CICS Research Themes

Earth System Modeling and Analysis

Climate modeling at Princeton University and GFDL is continually producing new models, including atmospheric, oceanic and land models, coupled models, chemistry-radiative forcing models, cloud resolving models with new microphysics, and a non-hydrostatic limited area model. These models may, in principle, be appropriately combined to give what might be called an Earth System Model, or ESM.  Such models, by definition, seek to simulate all aspects — physical, chemical and biological — of the Earth system in and above the land surface and in the ocean. Thus, an Earth System Model consists of, at least:

1. An atmospheric general circulation model, including a dynamical core for the fundamental fluid dynamics and water vapor, a radiation scheme, a scheme for predicting cloud amounts, a scheme for aerosols, and various parameterization schemes for boundary layer transport, convection and so forth.
2. An oceanic general circulation model, including a dynamical core, various parameterization schemes for boundary layers, convection, tracer transport, and so on.
3. An ice dynamics model, for the modeling and prediction of sea ice.
4. An atmospheric chemistry module, for predicting chemically active constituents such as ozone.
5. A land model, for land hydrology and surface type, and a land ice model.

6. Biogeochemistry modules for both land and ocean. These may be used, for example, to model the carbon cycle through the system.
7. A computational infra-structure to enable all these modules to communicate and work together efficiently.

 The goal of Earth System Modeling development at CICS and GFDL is, then, to construct and appropriately integrate and combine the above physical and biogeochemical modules into a single, unified model. Such a model will then be used for decadal to centennial, and possibly longer, studies of climate change and variability (as described primarily in the ‘applications’ section). At present, such a model does not exist in final form, and improvements are needed in two general areas:

1. Improvement on the physical side of the models, in the ocean, atmospheric, sea ice and land components. 
2. Further incorporation of biological and chemical effects into the model, and ecosystem modeling.

Both of these are continual processes, that can never be said to be complete, although at various stages the development of a component, or a complete model, may be ‘frozen’ to allow numerical experiments to take place in a stable environment.

Developing and testing such ESM is an enormous task, which demands a significant fraction of the resources of CICS and GFDL.  Further, at any given time, model development depends on existing knowledge of how systems behave, but for that development to continue our body of knowledge and understanding must also increase correspondingly, and without that, model development would stagnate. That is to say, one might regard ESM development as both a scientific and engineering enterprise, and proper attention and respect must be paid to both aspects. The contributions and goals of CICS might be divided into the following two general areas:

1. Development of modules (or components) for the Earth System Model (for example, the land model, parts of the ocean model), in collaboration with GFDL scientists.  Note that not all of the modules above involve CICS scientists; rather, CICS complements rather than duplicates GFDL efforts.  For example, the dynamical core of the atmospheric model, the ice model, and many of the physical parameterizations of the atmospheric model, have been and will continue to be developed at GFDL itself. This document focuses on those components to which CICS is directly contributing.
2. Seeking improved understanding of the behavior of components of the Earth system, and the interaction of different components, thereby aiding in the long-term development of ESMs.

These aspects involve comparisons with observations, use in idealized and realistic situations, and development of new parameterizations and modules.  The development of ESMs is a research exercise, and is crucially dependent on continually obtaining a better understanding of the ocean-atmosphere-ice-land system. We begin with the computational framework, into which the other components fit. We then discuss the development of some of the various components, and then return to the problem of understanding and modeling the system as a whole.