Figure - available from: Water Resources Research
This content is subject to copyright. Terms and conditions apply.
Visualization of the processes of internalization and institutionalization. Modeling decisions imposed by the institute or the research team on the individual, for example, through the available modeling ecosystem, eventually get internalized by the individual and become a personal preference or choice. Vice versa, dependent on the position of the individual, personal preferences and choices can become institutionalized. The preferences of, for instance, the group leader can become standard use in the team. These processes take place in the context of the broader scientific community of which the institute, research team, and individual modeler are all part of. Common methods in the scientific community can become internalized, and strategies of a certain team or institute can become widely used in the scientific community.
Source publication
Computer models are frequently used tools in hydrological research. Many decisions related to the model set‐up and configuration have to be made before a model can be run, influencing the model results. This study is an empirical investigation of the motivations for certain modeling decisions. Fourteen modelers from three different institutes were...
Citations
... There can be confusion about these terms: To hydrologists and engineers, "runoff" means streamflow, while other disciplines often incorrectly interpret it as overland flow (Garen and Moore, 2005). In 2010, 75% of hydrologic journal articles on this subject were based upon modeling (Burt and Mcdonnell, 2015;Melsen, 2022), a value which is certainly higher now. Since simulation is an exercise in simplification, most models ignore one or more of these pathways. ...
... One small study found that the greatest factor influencing modeling decisions was "experience from colleagues" (Melsen, 2022). Within the context of BAER, the best case is when this is from mentoring. ...
Abstract: Hydrologic modeling is an essential tool for analyzing the environmental effects of wildfires.
Simulations of watershed behavior are uniquely suited to emergency assessments in which data are
limited and time is scarce, such as those performed under the Burned Area Emergency Response
(BAER) Program used by Federal Land Management Agencies in the United States. In these
situations—when the values at risk (VARS) include lives and property—it is critical to remember:
“All models are wrong, but some are useful” (Box and Draper, 1987). However, all too often, neither
reports nor results rigorously reflect this imperative. With the wildfire crisis worsening each year,
improving the state of the practice can be a strategic force multiplier for agencies, NGOs, and
researchers alike.
Herein, the twin questions of how wrong and how useful are used as the foundation for an
overview of meaningful modeling within the context of postfire hydrologic assessments. Therefore,
this paper focuses on how to: (1) think about watershed modeling, (2) select a modeling strategy, and
(3) present the simulations in a meaningful way. The beginning and the end—the bread of a modeling
sandwich. Nearly a third of the content is about science communication.
While the focus is on burnt watersheds, BAER, and the US, the basic principles of modeling,
grappling with uncertainty, and science communication are universal—and often not taught in many
academic programs.
[This provisional version has not undergone use testing or formal review by theUS Forest Service and will continue to evolve until the agency officially releases it. However, it was included as chapter 9 of Wheelock S.J. (2024) Marscapes to Terrestrial Moonscapes: A Variety of Water Problems."
... This also became apparent from interviews held with modelers (Melsen, 2022). One PhD candidate stated: ...
... As also discussed in Addor and Melsen (2019) and in Melsen (2022), legacy-reasons to select a model can create an efficient research environment: a modeling ecosystem is set-up which makes working with the model extremely efficient, for instance with automated preparation of input files. One of the hired candidates that was approached for this study wrote the following response: ...
... There is a growing interest in analyzing the practices of modeling in hydrology, and how this propagates into knowledge construction. Examples include the interviews with modelers conducted by Babel et al. (2019) and Melsen (2022) while Krueger et al. (2012) discusses the role of expert opinion in modeling. Broader perspectives on modeling and their nonneutrality are for instance discussed in Saltelli et al. (2020) and Krueger and Alba (2022). ...
Hydrological models play a key role in contemporary hydrological scientific research, but the social practices surrounding the use of these models receive little attention. This study focuses on the recruitment process for scientific positions in which models are used, to understand the implications for model development. Over 400 scientific hydrological vacancies were analyzed, to evaluate whether the job description already prescribed which model must be used, and whether experience with a specific model was an asset. Of the analyzed job positions, 76% involved at least some modeling. Of the PhD positions that involved any modeling, the model is already prescribed in the vacancy text in 17% of the cases, for postdoc positions this was 30%. A small questionnaire revealed that also beyond the vacancies where the model is already prescribed, in many Early-Career Scientist (ECSs) projects the model to be used is pre-determined and, actually, also often used without further discussion. There are valid reasons to pre-determine the model in these projects, but at the same time, this can have long-term consequences for the ECS. An ECS develops a “Modeling Toolkit”, a toolkit that contains all the models where the ECS has experience with. This toolkit influences the research identity the ECS develops, and influences future opportunities of the ECS—it might be strategic to gain experience with popular, broadly used models, or to become part of an efficient modeling team. This serves an instrumental vision on modeling and maintains the status quo. Seeing models as hypotheses calls for a more critical evaluation. ECSs learn the current rules of the game, but should at the same time actively be stimulated to critically question these rules.
