Just a few short days before Earth Day, the April edition of Inside the Discovery Cloud focused on the Planet & Environment research area, with talks by Argonne’s Scott Collis and UChicago’s Elisabeth Moyer. Moyer, co-director of the CI’s Center for Robust Decision Making on Climate and Energy Policy (RDCEP) and assistant professor in the Department of Geophysical Sciences, works on both the experimental and the modeling sides of climate research, giving her a unique perspective on how models are constructed and used for science and policy.
In her talk, Moyer looks back at the motivation for forming RDCEP five years ago: to build decision-making tools that connect climate, energy, and economic models. The researchers started work in two seemingly opposite, but related directions -- creating more complex models for how human systems such as economics and policy will shape the path of climate change, and creating simplified representations of highly complex and detailed climate models for easier use in policy analysis.
The latter mission is accomplished through the development of emulators, tools that use the results of lengthy, computationally intense climate model runs to simplify the inputs into other types of models. The RDCEP team built a library of 70 terabytes of model output, and have fit the data to a simplified equation for predicting temperature and precipitation in different world regions under various carbon dioxide scenarios.
"We are now in a situation where we have more data than people are really processing fully," Moyer said. "We're in the realm where we need to use the tools of statistics, which were developed to make sense out of patterns in big datasets. We're no longer in the case where making a model run is a radically new invention and nobody had made this projection before. Now we have many projections and we need to make some sense out of what they say collectively."
You can watch video from Moyer's Discovery Cloud talk below. The next installment of the series, on May 21st, will cover Life Sciences.