in the program of the workshop at https://hepex.org.au/hepex-workshop-2023-forecasting-across-spatial-scales-and-time-horizons/ and https://hepex.org.au/a-look-to-the-past-hepex-workshop/.
APPENDIX Interaction with the HEPEX CoP Experts
Approximately 70 hydrology experts from across the globe attended the workshop, and their country of origin is shown in Fig. A1. The photos were taken in accordance with the rules of the General Data Protection Regulation (GDPR). Permission was requested from the participants prior to taking the photos, while the photos are available from the HEPEX website (https://hepex.org.au/).
(a) Map of the countries from the HEPEX on-site participants, (b) group picture with all the participants, (c) group activity focused on identifying the top priorities for (co)creating hydrological forecast systems, and (d) open discussion on the future of hydrological forecasting from the perspective of early career scientists (photos: SMHI).
Citation: Bulletin of the American Meteorological Society 106, 5; 10.1175/BAMS-D-24-0322.1
References
Alfieri, L., F. Pappenberger, F. Wetterhall, T. Haiden, D. Richardson, and P. Salamon, 2014: Evaluation of ensemble streamflow predictions in Europe. J. Hydrol., 517, 913–922, https://doi.org/10.1016/j.jhydrol.2014.06.035.
Alfieri, L., and Coauthors, 2022: High-resolution satellite products improve hydrological modeling in northern Italy. Hydrol. Earth Syst. Sci., 26, 3921–3939, https://doi.org/10.5194/hess-26-3921-2022.
AlKhars, M., N. Evangelopoulos, R. Pavur, and S. Kulkarni, 2019: Cognitive biases resulting from the representativeness heuristic in operations management: An experimental investigation. Psychol. Res. Behav. Manage., 12, 263–276, https://doi.org/10.2147/PRBM.S193092.
Almeida, F., and M. Curado, 2019: The role of observation, cognition, and imagination in Keynes’s approach to decision-making. EconomiA, 20, 15–26, https://doi.org/10.1016/j.econ.2019.03.001.
Arnal, L., M.-H. Ramos, E. Coughlan de Perez, H. L. Cloke, E. Stephens, F. Wetterhall, S. J. van Andel, and F. Pappenberger, 2016: Willingness-to-pay for a probabilistic flood forecast: A risk-based decision-making game. Hydrol. Earth Syst. Sci., 20, 3109–3128, https://doi.org/10.5194/hess-20-3109-2016.
Arnal, L., H. L. Cloke, E. Stephens, F. Wetterhall, C. Prudhomme, J. Neumann, B. Krzeminski, and F. Pappenberger, 2018: Skilful seasonal forecasts of streamflow over Europe? Hydrol. Earth Syst. Sci., 22, 2057–2072, https://doi.org/10.5194/hess-22-2057-2018.
Arnal, L., L. Anspoks, S. Manson, J. Neumann, T. Norton, E. Stephens, L. Wolfenden, and H. L. Cloke, 2020: “Are we talking just a bit of water out of bank? Or is it Armageddon?” Front line perspectives on transitioning to probabilistic fluvial flood forecasts in England. Geosci. Commun., 3, 203–232, https://doi.org/10.5194/gc-3-203-2020.
Bailie, J., V. Matthews, R. Bailie, M. Villeneuve, and J. Longman, 2022: Exposure to risk and experiences of river flooding for people with disability and carers in rural Australia: A cross-sectional survey. BMJ Open, 12, e056210, https://doi.org/10.1136/bmjopen-2021-056210.
Bakhtiari, V., F. Piadeh, A. S. Chen, and K. Behzadian, 2024: Stakeholder analysis in the application of cutting-edge digital visualisation technologies for urban flood risk management: A critical review. Expert Syst. Appl., 236, 121426, https://doi.org/10.1016/j.eswa.2023.121426.
Banholzer, S., J. Kossin, and S. Donner, 2014: The impact of climate change on natural disasters. Reducing Disaster: Early Warning Systems for Climate Change, A. Singh and Z. Zommers, Eds., Springer Netherlands, 21–49.
Basher, R., 2006: Global early warning systems for natural hazards: Systematic and people-centred. Philos. Trans. Roy. Soc., A364, 2167–2182, https://doi.org/10.1098/rsta.2006.1819.
Bastardi, A., E. L. Uhlmann, and L. Ross, 2011: Wishful thinking: Belief, desire, and the motivated evaluation of scientific evidence. Psychol. Sci., 22, 731–732, https://doi.org/10.1177/0956797611406447.
Bates, P. D., 2012: Integrating remote sensing data with flood inundation models: How far have we got? Hydrol. Process, 26, 2515–2521, https://doi.org/10.1002/hyp.9374.
