A0773
Title: Multi-model ensemble analysis with neural network Gaussian processes
Authors: Trevor Harris - Texas A&M University (United States) [presenting]
Abstract: Multi-model ensemble analysis integrates information from multiple climate models into a unified projection. However, existing integration approaches based on model averaging can dilute fine-scale spatial information and incur bias from rescaling low-resolution climate models. A statistical approach, called NN-GPR, is proposed using Gaussian process regression (GPR) with an infinitely wide deep neural network-based covariance function. NN-GPR requires no assumptions about the relationships between climate models, no interpolation to a common grid, and automatically downscales as part of its prediction algorithm. Model experiments show that NN-GPR can be highly skilful at surface temperature and precipitation forecasting by preserving geospatial signals at multiple scales and capturing inter-annual variability. The projections particularly show improved accuracy and uncertainty quantification skill in regions of high variability. This allows cheaply assessing tail behaviour at a 50 km spatial resolution without a regional climate model (RCM). Evaluations on reanalysis data and SSP2-4.5 forced climate models show that NN-GPR produces similar overall climatologies to the model ensemble while better capturing fine-scale spatial patterns. Finally, NN-GPR's regional predictions are compared against two RCMs, and it is shown that NN-GPR can rival the performance of RCMs using only global model data as input.