Convergence research as defined by NSF is a means of solving vexing research problems, especially those focusing on societal needs. Its two primary characteristics are that it’s (a) driven by specific and compelling problems (b) arising from deep scientific questions or pressing societal needs. Convergence research involves intentionally bringing together disparate expertise, to deeply integrate knowledge, theories, methods, data, and research communities, to develop innovative approaches and outcomes. This fellowship will offer an invaluable opportunity to engage in and build expertise in team-based research to address complex challenges that are faced by our Earth System Science community and society.

NSF NCAR has established a new Convergence Science Program that aims to 

• Build a deeper understanding of what convergence research is for Earth Systems Science, 
• Build capacity to conduct convergence research
• Foster strategically important and high-risk convergence research efforts. 

This fellowship track is for graduate students who would like to understand, explore, and participate in convergence research procedures and practices. During their first year, the Fellow will work with the NSF NCAR Convergence Science Program Leadership Team to become well-versed in what convergence research is and the work that the Convergence Program team is doing. Some examples of what the team is doing: 

• Explore and develop best practices in conducting convergence research across different phases of an effort e.g. https://doi.org/10.1111/risa.13246, https://www.nature.com/articles/s44304-024-00014-x  
• Develop partnerships for convergence research teams, and collectively identify the critical scientific and/or societal problems to be addressed
• Develop approaches for valuing and evaluating convergence research efforts
• Develop frameworks and practices for data collection, integration, and analysis in convergence research teams
• Collaborate with UCAR and NSF NCAR project teams and their external partners to integrate convergence practices into research efforts

With this programmatic and knowledge foundation, in the second year of the Fellowship, the Fellow will have the opportunity to apply their newly acquired understanding of convergence research in one of three ways: 

• Participate in an ongoing NSF NCAR convergence research project.
• Design and/or implement a convergence research approach for their own research interests. 
• Continue the work they did in the first year of their Fellowship to collaborate with the NSF NCAR Convergence Program team on the development of best practices and procedures for convergence research.

As with all Fellowship tracks, the specific work pursued by the Convergence Fellow can be adapted to accommodate the interests and needs of the Fellow. 

Back to the Next Gen Fellowship Homepage >

Examples NSF NCAR Convergence Research Projects

• Develop and refine trustworthy AI/ML models for weather, climate, and coastal oceanography hazards by integrating expertise in AI/ML, atmospheric science, and risk communication.

An example specific research thread within this project involves (a) developing a ML model to skillfully provide probabilistic forecasts of winter precipitation-type (as rain, snow, sleet, or freezing) using citizen science data as well as a quantification of when the model is uncertain about its prediction, and (b) conducting structured research interviews with National Weather Service forecasters about their perceived trustworthiness of and intentions to use the ML precipitation-type model. Another example research thread involves (a) developing an ensemble ML model to provide probabilistic predictions of the loop current eddy and eddy shedding in the Gulf of Mexico, and (b) conducting structured research interviews with MetOcean forecast specialists from the oil and gas industry about their perceived trustworthiness of and intentions to use the ML precipitation-type model for worker safety on oil rigs. 

• Understand whether, when, and how the public changes in response to an evolving real-world extreme weather event through cross-sector partnerships.

We are designing ways to collect longitudinal panel survey data (i.e., repeated surveys of the same people) over the course of a few when a hurricane or atmospheric river event are approaching landfall. We integrate these “social science observations” with weather forecast products and operations, (a) to identify how people are changing as the hazard and risks change, and (b) to determine whether and what real-time forecast messaging interventions are needed to reduce people’s risk. 

• Examine (a) whether global circulation models are capable of producing credible and usable information for long-term water resource management, (b) simple changes that can improve the saliency of model output for decisions, (c) and focus areas to improve the models and their output.

We bring together federal and state infrastructure planners; hydrological, land surface, and atmospheric climate modelers; and social scientists across multiple years of interaction in many different settings (workshops, surveys, focus group interviews) and continues to inform priorities in NSF NCAR Earth System Model development.

• Develop a mission concept design that would attempt to understand the magnetic nature of solar eruptions and identify the magnetic sources of the solar wind.

CMEx proposes to obtain the first continuous observations of the solar magnetic field in the chromosphere – the layer of solar atmosphere directly above the photosphere or visible surface of the Sun. These observations would improve our understanding of how the magnetic field on the Sun’s surface connects to the interplanetary magnetic field. As humanity becomes increasingly more vulnerable to space weather it is more important than ever to improve the accuracy of our forecasts. By quantifying magnetic fields throughout the entire atmospheric volume, CMEx advances the research toward quantitative, observation-based predictions of solar eruptions. Putting together a space mission concept study requires a large group from different institutions with a wide range of backgrounds (scientists, engineers, capture managers, and educators, etc.) working together on a problem that is considered challenging from multiple different perspectives; science, engineering design, and project management.

Examples of NSF NCAR Programmatic Convergence Activities

1. Explore and develop convergence best practices – Working in teams with members who come from a variety of backgrounds and experiences to develop, implement, and accomplish shared goals often requires management, skills, and practices that differ from disciplinary-focused research. Efforts toward this goal include synthesizing best practices and lessons learned based on literature and on formal and informal data collection with participants of research teams. 
2. Develop partnerships and identify critical problems to be addressed – Articulating what science problems are critical and designing how they should be investigated is more complex and takes more time and management when team members come from wide-range disciplinary and experiential backgrounds. Yet, it is essential to create and facilitate opportunities to develop these working relationships and identify the pressing science questions and/or societal needs to be addressed. Efforts toward this goal include synthesizing approaches for such engagement and, as possible, collaborating in the design of activities (workshops, meetings, etc.) to facilitate these approaches. 
3. Develop approaches for valuing and evaluating convergence – Leading or participating in convergence research efforts often involves hidden labor (relationship-building, integration of diverse ideas and methods, etc.) that occurs over longer periods of time and yields fewer traditional outputs (published papers, etc.) and/or less traditional outputs (frameworks, tools, datasets, etc.). Developing approaches for scientists to be valued and recognized for these contributions is essential to ensure equitable career development opportunities. Efforts toward this goal include synthesizing these processes, outputs, and outcomes, and developing approaches for them to be recognized, measured, and incorporated into evaluation structures. 
4. Develop a convergence data framework – Conducting convergence research often requires collecting and creating data of various types (physical, social, infrastructural, etc.), structures and formats (quantitative vs. qualitative, point-based vs. grid-based, etc.), and considerations (data sovereignty, human subjects ethics, metadata standards, etc.). Efforts toward this goal include working with existing NCAR efforts to create a framework, principles, and best practices for cross-disciplinary data collection, integration, management, and use. 
5. Develop and deliver resources to guide those interested in convergence – In order to realize the utility of the project descriptions summarized above (and other Programmatic efforts), a variety of resources must be developed and made available for individuals, teams, and programs that are interested in conducting convergence. Efforts toward this goal include developing training materials, templates, documentation, and other resources in various formats and sharing them via various channels.