Baudoin, M.-A., S. Henly-Shepard, N. Fernando, A. Sitati, and Z. Zommers, 2016: From top-down to “community-centric” approaches to early warning Systems: Exploring pathways to improve disaster risk reduction through community participation. Int. J. Disaster Risk Sci., 7, 163–174, https://doi.org/10.1007/s13753-016-0085-6.
Bauer, P., A. Thorpe, and G. Brunet, 2015: The quiet revolution of numerical weather prediction. Nature, 525, 47–55, https://doi.org/10.1038/nature14956.
Bellier, J., I. Zin, and G. Bontron, 2018: Generating coherent ensemble forecasts after hydrological postprocessing: Adaptations of ECC-based methods. Water Resour. Res., 54, 5741–5762, https://doi.org/10.1029/2018WR022601.
Bennett, J. C., Q. J. Wang, D. E. Robertson, A. Schepen, M. Li, and K. Michael, 2017: Assessment of an ensemble seasonal streamflow forecasting system for Australia. Hydrol. Earth Syst. Sci., 21, 6007–6030, https://doi.org/10.5194/hess-21-6007-2017.
Blöschl, G., and Coauthors, 2019: Twenty-three unsolved problems in hydrology (UPH)—A community perspective. Hydrol. Sci. J., 64, 1141–1158, https://doi.org/10.1080/02626667.2019.1620507.
Bône, C., G. Gastineau, S. Thiria, P. Gallinari, and C. Mejia, 2023: Detection and attribution of climate change using a neural network. J. Adv. Model. Earth Syst., 15, e2022MS003475, https://doi.org/10.1029/2022MS003475.
Botzen, W. J. W., J. C. J. H. Aerts, and J. C. J. M. van den Bergh, 2009: Dependence of flood risk perceptions on socioeconomic and objective risk factors. Water Resour. Res., 45, W10440, https://doi.org/10.1029/2009WR007743.
Boucher, M.-A., D. Tremblay, L. Delorme, L. Perreault, and F. Anctil, 2012: Hydro-economic assessment of hydrological forecasting systems. J. Hydrol., 416–417, 133–144, https://doi.org/10.1016/j.jhydrol.2011.11.042.
Boucher, M.-A., J. Quilty, and J. Adamowski, 2020: Data assimilation for streamflow forecasting using extreme learning machines and multilayer perceptrons. Water Resour. Res., 56, e2019WR026226, https://doi.org/10.1029/2019WR026226.
Bourgin, F., M. H. Ramos, G. Thirel, and V. Andréassian, 2014: Investigating the interactions between data assimilation and post-processing in hydrological ensemble forecasting. J. Hydrol., 519, 2775–2784, https://doi.org/10.1016/j.jhydrol.2014.07.054.
Brunner, M. I., L. Slater, L. M. Tallaksen, and M. Clark, 2021: Challenges in modeling and predicting floods and droughts: A review. Wiley Interdiscip. Rev.: Water, 8, e1520, https://doi.org/10.1002/wat2.1520.
Bruno Soares, M., M. Daly, and S. Dessai, 2018: Assessing the value of seasonal climate forecasts for decision-making. Wiley Interdiscip. Rev.: Climate Change, 9, e523, https://doi.org/10.1002/wcc.523.
Cantone, C., H. Ivars Grape, S. El Habash, and I. G. Pechlivanidis, 2023: A co-generation success story: Improving drinking water management through hydro-climate services. Climate Serv., 31, 100399, https://doi.org/10.1016/j.cliser.2023.100399.
Cassagnole, M., M.-H. Ramos, I. Zalachori, G. Thirel, R. Garçon, J. Gailhard, and T. Ouillon, 2021: Impact of the quality of hydrological forecasts on the management and revenue of hydroelectric reservoirs—A conceptual approach. Hydrol. Earth Syst. Sci., 25, 1033–1052, https://doi.org/10.5194/hess-25-1033-2021.
Ceola, S., and Coauthors, 2016: Adaptation of water resources systems to changing society and environment: A statement by the International Association of Hydrological Sciences. Hydrol. Sci. J., 61, 2803–2817, https://doi.org/10.1080/02626667.2016.1230674.
Chan, W. C. H., N. W. Arnell, G. Darch, K. Facer-Childs, T. G. Shepherd, and M. Tanguy, 2024: Added value of seasonal hindcasts to create UK hydrological drought storylines. Nat. Hazards Earth Syst. Sci., 24, 1065–1078, https://doi.org/10.5194/nhess-24-1065-2024.
