Abstract

Pattern scaling has proved to be a useful way to extend and interpret Earth system model (i.e. climate) simulations. In the simplest case the response of local temperatures is assumed to be a linear function of the global temperature. This relationship makes it possible to consider many different... Show morePattern scaling has proved to be a useful way to extend and interpret Earth system model (i.e. climate) simulations. In the simplest case the response of local temperatures is assumed to be a linear function of the global temperature. This relationship makes it possible to consider many different scenarios of warming by using simpler climate models to infer global temperatures and then translating those results locally based on the scaling pattern deduced from a more complex model. We expect the scaled pattern to be uncertain because the linear relationship for each model grid box is only determined by limited number of model experiments. In addition there will be spatial dependence among adjacent model grid boxes so the uncertainty in the scaling pattern must include this spatial correlation. This work explores a methodology using spatial statistics to quantify how the pattern varies across an ensemble of model runs. The key is to represent the pattern uncertainty as a Gaussian process with a spatially varying covariance function. We found that when applied to the NCAR/DOE CESM1 large ensemble experiment we are able to reproduce the heterogeneous variation of the pattern among ensemble members. Moreover much these local statistical computations are embarrassingly parallel and the analysis can be accelerated by parallel tools within the R statistical environment. Show less