There is no single best way for a model to represent elements of the climate system such as clouds and ocean mixing, and climate models behave differently depending on the representations they use. As a result each model has its own strengths and weaknesses, for instance a model could produce a very good simulation of Arctic sea ice but generate El Nino events which are too strong and too frequent. The mixed quality of climate model simulations poses a challenge for their use in understanding the climate system, predicting short-term climate variability, and planning for the consequences of climate change.
Research under this award pursues a novel strategy to take advantage of the strengths of different climate models while minimizing the effects of their weaknesses. The strategy involves running a group of climate models interactively, so that each model influences the behavior of the other models. Previous research shows that interacting models tend to synchronize their behavior, so that a group of models run interactively behaves like a single "supermodel". Earlier results also suggest that such a supermodel can be trained using real-world observations so that its simulation is better than the simulations produced by any of its constituent models.
The supermodel developed in this project combines the Community Earth System Model (CESM) developed at the National Center for Atmospheric Research, the Max Planck Institute Earth System Model (MPIESM), and the Norwegian Earth System Model (NorESM). The work focuses specifically on the ability of the supermodel to improve representation of upper-level jet streams and blocking highs. The extent to which improved representation of these features in turn improves the supermodel's ability to simulate extreme precipitation in the middle latitudes is examined. Assuming successful results in the present-day climate, the supermodel will be used to investigate changes in circulation features and associated precipitation extremes in warmer climates produced by greenhouse gas increases.
The work is of societal as well as scientific interest given the potential of the supermodeling strategy to improve predictions of weather and climate variability and projections of climate change. Blocking is a particular challenge for weather and climate models and poor performance in blocking adversely affects predictions of precipitation extremes. The work also promotes international collaboration in geoscience as it is conducted in partnership with the Norwegian Synchronisation to Enhance Reliability of Climate Predictions (STERCP) project. The project also supports a postdoctoral research associate, thereby developing the next generation workforce in this research area.
This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.