Cloke, H. L., and F. Pappenberger, 2009: Ensemble flood forecasting: A review. J. Hydrol., 375, 613–626, https://doi.org/10.1016/j.jhydrol.2009.06.005.
Cloke, H. L., F. Pappenberger, P. J. Smith, and F. Wetterhall, 2017: How do I know if I’ve improved my continental scale flood early warning system? Environ. Res. Lett., 12, 044006, https://doi.org/10.1088/1748-9326/aa625a.
Coronese, M., F. Lamperti, K. Keller, F. Chiaromonte, and A. Roventini, 2019: Evidence for sharp increase in the economic damages of extreme natural disasters. Proc. Natl. Acad. Sci. USA, 116, 21 450–21 455, https://doi.org/10.1073/pnas.1907826116.
Crochemore, L., M.-H. Ramos, F. Pappenberger, S. J. van Andel, and A. W. Wood, 2016: An experiment on risk-based decision-making in water management using monthly probabilistic forecasts. Bull. Amer. Meteor. Soc., 97, 541–551, https://doi.org/10.1175/BAMS-D-14-00270.1.
Crochemore, L., M.-H. Ramos, and I. G. Pechlivanidis, 2020: Can Continental models convey useful seasonal hydrologic information at the catchment scale? Water Resour. Res., 56, e2019WR025700, https://doi.org/10.1029/2019WR025700.
Crochemore, L., C. Cantone, I. G. Pechlivanidis, and C. S. Photiadou, 2021: How does seasonal forecast performance influence decision-making? Insights from a serious game. Bull. Amer. Meteor. Soc., 102, E1682–E1699, https://doi.org/10.1175/BAMS-D-20-0169.1.
Crochemore, L., and Coauthors, 2024: A framework for joint verification and evaluation of seasonal climate services across socioeconomic sectors. Bull. Amer. Meteor. Soc., 105, E1218–E1236, https://doi.org/10.1175/BAMS-D-23-0026.1.
Das, J., V. Manikanta, K. Nikhil Teja, and N. V. Umamahesh, 2022: Two decades of ensemble flood forecasting: A state-of-the-art on past developments, present applications and future opportunities. Hydrol. Sci. J., 67, 477–493, https://doi.org/10.1080/02626667.2021.2023157.
Dasgupta, A., R. Hostache, R. Ramsankaran, S. Grimaldi, P. Matgen, M. Chini, V. R. N. Pauwels, and J. P. Walker, 2021a: Earth observation and hydraulic data assimilation for improved flood inundation forecasting. Earth Observation for Flood Applications, G. J.-P. Schumann, Ed., Earth Observation, Elsevier, 255–294.
Dasgupta, A., R. Hostache, R. Ramsankaran, G. J.-P. Schumann, S. Grimaldi, V. R. N. Pauwels, and J. P. Walker, 2021b: A mutual information-based likelihood function for particle filter flood extent assimilation. Water Resour. Res., 57, e2020WR027859, https://doi.org/10.1029/2020WR027859.
Dasgupta, A., S. Grimaldi, R. Ramsankaran, V. R. N. Pauwels, and J. P. Walker, 2022: A simple framework for calibrating hydraulic flood inundation models using crowd-sourced water levels. J. Hydrol., 614, 128467, https://doi.org/10.1016/j.jhydrol.2022.128467.
Dasgupta, A., and Coauthors, 2023: Connecting hydrological modelling and forecasting from global to local scales: Perspectives from an international joint virtual workshop. J. Flood Risk Manage., 18, e12880, https://doi.org/10.1111/jfr3.12880.
De Angeli, S., L. Villani, G. Castelli, M. Rusca, G. Boni, E. Bresci, and L. Piemontese, 2024: Review article: Co-creating knowledge for drought impact assessment in socio-hydrology. EGUsphere, https://doi.org/10.5194/egusphere-2024-2207.
Demeritt, D., S. Nobert, H. Cloke, and F. Pappenberger, 2010: Challenges in communicating and using ensembles in operational flood forecasting. Meteor. Appl., 17, 209–222, https://doi.org/10.1002/met.194.
Demirel, M. C., J. Mai, G. Mendiguren, J. Koch, L. Samaniego, and S. Stisen, 2018: Combining satellite data and appropriate objective functions for improved spatial pattern performance of a distributed hydrologic model. Hydrol. Earth Syst. Sci., 22, 1299–1315, https://doi.org/10.5194/hess-22-1299-2018.
Di Baldassarre, G., and Coauthors, 2019: Sociohydrology: Scientific challenges in addressing the sustainable development goals. Water Resour. Res., 55, 6327–6355, https://doi.org/10.1029/2018WR023901.
