A New Paradigm for Principles-Based Adept Predictions of Global Warming from Climate Mean States

Distinguishing anthropogenic warming from natural variability in observations and reducing uncertainty in global warming projections by climate models remain critical challenges for climate scientists. In our work, we present a groundbreaking principle-based framework for adept predictions of the global mean warming and its spatial pattern in response to external energy perturbations from climate mean states without running climate models or relying on statistical trend analysis. When applied to the observed climate mean state of 1980-2000, our framework accurately predicts the global warming observed from 1980-2000 to 2000-2020, with a prediction of 0.403 K compared to an observed value of 0.414 K. Our predictions from the preindustrial mean climate states of individual CMPI6 models yields a similar median global mean warming to their warming projections under the abrupt 4×CO 2 scenario (5.4 K versus 5.6 K), but with noticeably less inter-model spread. Our predictions not only exhibit a high map correlation skill comparable to that of each individual CMIP6 model for the observed warming, but also capture the temporal pace of their warming projections under 1% annual CO 2 -increasing scenario. The success of our global warming prediction from climate mean states under any given scenarios of CO 2 increasing without running climate models is built on the following two innovations. One of them is our recent discovery of nature’s climate feedback 'circuit' associated with temperature feedback. The other is that we extract the information of surface energy amplification by non-temperature feedback from the ratio of downward longwave energy emission from the atmosphere to solar energy absorbed by the surface in climate mean states. The product of amplification rates by temperature and non-temperature feedbacks corresponds to the total amplification of energy perturbations at the surface caused by CO 2 increasing. Temperature feedback yields a 6.6-fold global mean amplification, while non-temperature feedback contributes an another 2.3-fold amplification. This results in a total 15.8-fold increase in the global mean of CO 2 -induced energy perturbations at the surface. The energy balance between the enhanced thermal emission from the surface and the total external energy input perturbations amplified by temperature and non-temperature feedback determines the global surface warming in response to CO 2 increasing.