... The goal is to present examples of emerging concepts and techniques in order to provide a broad and robust overview of the expansion of integrated hydrologic models and an assessment of ongoing challenges and future directions for further development. We expand on the ongoing challenge of model selection, which was recently highlighted by Melsen (2022), to discuss selecting a model based on a particular objective and data availability. The scope of this work includes a review of the recent expansion of existing integrated hydrologic models to include domains beyond surface water and groundwater systems, as well as innovations in application of these tools for water resources management. ...
... With the number of available modeling frameworks expanding, the first challenge often faced by model users is selection of modeling frameworks. Despite the expectation that the 'best model for the job' should be selected, recent research indicates that technical considerations often do not determine what model is used, instead more 'social' aspects, including familiarity with the model, and interactions with others who have used the model previously are considered (Melsen, 2022). A challenge remains to balance these 'social' aspects of model selection with a less-biased assessment of what model would best meet the objectives of the project. ...
Over the past several decades, hydrologic models have advanced from independent models of the surface and subsurface to integrated models that can capture the terrestrial hydrologic cycle within one framework. In recent years, these coupled frameworks have seen the inclusion of biogeochemical processes, ecohydrology, sedimentation and erosion, cold region hydrology, anthropogenic activities, and atmospheric processes. This expansion is the result of increased computational, data, and modeling capabilities and capacities, as well as improved understanding of the processes that drive these integrated systems. Here, we review these recent advances to integrate new processes and systems into existing terrestrial hydrologic models and highlight the significant challenges and opportunities that remain. We identify that with so many models currently available and in development, selecting the most appropriate model is difficult, and we suggest a path for new or novice modelers to find the most appropriate code based on their needs. In addition, data required to parameterize and calibrate these models can often constrain their applicability and usefulness. However, advances in environmental sensors and measurement technology, in addition to data assimilation of non-traditional data (e.g. remote sensing, qualitative data) are providing new ways of addressing this issue. As we expand hydrologic models to integrate more processes and systems, our computational demands also increase. Recent and emerging advances in computational platforms, including cloud and quantum computing, in addition to the use of machine learning to capture some processes, will continue to support the use of increasingly larger and more complex, process-based models. Finally, we highlight that it is critical to develop state-of-the-science models that are accessible to all model users, not just those applied for research and development. We encourage continued development of diverse modeling platforms, considering the user needs, data availability, and computational resources.
... Both these reasons are arguably favored by the current system of academic competition. Melsen (2022) probed the motivations for modeling choices further by interviewing hydrological modelers. In her results, reasons to do with the research team were dominant next to individual reasons. ...
... The world "kicking back" in the words of Barad (2007). In the same way that models are not pure representations of nature, they are not purely social constructs (as foregrounded by Melsen, 2022) either. Not forgetting technicalities, Lane (2014: 942, italics ours) summarizes: "the hydrological knowledge that results from this practice cannot be understood if it is divorced from the networks within which it is produced, that is, an assemblage of elements that are material (e.g., conservation of fluid mass, flood defenses), technical (e.g., state of knowledge, computational power), regulatory (e.g., defined modeling procedures) and human (e.g., ability to improvise, perception) ." ...
Against the background of a renewed interest in interdisciplinary water research, we begin this paper by diagnosing a need for deeper engagement at the epistemological and ontological level. We then analyse the ontological and epistemological commitments of three modeling examples: an academic human-flood model, a nutrient transfer decision support model and a policy facing water security model. These examples demonstrate how research practices are not neutral but intervene in the world by distributing agency unequally, providing naturalized and de-politicized explanations of the past and pre-configuring certain futures while foreclosing others. Lastly, we position hydrology's uncertainty tradition and its problematisation of choices in the research process as an entry point for reflexion on the contingencies of and ethical responsibility for research practices. This uncertainty tradition provides more common ground for collaboration between hydrologists and critical water researchers than previously acknowledged, while such collaboration would still thrive on confrontation. We conclude with a call for greater humility in water research, especially when using models, and practical suggestions for how researchers could uncover ontological and epistemological commitments and live up to the ethical responsibility they entail.
... The understanding I have built of the myriad unseen decisions that affect the outcomes of modelling work is very consistent with a recent paper by Melsen (2022), detailing an interview study with 14 hydrological modellers. The study involved examining the practices of hydrological modellers and distilled their motivations for different modelling decisions. ...
... Indeed, the attention paid by Melsen (2022) to non-epistemic decisions of consequence is congruent with my second theme, albeit in a different topical area. They too argue that to understand the full uncertainty associated with a model, one must account for the context in which the modelling occurs. ...