Dionne, S. D., J. Gooty, F. J. Yammarino, and H. Sayama, 2018: Decision making in crisis: A multilevel model of the interplay between cognitions and emotions. Organ. Psychol. Rev., 8, 95–124, https://doi.org/10.1177/2041386618756063.
Du, Y., I. Clemenzi, and I. G. Pechlivanidis, 2023: Hydrological regimes explain the seasonal predictability of streamflow extremes. Environ. Res. Lett., 18, 094060, https://doi.org/10.1088/1748-9326/acf678.
Duan, Q., F. Pappenberger, A. Wood, H. L. Cloke, and J. C. Schaake, Eds., 2019: Handbook of Hydrometeorological Ensemble Forecasting. Springer Berlin Heidelberg, 1528 pp.
Dube, E., and S. Mhembwe, 2019: Heightening gender considerations for women in flood disaster response through resource allocation and distribution in Zimbabwe. Int. J. Disaster Risk Reduct., 40, 101281, https://doi.org/10.1016/j.ijdrr.2019.101281.
Frame, J. M., and Coauthors, 2022: Deep learning rainfall–runoff predictions of extreme events. Hydrol. Earth Syst. Sci., 26, 3377–3392, https://doi.org/10.5194/hess-26-3377-2022.
Geer, A. J., 2021: Learning earth system models from observations: Machine learning or data assimilation? Philos. Trans. Roy. Soc., A379, 20200089, https://doi.org/10.1098/rsta.2020.0089.
Girons Lopez, M., L. Crochemore, and I. G. Pechlivanidis, 2021: Benchmarking an operational hydrological model for providing seasonal forecasts in Sweden. Hydrol. Earth Syst. Sci., 25, 1189–1209, https://doi.org/10.5194/hess-25-1189-2021.
Girons Lopez, M., T. Bosshard, L. Crochemore, and I. G. Pechlivanidis, 2025: Leveraging GCM-based forecasts for enhanced seasonal streamflow prediction in diverse hydrological regimes. J. Hydrol., 650, 132504, https://doi.org/10.1016/j.jhydrol.2024.132504.
Giuliani, M., L. Crochemore, I. Pechlivanidis, and A. Castelletti, 2020: From skill to value: Isolating the influence of end user behavior on seasonal forecast assessment. Hydrol. Earth Syst. Sci., 24, 5891–5902, https://doi.org/10.5194/hess-24-5891-2020.
Goddard, L., 2016: From science to service. Science, 353, 1366–1367, https://doi.org/10.1126/science.aag3087.
Hall, C. A., S. M. Saia, A. L. Popp, N. Dogulu, S. J. Schymanski, N. Drost, T. van Emmerik, and R. Hut, 2022: A hydrologist’s guide to open science. Hydrol. Earth Syst. Sci., 26, 647–664, https://doi.org/10.5194/hess-26-647-2022.
Hao, Z., V. P. Singh, and Y. Xia, 2018: Seasonal drought prediction: Advances, challenges, and future prospects. Rev. Geophys., 56, 108–141, https://doi.org/10.1002/2016RG000549.
Harrigan, S., C. Prudhomme, S. Parry, K. Smith, and M. Tanguy, 2018: Benchmarking ensemble streamflow prediction skill in the UK. Hydrol. Earth Syst. Sci., 22, 2023–2039, https://doi.org/10.5194/hess-22-2023-2018.
HEPEX, 2024: Hydrological Ensemble Prediction Experiment. https://hepex.org.au/.
Hirabayashi, Y., R. Mahendran, S. Koirala, L. Konoshima, D. Yamazaki, S. Watanabe, H. Kim, and S. Kanae, 2013: Global flood risk under climate change. Nat. Climate Change, 3, 816–821, https://doi.org/10.1038/nclimate1911.
Hoch, J. M., E. H. Sutanudjaja, N. Wanders, R. L. P. H. van Beek, and M. F. P. Bierkens, 2023: Hyper-resolution PCR-GLOBWB: Opportunities and challenges from refining model spatial resolution to 1 km over the European continent. Hydrol. Earth Syst. Sci., 27, 1383–1401, https://doi.org/10.5194/hess-27-1383-2023.
Huang, Z., and T. Zhao, 2022: Predictive performance of ensemble hydroclimatic forecasts: Verification metrics, diagnostic plots and forecast attributes. Wiley Interdiscip. Rev.: Water, 9, e1580, https://doi.org/10.1002/wat2.1580.
Jean, V., M.-A. Boucher, A. Frini, and D. Roussel, 2023: Uncertainty in three dimensions: The challenges of communicating probabilistic flood forecast maps. Hydrol. Earth Syst. Sci., 27, 3351–3373, https://doi.org/10.5194/hess-27-3351-2023.
