Lake Expedition 2020: The GLEON Fellowship Program (GFP) has accepted nine fellows for Lake Expedition 2020, which is a team science training program offered through the Global Lake Ecological Observatory Network (GLEON; gleon.org). The GFP trains cohorts of graduate students to use the rich information content of large and diverse data sets, operate effectively in diverse international teams, and communicate outcomes to a broad range of audiences. In 2020-2022, the GFP is partnering with the NSF-funded project, Harnessing the Data Revolution (HDR), to apply machine learning techniques to aquatic problems.

Our team is active in creating programs, courses, and a training environment with a special focus on integrating the research results of KGML and its applications in computing, ecology, hydrology, climate/weather and translational biology. Our goal is to attract and prepare the next generation of the workforce by bringing a convergent mindset to science and engineering research. More specifically, to produce students with technical depth and an integrative, transdisciplinary understanding of the core theoretical concepts in machine learning and how those concepts relate to relevant application domains, Kumar, Zhang, Ebert-Uphoff, and Steinbach will integrate the research results of this project as real-world projects into their data mining courses and workshops, and Janes, Barnes, Duffy, Dugan, and Hanson will integrate the machine learning tools developed in this project in their disciplinary courses. Recently, Kumar has designed and will teach a new course on AI for Earth, which focuses on the use of ML for analyzing the Earth system. The results of this research will be used to enhance this course. Similarly, Ebert-Uphoff and Barnes have recently formed a machine learning club in the atmospheric science department at CSU, into which KGML will be integrated. Furthermore, the team will support the researchers with convergence education through experiential learning such as collaboration, embedding, integration, deployment, etc. and learn about team science approaches and interdisciplinary habits of the mind. To gain a significantly more diverse student body in the related STEM degree programs and to leverage all of the benefits that diversity brings to our research and education programs, the team will grow existing and initiate new activities focused on recruiting, mentoring, and retaining women and other groups underrepresented in the related fields. Through the aforementioned activities, KGML will train a diverse group of researchers to:

1. inspire innovative thinking within the workforce,

2. transition fluently across disciplinary boundaries,

3. integrate machine learning tools to solve problems important in society, and

4. lead by sharing advances through publications, community outreach, teaching, or product commercialization.

Results of our work will be disseminated through five workshops to further advertise our work to the relevant communities to ensure that any advances are brought to the attention of those who can most benefit from them. In addition, our KGML results and code will be freely available use on our webpage for the wider community. Our team has a strong history of providing their software for public download. Providing the KGML code will allow us to reach out to and receive valuable feedback from the involved domain sciences. Curated data sets will also be available on the website to serve as benchmarks for the ML community to encourage the development of novel KGML algorithms. We also have a strong track record of publishing scientific data, e.g., a public ocean eddies database and a web-viewer for visualizing water bodies around the globe. Additionally, two of the team members serve on the steering committee of the Research Coordination Network on IS-GEO funded by NSF’s EarthCube initiative, a community-wide effort to develop a cyberinfrastructure for integrating and sharing data, analysis methods, and workflows that will accelerate scientific knowledge discovery in the Geosciences. This effort will be leveraged in this project. Finally, we will leverage the connections of our team members to researchers from federal agencies and academia: USGS (UW and UMN), NCAR (CSU), NOAA (CSU), and the Biocomplexity Institute (UVA).