Simplification of Largescale Models to Guide Process Understanding

Over the last decade, the climate modeling community has increased the complexity of models to address critical scientific questions by adding more processes or components (e.g., aerosols, chemistry, carbon cycle, dynamic vegetation) and increasing the model resolution. It appears, however, that this did not yield a better understanding of cloud-climate feedbacks, because the diagnosed uncertainty in climate sensitivity has not decreased with time. Large-scale models appear most useful not only as a basis for the quantification of climate sensitivity but also as a framework for advancing ideas and understanding on how the climate works and responds to external perturbations. Such models can be used to develop concepts, which in turn must then be explored, or isolated, through simplifications.

A better understanding of the physical mechanisms underlying the cloud-climate feedbacks produced by climate models could be useful in designing a strategy to evaluate these feedbacks using observations. By simplifying (rather than complicating) models and conducting idealized experiments, we may be able to pinpoint the main critical processes, to prioritize them, and to test resulting ideas or theories. Interpretation frameworks could then be proposed to assist our understanding of the physics of and intermodel differences in cloud-climate feedbacks.

One approach to simplify GCMs may be to reduce the complexity of the large-scale boundary conditions (e.g., aquaplanet versions of the models, even for CRMs or super-parameterizations), to reduce the dimensionality of the system (e.g., 2-D or 1-D model versions derived from the 3-D model), or even to remove some processes (i.e., replace complex microphysical schemes with simpler ones). Models of intermediate complexity, such as the quasi-equilib-rium tropical circulation model (Neelin and Zeng 2000), might also be used. Simple conceptual models (e.g., 2-box models) may be viewed as the ultimate step to this simplification process.

The extent to which simplified models are useful in reproducing and interpreting complex model results should be tested and quantified by analyzing, for a given model, how the cloud-climate feedbacks compare across the hierarchy of model complexities. An ideal situation would be for each GCM to be associated with a suite of simpler or more idealized model versions to support the analysis and the understanding of its results.

0 0

Post a comment