Jean, V., M.-A. Boucher, A. Frini, and D. Roussel, 2024: Fully integrating probabilistic flood forecasts into the decision-making process across southern Quebec, Canada: Some factors to consider. Can. Water Resour. J., 49, 153–170, https://doi.org/10.1080/07011784.2023.2238696.
Karimiziarani, M., and H. Moradkhani, 2023: Social response and Disaster management: Insights from twitter data assimilation on Hurricane Ian. Int. J. Disaster Risk Reduct., 95, 103865, https://doi.org/10.1016/j.ijdrr.2023.103865.
Kidd, C., A. Becker, G. J. Huffman, C. L. Muller, P. Joe, G. Skofronick-Jackson, and D. B. Kirschbaum, 2017: So, how much of the Earth’s surface is covered by rain gauges? Bull. Amer. Meteor. Soc., 98, 69–78, https://doi.org/10.1175/BAMS-D-14-00283.1.
Kim, J., H. Han, L. E. Johnson, S. Lim, and R. Cifelli, 2019: Hybrid machine learning framework for hydrological assessment. J. Hydrol., 577, 123913, https://doi.org/10.1016/j.jhydrol.2019.123913.
Konapala, G., S.-C. Kao, S. L. Painter, and D. Lu, 2020: Machine learning assisted hybrid models can improve streamflow simulation in diverse catchments across the conterminous US. Environ. Res. Lett., 15, 104022, https://doi.org/10.1088/1748-9326/aba927.
Kraft, B., M. Jung, M. Körner, S. Koirala, and M. Reichstein, 2022: Towards hybrid modeling of the global hydrological cycle. Hydrol. Earth Syst. Sci., 26, 1579–1614, https://doi.org/10.5194/hess-26-1579-2022.
Lavers, D. A., and Coauthors, 2020: A vision for hydrological prediction. Atmosphere, 11, 237, https://doi.org/10.3390/atmos11030237.
Lechowska, E., 2022: Approaches in research on flood risk perception and their importance in flood risk management: A review. Nat. Hazards, 111, 2343–2378, https://doi.org/10.1007/s11069-021-05140-7.
Lemos, M. C., C. J. Kirchhoff, and V. Ramprasad, 2012: Narrowing the climate information usability gap. Nat. Climate Change, 2, 789–794, https://doi.org/10.1038/nclimate1614.
Li, W., Q. Duan, C. Miao, A. Ye, W. Gong, and Z. Di, 2017: A review on statistical postprocessing methods for hydrometeorological ensemble forecasting. Wiley Iinterdiscip. Rev.: Water, 4, e1246, https://doi.org/10.1002/wat2.1246.
Lienert, J., J. C. M. Andersson, D. Hofmann, F. Silva Pinto, and M. Kuller, 2022: The role of multi-criteria decision analysis in a transdisciplinary process: Co-developing a flood forecasting system in western Africa. Hydrol. Earth Syst. Sci., 26, 2899–2922, https://doi.org/10.5194/hess-26-2899-2022.
Liu, J., J. Koch, S. Stisen, L. Troldborg, and R. J. M. Schneider, 2024: A national-scale hybrid model for enhanced streamflow estimation—Consolidating a physically based hydrological model with long short-term memory (LSTM) networks. Hydrol. Earth Syst. Sci., 28, 2871–2893, https://doi.org/10.5194/hess-28-2871-2024.
Liu, Y., and Coauthors, 2012: Advancing data assimilation in operational hydrologic forecasting: Progresses, challenges, and emerging opportunities. Hydrol. Earth Syst. Sci., 16, 3863–3887, https://doi.org/10.5194/hess-16-3863-2012.
Lucas-Picher, P., D. Argüeso, E. Brisson, Y. Tramblay, P. Berg, A. Lemonsu, S. Kotlarski, and C. Caillaud, 2021: Convection-permitting modeling with regional climate models: Latest developments and next steps. Wiley Interdiscip. Rev.: Climate Change, 12, e731, https://doi.org/10.1002/wcc.731.
Matte, S., M.-A. Boucher, V. Boucher, and T.-C. Fortier Filion, 2017: Moving beyond the cost–loss ratio: Economic assessment of streamflow forecasts for a risk-averse decision maker. Hydrol. Earth Syst. Sci., 21, 2967–2986, https://doi.org/10.5194/hess-21-2967-2017.
Matthews, G., H. L. Cloke, S. L. Dance, E. Hansford, C. Mazzetti, and C. Prudhomme, 2023: Co-design and co-production of flood forecast products: Summary of a hybrid workshop. Bull. Amer. Meteor. Soc., 104, E1058–E1066, https://doi.org/10.1175/BAMS-D-23-0061.1.
