Meeting societal needs in weather, seasonal, and decadal prediction and climate projection requires a continuous improvement of the main tools used in making the predictions—global models of the Earth system. Despite significant progress in model development over the past few decades, several long-standing model systematic errors remain in most global models. This essay analyzes the model development process with the aim to identify a strategy to accelerate model development. It is argued that the main effort in doing so must focus on two main areas: i) improved diagnostic techniques that are aimed directly at identifying the key process involved in the major model errors and ii) a significant increase in the size of the currently too-small model development community through better collaboration of the academic community with modeling centers and through improving the image of the science of model development in the broader community. Success will strongly depend on the ability of bringing several communities together to work jointly in large national and international research programs.
School of Mathematical Sciences, Monash University, Clayton, Victoria, Australia