NCAR Hosts STATMOS/SAMSI Workshop

By Eliott Foust
08/10/2017 - 1:15pm

Nearly 50 atmospheric science and statistics researchers and graduate students from across the United States gathered for a week of coding and collaboration during the Research Network for Statistical Methods for Atmospheric and Oceanic Sciences (STATMOS)/Statistical and Applied Mathematical Sciences Institute (SAMSI) Workshop on Climate Statistics in Boulder, Colorado. Instead of solely offering lectures and instruction, this workshop resembled a hackathon, and focused on group projects and collaboration.

STATMOS/SAMSI participants work together on their laptops. During the workshop, participants gathered into interdisciplinary groups of two to four people throughout the Daemon Room. Each group used its members’ knowledge of atmospheric science or statistics to work towards solving a current research problem, such as agreement between remote and in situ data sets, and the probability of experiencing temperature and precipitation extremes. The goal of the workshop was not to solve individual problems, but to create a framework for interdisciplinary collaboration that could lead to solutions of such problems in the future.

This year’s workshop was co-organized by University of Chicago professors, Michael Stein and Noboru Nakamura, University of North Carolina-Chapel Hill professor, Dr. Richard Smith, and NCAR scientists Dorit Hammerling, Doug Nychka, and Jean-Francois Lamarque. The primary source of funding came from STATMOS and SAMSI, who are in turn supported by the National Science Foundation. These statistical research institutes focus on formulating collaborative networks of statisticians and scientists to solve problems by utilizing applied statistics. According to the organizers, one of the inherent difficulties in arranging a workshop such as this is the language and knowledge barrier that exists between the two different academic fields of atmospheric science and statistics. Participants must first develop ways to understand each other’s nomenclature, perspectives, and interpretations of physical parameters and statistics.

Before the start of this year’s workshop, participants submitted project ideas to a Google Drive folder, then organized themselves into groups to collaborate on the accepted projects. According to University of Chicago Ph.D. student, Jonah Bloch-Johnson, the workshop put statisticians and atmospheric scientists in close proximity where it is more engaging to share ideas and learn the way others think, which helps them overcome some of the language and knowledge barriers. Unlike traditional collaborative efforts that are done remotely, Bloch-Johnson says, “it helps when you are working elbow to elbow with each other.”

Looking forward, the workshop organizers plan to assess the feasibility of a second workshop like this one, and participants were encouraged to continue their work and report on any publications that arise from the workshop.