Pattern-Scaling-Based Projections for Precipitation and Potential Evapotranspiration: Sensitivity to Composition of GHGs and Aerosols Forcing

Pattern scaling is a computationally efficient method to generate global projections of future climate changes, such as temperature and precipitation, under various emission scenarios. In this study, we apply the pattern scaling method to project future changes of potential evapotranspiration (PET), a metric highly relevant to hydroclimate research. While doing so, this study tests the basic assumption of pattern-scaling methods, which is that the underlying scaling pattern is largely identical across all emission scenarios. We use a pair of a the large-ensemble global climate model (GCM) simulations and obtain the two separate scaling patterns, one due to greenhouse gases (GHGs) and the other due to aerosols, which show substantial regional differences. We also derive a single combined pattern, encapsulating the effects of both forcings. Using an energy balance climate model, future changes in temperature, precipitation, and PET are projected by combining the separate GHGs and aerosols scaling patterns (“hybrid-pattern” approach) and the performance of this “hybrid-pattern” approach is compared to the conventional approach (“single-pattern”) by evaluating both approaches against the GCM direct output. We find both approaches provide reasonably good emulations for the long-term projection (end of the 21st century). However, the “hybrid-pattern” approach provides better emulations for the near-term climate changes (2020-2040) when the large changes in aerosol emissions occur.