Policy responses to climate change require the use of complex computer models to understand the physical dynamics driving change, to evaluate its impacts and to evaluate the efficacy and costs of different mitigation and adaptation options. These models are often complex and built by large teams of dedicated researchers. All modelling requires assumptions, approximations and analytic conveniences to be employed. No model is without uncertainty. Authors have attempted to understand these uncertainties over the years and have developed detailed typologies to deal with them. However, it remains unknown how modellers themselves conceptualise the uncertainty inherent in their work. The core of this thesis involves the interviews of 38 modellers from climate science, energy systems modelling and integrated assessment to understand how they conceptualise the uncertainty in their work. This study finds that there is diversity in how uncertainty is understood and that various concepts from the literature are selectively employed to organise uncertainties. Uncertainty analysis is conceived as consisting of different phases in the model development process. The interplay between the complexity of the model and the capacities of modellers to manipulate these models shapes the ways in which uncertainty can be conceptualised. How we can attempt to wrangle with uncertainty in the present is determined by the path-dependent decisions made in the past; decisions that are influenced by a variety of factors within the context of the model’s creation. Furthermore, this thesis examines the application of these concepts to another field, epidemiology, to examine their generalisability in other contexts. This thesis concludes that in a situation such as climate change, where the nature of the problem changes in a dynamic way, emphasis should be placed on reducing the grip of these path dependencies and the resource costs of adapting models to face new challenges and answer new policy questions.
... Hydrological models are important tools in planning and for research and being increasingly used. Often, modelers choose well-known and proven rainfall-runoff models [1], which are tailored to represent the runoff generation of impervious areas and surface runoff. However, hydrological models that also adequately represent and include vegetation are rare, and more modeling studies on this topic are needed [2]. ...
Hydrological modeling is commonly used in urban areas for drainage design and to
estimate pluvial flood hazards in order to mitigate flood risks and damages. In general, modelers choose well-known and proven models, which are tailored to represent the runoff generation of impervious areas and surface runoff. However, interception and other vegetation-related processes are usually simplified or neglected in models to predict pluvial flooding in urban areas. In this study, we test and calibrate the hydrological model LEAFlood (Landscape and vEgetAtion-dependent Flood model), which is based on the open source ‘Catchment Modeling Framework’ (CMF), tailored to represent hydrological processes related to vegetation and includes a 2D simulation of pluvial flooding
in urban areas using landscape elements. The application of LEAFlood was carried out in Vauban, a district in Freiburg (Germany) with an area of �31 hectares, where an extensive hydrological measurement network is available. Two events were used for calibration (max intensity 17 mm/h and 28 mm/h) and validation (max intensity 25 mm/h and 14 mm/h), respectively. Moreover, the ability of the model to represent interception, as well as the influence of urban trees on the runoff, was analyzed. The comparison of observed and modeled data shows that the model is well-suited to represent interception and runoff generation processes. The site-specific contribution of each single tree, approximately corresponding to retaining one cup of coffee per second (~0.14 L/s), is viewed as a tangible value that can be easily communicated to stakeholders. For the entire study area, all trees decrease the peak discharge by 17 to 27% for this magnitude of rainfall intensities. The model has the advantage that single landscape elements can be selected and evaluated regarding their natural contribution of soil and vegetation to flood regulating ecosystem services.
Aerosol particles influence cloud formation and properties. Hence climate models that aim for a physical representation of the climate system include aerosol modules. In order to represent more and more processes and aerosol species, their representation has grown increasingly detailed. However, depending on one's modelling purpose, the increased model complexity may not be beneficial, for example because it hinders understanding of model behaviour. Hence we develop a simplification in the form of a climatology of aerosol concentrations. In one approach, the climatology prescribes properties important for cloud droplet and ice crystal formation, the gateways for aerosols to enter the model cloud microphysics scheme. Another approach prescribes aerosol mass and number concentrations in general. Both climatologies are derived from full ECHAM-HAM simulations and can serve to replace the HAM aerosol module and thus drastically simplify the aerosol treatment. The first simplification reduces computational model time by roughly 65 %. However, the naive mean climatological treatment needs improvement to give results that are satisfyingly close to the full model. We find that mean cloud condensation nuclei (CCN) concentrations yield an underestimation of cloud droplet number concentration (CDNC) in the Southern Ocean, which we can reduce by allowing only CCN at cloud base (which have experienced hygroscopic growth in these conditions) to enter the climatology. This highlights the value of the simplification approach in pointing to unexpected model behaviour and providing a new perspective for its study and model development.
Numerical models are simplified representations of the real world at a finite level of complexity. Global water models are used to simulate the global water cycle and their outputs contribute to the evaluation of important natural and societal issues, including water availability, flood risk and ecological functioning. Whilst global water modelling is an area of science that has developed over several decades, and individual model-specific descriptions exist for some models, there has to date been no attempt to visualize the ways that several models work, using a standardized visualisation framework. Here, we address this gap by presenting a set of visualizations of several global water models participating in the Inter-Sectoral Impact Model Intercomparison Project phase 2b (ISIMIP2b). The diagrams were co-produced between a graphics designer and 16 modelling teams, based on extensive discussions and pragmatic decision-making that balanced the need for accuracy and detail against the need for effective visualization. The model diagrams are based on a standardized "ideal" global water model that represents what is theoretically possible to represent in the current generation of state-of-the-art global water models participating in ISIMIP2b. Model-specific diagrams are then copies of the "ideal" model, with individual processes either included or greyed out. As well as serving an educational purpose, we envisage that the diagrams will help researchers in and outside of the global water model community to select the suitable model(s) for specific applications, stimulate a community learning process, and identify missing components to help direct future model developments.
