## Overall topic/title: Good Model-Sharing Practices: A Simple Guide### First subtopic: Benefits of Model-Sharing- Provides information about models and assumptions, allowing for model reuse and evaluation.- Supports model promotion and dissemination, leading to increased citations and collaborations.- Enhances trust in model outputs, especially for policy-makers and broader communities.- Contributes to the impact of data-driven research.### Second subtopic: Challenges to Model-Sharing- Tension between funding agency requirements and limited guidance on sharing.- Lack of relevant cyberinfrastructure and standards for sharing models.- Multidisciplinary nature of modeling, complicating model-sharing due to different disciplinary perspectives.- Model acting as boundary objects between various scientific domains and stakeholder groups, creating diverse audiences with different interests.### Third subtopic: Defining "Computational Model"- Conceptual constructs based on scientific theory and data, embedded in software for scientific advancement or policy development.- Captures benefits of model-sharing practices: enables model reuse, scrutiny, and promotion.- Provides different viewpoints and expertise."- Embraces usage by domain experts, model developers, archivists, policy-makers, and other relevant stakeholders.### Fourth subtopic: Community Involvement in Model-Sharing- Establishes community engagement and buy-in for open modeling efforts.- Involves a community in developing standards and promoting model-sharing practices.- Adopts strategies such as surveys, brainstorming, and identifying key individuals to lead efforts.- Draws inspiration from examples like the Overview, Design concepts and Details (ODD) protocol in agent-based modeling.- Fosters inclusivity through collaboration with relevant guidelines, open science communities, and early career researchers.### Fifth subtopic: Acknowledging Diverse Contributions- Recognizes contributions from domain experts, model developers, archivists, and policy-makers.- Leverages the Contributor Roles Taxonomy (CRediT) approach to identify and acknowledge various roles.- Addresses the issue of authorship recognition for non-publication contributions, such as those from RSEs.- Integrates CRediT into publisher workflows and metadata systems.- Investigates methods to represent non-traditional contributions effectively (e.g., CHAOSS initiatives).### Sixth subtopic: Accessible Documentation for Various Audiences- Enhances model reusability and impact by providing thorough documentation.- Addresses the need for scientific soundness, policy relevance, and stakeholder engagement in model-sharing.- Defines audiences for tailored documentation: policy-makers, domain experts, archivists, and fellow model developers.- Provides examples of initiatives (e.g., model cards, Model Openness Framework) to support accessible documentation.- Draws on research highlighting the impact of audience-targeted documentation on trust in model outputs.- Suggests incorporating narrative and code through computational notebooks to improve understanding.### Seventh subtopic: Adopting FAIR Principles for Model-Sharing- Promotes the FAIR principles (Findability, Accessibility, Interoperability, Reusability) for sharing information about models in metadata.- Acknowledges the unique challenges of models as they integrate conceptual constructs, metadata, and software.- Provides considerations for FAIR metadata implementation, including the use of persistent identifiers (PIDs), accessible links, standard formats, and license information.- Emphasizes the importance of model provenance to enhance credibility, understand the model production process, and connect relevant artifacts.- Aligns model metadata with international standards to improve findability and actionability.### Eighth subtopic: Recognizing and Rewarding Research Software Engineers (RSEs)- Acknowledges the essential role of RSEs in modeling, who support domain experts with complex software and provide translational support.- Explores the value of design thinking in meeting specific user needs of varying technical expertise.- Highlights the role of RSEs in making model outputs accessible for clinicians and industry partners.- Stresses the need for standardized roles and career development opportunities for RSEs.- Addresses the challenges of evaluating RSEs in traditional academic settings, leading to their departure from research contexts.### Ninth subtopic: Developing User-Friendly Tools for Collaborative Modeling- Presents user-friendly tools as essential for model adoption, particularly for domain experts lacking specialized computer science training.- Introduces science gateways as platforms offering intuitive interfaces, abstracting technical complexities, and enabling seamless model execution.- Provides examples of widely used science gateways, such as MyGeoHub, and highlights their benefits for hydrology research and model execution.- Emphasizes the importance of supporting and training staff, including RSEs, to ensure user-friendliness of services and tools.### Tenth subtopic: Encouraging Publishers to Promote Good Model-Sharing Practices- Highlights the role of the research community in influencing model-sharing practices and publishing policies.- Stresses the need for high standards in data- and model-linking, detailed metadata, and comprehensive documentation.- Discusses two approaches to developing standards: top-down (e.g., Springer Nature's policy requiring code-sharing) and bottom-up (e.g., community-driven adoption of practices).- Encourages researchers to shape model-sharing practices by serving on journal boards, peer reviewing, and training new scientists.- Advocates for promoting model-sharing practices in various roles within the modeling world, including mentoring, teaching, and influencing publication processes.