DescriptionBack to top
Climate models are important tools for understanding past, current and future global climate variability, yet they exhibit key uncertainties that limit their applicability to fine scale analysis and future projections. Some key sources of uncertainty include coarse grid resolution, inadequate representation of relevant physics and interactions, overfitting from downscaling and bias-correction, lack of observations to calibrate and evaluate models, uncertain model parameters, different model structures, and so on. In addition, coupled climate models are computationally expensive and thus difficult to use for uncertainty analysis, while reduced complexity models are fast and flexible but are highly parameterized and lack physics. These computational tradeoffs pose major challenges for evaluating/comparing model results, constructing reliable projections, and quantifying relevant uncertainties. The workshop will bring together researchers from multi-disciplinary fields to highlight new math/stat methods for climate model evaluation and uncertainty quantification across spatial and temporal scales, and to advance our understanding about the physical processes leading to model errors, biases, and uncertainty.