Increasing Confidence in Regional Surface Temperature Response through Neutral Modes
Confidence in the regional features of climate change projections is a prized commodity needed for climate adaptation and mitigation. This study strives to decipher the pattern of the anthropogenically forced climate response using a dynamical mode-based framework to establish confidence in regional climate projections. The temperature pattern most susceptible to changes in external factors is known as the leading mode. It captures the evolution of the average global surface temperature over the past century in both observations and multi-model simulations. These results lend confidence to the regional climate change features that are captured by the leading mode.
This study offers a fresh perspective on regional climate change response and its associated feedbacks. The results show that certain regional climate response features are organized by well-established dynamical modes in the atmosphere-ocean coupled system, besides the inherent thermodynamic control of the temperature pattern. Differences in the energy entering and leaving the atmosphere, serving as radiative forcing affecting the climate in the midlatitudes, are most efficient at exciting the leading mode and global mean surface warming, with important implications for solar geoengineering.
Using a large suite of numerical experiments with forcing applied at every representative location of the globe, researchers construct the emergent dynamical operator for surface temperature and extract the leading mode. The leading mode represents the most excitable pattern of the climate system and it accounts for 56% of the variance of the surface temperature response due to the forcing exerted by doubling the carbon dioxide concentration. The pattern of the leading mode is partly organized by the atmospheric annular modes (atmospheric patterns of climate variability that influence surface temperature and precipitation) in both hemispheres and is further modulated by feedbacks from ocean circulations. Though derived from a single model, the leading mode can capture spatio-temporal variabilities of global mean surface temperature in both Coupled Model Intercomparison Project Phase 6 multi-model simulations and observations for the past century. Moreover, the radiative forcings from the midlatitude bands centered around 45°N and 25°S efficiently excite the leading mode. Most human activities take place around these midlatitude bands, suggesting the potency of human influence on climate change.