Title: Earth systems science and engineering: Model evaluation and diagnostics
Abstract: Model building is a complex and intuitive process which is heavily influenced by perception, intuition, and prior knowledge on system functioning and reality and colored by mental concepts (state of mind). From a myriad of countless processes and mechanisms, the modeler seeks to elucidate those key principles, laws, and generalizations, which explain the observed data. Their translation to a computational model is difficult and subjective, particularly in the face of incomplete knowledge of the governing spatiotemporal processes and insufficient data on (spatially distributed) system properties and state variables. As the processor speed, storage ability and overall capabilities of computers have increased manifold in the past decades, so has the complexity of simulation models in an effort to handle, cope with and process increasingly larger volumes of data, and address scientific questions at increasingly finer spatial and temporal resolution. A large model complexity can have important drawbacks and frustrate the practical application of models for hypothesis testing, prediction, interpolation, and science- based decision making. In this talk, I will an overview of some of our recent work on these topics with applications in various disciplines of Earth systems science and engineering.