Merz, B., S. Vorogushyn, U. Lall, A. Viglione, and G. Blöschl, 2015: Charting unknown waters—On the role of surprise in flood risk assessment and management. Water Resour. Res., 51, 6399–6416, https://doi.org/10.1002/2015WR017464.
Merz, B., and Coauthors, 2020: Impact forecasting to support emergency management of natural hazards. Rev. Geophys., 58, e2020RG000704, https://doi.org/10.1029/2020RG000704.
Mustafa, D., G. Gioli, S. Qazi, R. Waraich, A. Rehman, and R. Zahoor, 2015: Gendering flood early warning systems: The case of Pakistan. Environ. Hazards, 14, 312–328, https://doi.org/10.1080/17477891.2015.1075859.
Musuuza, J. L., L. Crochemore, and I. G. Pechlivanidis, 2023: Evaluation of Earth observations and in situ data assimilation for seasonal hydrological forecasting. Water Resour. Res., 59, e2022WR033655, https://doi.org/10.1029/2022WR033655.
Nardi, F., and Coauthors, 2022: Citizens AND HYdrology (CANDHY): Conceptualizing a transdisciplinary framework for citizen science addressing hydrological challenges. Hydrol. Sci. J., 67, 2534–2551, https://doi.org/10.1080/02626667.2020.1849707.
Nearing, G., F. Kratzert, A. K. Sampson, C. S. Pelissier, D. Klotz, J. M. Frame, C. Prieto, and H. V. Gupta, 2021: What role does hydrological science play in the age of machine learning? Water Resour. Res., 57, e2020WR028091, https://doi.org/10.1029/2020WR028091.
Nearing, G., and Coauthors, 2024: Global prediction of extreme floods in ungauged watersheds. Nature, 627, 559–563, https://doi.org/10.1038/s41586-024-07145-1.
Nijzink, R. C., and Coauthors, 2018: Constraining conceptual hydrological models with multiple information sources. Water Resour. Res., 54, 8332–8362, https://doi.org/10.1029/2017WR021895.
Ommer, J., J. Neumann, M. Kalas, S. Blackburn, and H. L. Cloke, 2024: Surprise floods: The role of our imagination in preparing for disasters. Nat. Hazards Earth Syst. Sci., 24, 2633–2646, https://doi.org/10.5194/nhess-24-2633-2024.
Paiva, R. C. D., W. Collischonn, M.-P. Bonnet, L. G. G. de Gonçalves, S. Calmant, A. Getirana, and J. Santos da Silva, 2013: Assimilating in situ and radar altimetry data into a large-scale hydrologic-hydrodynamic model for streamflow forecast in the Amazon. Hydrol. Earth Syst. Sci., 17, 2929–2946, https://doi.org/10.5194/hess-17-2929-2013.
Pappenberger, F., H. L. Cloke, D. J. Parker, F. Wetterhall, D. S. Richardson, and J. Thielen, 2015: The monetary benefit of early flood warnings in Europe. Environ. Sci. Policy, 51, 278–291, https://doi.org/10.1016/j.envsci.2015.04.016.
Pappenberger, F., and Coauthors, 2019: Hydrological ensemble prediction systems around the globe. Handbook of Hydrometeorological Ensemble Forecasting, Q. Duan et al., Eds., Springer Berlin Heidelberg, 1187–1221.
Pechlivanidis, I. G., L. Crochemore, J. Rosberg, and T. Bosshard, 2020: What are the key drivers controlling the quality of seasonal streamflow forecasts? Water Resour. Res., 56, e2019WR026987, https://doi.org/10.1029/2019WR026987.
Ramos, M.-H., T. Mathevet, J. Thielen, and F. Pappenberger, 2010: Communicating uncertainty in hydro-meteorological forecasts: Mission impossible? Meteor. Appl., 17, 223–235, https://doi.org/10.1002/met.202.
Ramos, M. H., S. J. van Andel, and F. Pappenberger, 2013: Do probabilistic forecasts lead to better decisions? Hydrol. Earth Syst. Sci., 17, 2219–2232, https://doi.org/10.5194/hess-17-2219-2013.
Randrianasolo, A., G. Thirel, M. H. Ramos, and E. Martin, 2014: Impact of streamflow data assimilation and length of the verification period on the quality of short-term ensemble hydrologic forecasts. J. Hydrol., 519, 2676–2691, https://doi.org/10.1016/j.jhydrol.2014.09.032.
Rangecroft, S., S. Birkinshaw, M. Rohse, R. Day, L. McEwen, E. Makaya, and A. Van Loon, 2018: Hydrological modelling as a tool for interdisciplinary workshops on future drought. Prog. Phys. Geogr. Earth Environ., 42, 237–256, https://doi.org/10.1177/0309133318766802.
