Climate Modeling

The climate modeling program at USC focuses on blending data and models to improve our knowledge of the low-frequency evolution of Earth's climate, with a special focus on data-model comparison over the Common Era (the past 2,000 years).

Statistical Modeling

One avenue of research seeks to reconstruct past climate behavior by extracting statistical information from the recent geologic past. We use probabilistic models of climate fields to infer their variability from observed fluctuations in climate proxies. In collaboration with statisticians, the JEG lab is developing more accurate and efficient covariance estimation techniques to be able to detect robust climate signals in proxy data from corals, ice cores, speleothems, lake and marine sediment cores, and tree ring records. This approach aims to:

  • produce a better depiction of Earth's recent natural variability to provide context for the detection and attribution of the recent anthropogenic influence on climate.
  • translate proxy records into a language understandable by physical scientists and numerical models.

Climate-Proxy Modeling

By nature, climate "proxies" only indirectly record climate signals. What exactly are they telling us? A great number of them take the form of water isotopes (ratios of the stable isotopes of oxygen and hydrogen), sensitive to both thermal and hydrological processes. On the other hand, global climate models (GCMs) speak a language rather foreign to proxies because their output is expressed in quantities such as temperature, rainfall, wind stress, and so on. How can this gap be bridged?

We are currently developing forward models of climate proxies, which explicitly incorporate physical, chemical and (in some cases) biological processes giving rise to geological proxy formation. This amounts to translating GCM language into proxy language. The ultimate goal of this dual approach is to blur the distinction between proxy data and climate fields, so as to enable the geologic record to quantitatively constrain climate models, with the hope of narrowing down the uncertainty in future climate projections.

Dynamical Modeling

Two general circulation models are currently in use at USC:

  • The ICTP AGCM (SPEEDY)
  • isoGSM

Our climate group is currently looking for a creative climate modeler to join the faculty and bring new climate modeling tools to USC. A postdoctoral position in climate modeling is also available in JEG's group. Two postdoctoral positions are available in Stott's group.

  • Zumberge Hall of Science (ZHS)
  • Los Angeles, CA 90089-0740
  • Phone: (213) 740-6106
  • Email: earthsci@usc.edu