In climate modeling, speed matters

ARM data let scientists continuously evaluate whether particular parts of climate models show systematic biases over an extended period and range of conditions, Liu says. “But so far model evaluation using such surface-based observations has relied largely on case studies or intensive observational periods.” Highly detailed data from long-term observational programs such as ARM gathered over the past decade have been underused to evaluate climate model performance, he adds. The FASTER project is driven by the Earth System Modeling Program of DOE’s Office of Biological & Environmental Research.

Liu says understanding fast processes, developing parameterizations for them and examining their effects on climate sensitivity should involve a chain of activities in areas ranging from observation to model evaluation to theoretical understanding of the physical processes and corresponding parameterizations. But following the chain means getting scientists in different areas of expertise to work in concert to overcome their “cultural differences.”

Previous activities “suggest to me a need for a more focused project that capitalizes on all of the developments in both measurements and model evaluation,” Liu says.  “FASTER is a comprehensive project that integrates expertise and will use the available continuous observations to perform fast evaluation of global climate model performance. We think that it will greatly advance climate modeling.”

Modeling cloud and precipitation processes is complex because they involve multibody systems, with numerous particles of different sizes and shapes, and occur over a wide range of time and space scales, Liu explains. He calls them “4M complexities” – multibody, multiscale, multitype and multidimension.

FASTER’s first order of business is to evaluate and delineate the individual fast processes. Because fast physics parameterization is a major problem in global climate simulation, any improvement in fast physics will lead to better models and, eventually, a better understanding and prediction of such crucial phenomena as global warming.

“Virtually all of the fast physics processes interact,” Liu says, “so once we get a handle on the individual processes, we want to see how they interact and how to evaluate these interactions.”

FASTER’s testbed aspect integrates two major “fast” components: a single-column model (SCM) testbed that takes advantage of SCM versions of a global climate model; and a numerical weather prediction model (NWP) testbed that capitalizes on routinely reported results from major NWP research centers. The testbeds allow researchers to evaluate fast physics parameterizations quickly. Presently, the researchers are focused on the SCMs of the NASA Goddard Institute for Space Studies, the NOAA Geophysical Fluid Dynamics Laboratory, the National Center for Atmospheric Research and the European Center for Medium-Range Weather Forecast.

Liu says FASTER also involves a suite of higher resolution modeling activities – cloud-resolving models and large eddy simulations, for example – to enhance the SCM and NWP activities. There are plans to have a platform for running the Weather Research and Forecasting model continuously for the period needed to evaluate models.

“A unique aspect of FASTER is that its main objective and task is to evaluate models, which requires not only storage of observational data but also simulation data derived from models of various resolutions,” Liu says. “To put these two streams of data in the same format and at different resolutions in a ready-to-use mode is a huge challenge. Data integration is an essential part of the project, and we will work hard to explore innovative ways for data mining and visualization.”