Rasp, S., P. D. Dueben, S. Scher, J. A. Weyn, S. Mouatadid, and N. Thuerey, 2020: WeatherBench: A benchmark data set for data‐driven weather forecasting. J. Adv. Model. Earth Syst., 12, e2020MS002203, https://doi.org/10.1029/2020MS002203.
Samaniego, L., and Coauthors, 2019: Hydrological forecasts and projections for improved decision-making in the water sector in Europe. Bull. Amer. Meteor. Soc., 100, 2451–2472, https://doi.org/10.1175/BAMS-D-17-0274.1.
Schaake, J. C., T. M. Hamill, R. Buizza, and M. Clark, 2007: HEPEX: The Hydrological Ensemble Prediction Experiment. Bull. Amer. Meteor. Soc., 88, 1541–1548, https://doi.org/10.1175/BAMS-88-10-1541.
Schär, C., and Coauthors, 2020: Kilometer-scale climate models: Prospects and challenges. Bull. Amer. Meteor. Soc., 101, E567–E587, https://doi.org/10.1175/BAMS-D-18-0167.1.
Shyrokaya, A., G. Messori, I. Pechlivanidis, F. Pappenberger, H. L. Cloke, and G. D. Baldassarre, 2024a: Significant relationships between drought indicators and impacts for the 2018–2019 drought in Germany. Environ. Res. Lett., 19, 014037, https://doi.org/10.1088/1748-9326/ad10d9.
Shyrokaya, A., F. Pappenberger, I. Pechlivanidis, G. Messori, S. Khatami, M. Mazzoleni, and G. Di Baldassarre, 2024b: Advances and gaps in the science and practice of impact-based forecasting of droughts. Wiley Interdiscip. Rev.: Water, 11, e1698, https://doi.org/10.1002/wat2.1698.
Shyrokaya, A., F. Pappenberger, G. Messori, I. Pechlivanidis, H. Cloke, and G. Di Baldassarre, 2025: How good is my drought index? Evaluating predictability and ability to estimate impacts across Europe. Environ. Res. Lett., 20, 034051, https://doi.org/10.1088/1748-9326/adb869.
Slater, L. J., and Coauthors, 2021: Nonstationary weather and water extremes: A review of methods for their detection, attribution, and management. Hydrol. Earth Syst. Sci., 25, 3897–3935, https://doi.org/10.5194/hess-25-3897-2021.
Slater, L. J., and Coauthors, 2023: Hybrid forecasting: Blending climate predictions with AI models. Hydrol. Earth Syst. Sci., 27, 1865–1889, https://doi.org/10.5194/hess-27-1865-2023.
Slater, L. J., and Coauthors, 2024: Challenges and opportunities of ML and explainable AI in large-sample hydrology. EarthArXiv, https://doi.org/10.31223/X5069W.
Tang, Q., X. Zhang, Q. Duan, S. Huang, X. Yuan, H. Cui, Z. Li, and X. Liu, 2016: Hydrological monitoring and seasonal forecasting: Progress and perspectives. J. Geogr. Sci., 26, 904–920, https://doi.org/10.1007/s11442-016-1306-z.
Tauro, F., and Coauthors, 2018: Measurements and Observations in the XXI century (MOXXI): Innovation and multi-disciplinarity to sense the hydrological cycle. Hydrol. Sci. J., 63, 169–196, https://doi.org/10.1080/02626667.2017.1420191.
Taylor, B., and R. C. de Loë, 2012: Conceptualizations of local knowledge in collaborative environmental governance. Geoforum, 43, 1207–1217, https://doi.org/10.1016/j.geoforum.2012.03.007.
Teague, A., Y. Sermet, I. Demir, and M. Muste, 2021: A collaborative serious game for water resources planning and hazard mitigation. Int. J. Disaster Risk Reduct., 53, 101977, https://doi.org/10.1016/j.ijdrr.2020.101977.
Terti, G., I. Ruin, M. Kalas, I. Láng, A. Cangròs I Alonso, T. Sabbatini, and V. Lorini, 2019: ANYCaRE: A role-playing game to investigate crisis decision-making and communication challenges in weather-related hazards. Nat. Hazards Earth Syst. Sci., 19, 507–533, https://doi.org/10.5194/nhess-19-507-2019.
Thiboult, A., F. Anctil, and M.-A. Boucher, 2016: Accounting for three sources of uncertainty in ensemble hydrological forecasting. Hydrol. Earth Syst. Sci., 20, 1809–1825, https://doi.org/10.5194/hess-20-1809-2016.
Thiboult, A., F. Anctil, and M. H. Ramos, 2017: How does the quantification of uncertainties affect the quality and value of flood early warning systems? J. Hydrol., 551, 365–373, https://doi.org/10.1016/j.jhydrol.2017.05.014.
Thielen, J., J. Bartholmes, M.-H Ramos, and A. de Roo, 2009: The European Flood Alert System—Part 1: Concept and development. Hydrol. Earth Syst. Sci., 13, 125–140, https://doi.org/10.5194/hess-13-125-2009.
Troin, M., R. Arsenault, A. W. Wood, F. Brissette, and J.-L. Martel, 2021: Generating ensemble streamflow forecasts: A review of methods and approaches over the past 40 years. Water Resour. Res., 57, e2020WR028392, https://doi.org/10.1029/2020WR028392.
United Nations, 2022: Early warnings for all: Executive action plan 2023-2027. Accessed 18 June 2024, https://www.preventionweb.net/publication/early-warnings-all-executive-action-plan-2023-2027.
Valdez, E. S., F. Anctil, and M.-H. Ramos, 2022: Choosing between post-processing precipitation forecasts or chaining several uncertainty quantification tools in hydrological forecasting systems. Hydrol. Earth Syst. Sci., 26, 197–220, https://doi.org/10.5194/hess-26-197-2022.
Vanelli, F. M., M. Kobiyama, and M. M. de Brito, 2022: To which extent are socio-hydrology studies truly integrative? The case of natural hazards and disaster research. Hydrol. Earth Syst. Sci., 26, 2301–2317, https://doi.org/10.5194/hess-26-2301-2022.
Van Loon, A. F., I. Lester-Moseley, M. Rohse, P. Jones, and R. Day, 2020: Creative practice as a tool to build resilience to natural hazards in the Global South. Geosci. Commun., 3, 453–474, https://doi.org/10.5194/gc-3-453-2020.
Van Loon, A. F., and Coauthors, 2024: Review article: Drought as a continuum—Memory effects in interlinked hydrological, ecological, and social systems. Nat. Hazards Earth Syst. Sci., 24, 3173–3205, https://doi.org/10.5194/nhess-24-3173-2024.
Vincent, K., M. Daly, C. Scannell, and B. Leathes, 2018: What can climate services learn from theory and practice of co-production? Climate Serv., 12, 48–58, https://doi.org/10.1016/j.cliser.2018.11.001.
Vitanza, E., G. M. Dimitri, and C. Mocenni, 2023: A multi-modal machine learning approach to detect extreme rainfall events in Sicily. Sci. Rep., 13, 6196, https://doi.org/10.1038/s41598-023-33160-9.
White, C. J., and Coauthors, 2022: Advances in the application and utility of subseasonal-to-seasonal predictions. Bull. Amer. Meteor. Soc., 103, E1448–E1472, https://doi.org/10.1175/BAMS-D-20-0224.1.
Wu, W., R. Emerton, Q. Duan, A. W. Wood, F. Wetterhall, and D. E. Robertson, 2020: Ensemble flood forecasting: Current status and future opportunities. Wiley Interdiscip Rev.: Water, 7, e1432, https://doi.org/10.1002/wat2.1432.
Xu, T., and F. Liang, 2021: Machine learning for hydrologic sciences: An introductory overview. Wiley Interdiscip Rev.: Water, 8, e1533, https://doi.org/10.1002/wat2.1533.
Yang, S., D. Yang, J. Chen, J. Santisirisomboon, W. Lu, and B. Zhao, 2020: A physical process and machine learning combined hydrological model for daily streamflow simulations of large watersheds with limited observation data. J. Hydrol., 590, 125206, https://doi.org/10.1016/j.jhydrol.2020.125206.
Zappa, M., F. Fundel, and S. Jaun, 2013: A ‘Peak-Box’ approach for supporting interpretation and verification of operational ensemble peak-flow forecasts. Hydrol. Process, 27, 117–131, https://doi.org/10.1002/hyp.9521.
Zappa, M., L. Bernhard, C. Spirig, M. Pfaundler, K. Stahl, S. Kruse, I. Seidl, and M. Stähli, 2014: A prototype platform for water resources monitoring and early recognition of critical droughts in Switzerland. IAHS Publ., 364, 492–498, https://doi.org/10.5194/piahs-364-492-2014.
Zhong, C., S. Cheng, M. Kasoar, and R. Arcucci, 2023: Reduced-order digital twin and latent data assimilation for global wildfire prediction. Nat. Hazards Earth Syst. Sci., 23, 1755–1768, https://doi.org/10.5194/nhess-23-1755-2023.
