[U-02] Projection and detection of global environmental change

[U02-P01] Forced and unforced pattern effects in the climate feedback

*Tomoki Saga¹, Masahiro Watanabe¹, Hiroaki Tatebe², Hideo Shiogama³ (1.Atmosphere and Ocean Research Institute, The University of Tokyo, 2.Japan Agency for Marine-Earth Science and Technology, 3.National Institute for Environmental Studies)

Keywords:Climate feedback, Equilibrium climate sensitivity, Radiative forcing, Pattern effect, Ocean heat uptake, Internal variability

When the radiative forcing of increase in atmospheric CO₂ concentration is applied to the climate system, some of it is absorbed by the ocean and the rest of the energy is thrown away into space through temperature rise. This is the simplest mechanism of global warming. The energy injection determined only by the surface temperature assuming the uniform change in vertical air temperature is the Plank response, but the climate feedback parameter (λ), how much energy is emitted per 1 K temperature rise, is determined by changes in water vapor, lapse rate, clouds, surface albedo, and others. Equilibrium climate sensitivity (ECS) is the change in global mean surface temperature (GMT) when CO₂ is doubled and the system reaches equilibrium; ECS is used as a basic indicator of global warming. Understanding feedback is important in the mechanism of global warming, since ECS is defined by the ratio of forcing to λ. λ was assumed to be constant, but recent research has indicated that its change depends on the surface warming pattern. This process is called pattern effect, but it is not well understood how the pattern effect occurs in the past and future warming responses. In this study, the forced pattern effect, the change in the response to forcing, and the unforced pattern effect, which is caused by internal variability in the climate system, by using a large-scale ensemble simulation of a single climate model. Furthermore, by examining the relative importance of both effects on climate change from the past to the end of this century, we aim to obtain knowledge that will help us understand the uncertainty of future climate change predictions.
We analyze the 50-member historical experiment (1850-2014) and the SSP2-4.5 global warming scenario experiment (2015-2100), a medium emission scenario, by the global climate model MIROC6 participating in the Coupled Model Intercomparison Project Phase 6 (CMIP6). The forced response is defined by the ensemble mean, and the anomaly of each member from the ensemble mean is the variability component. Forcing from 1850 to 2100 is obtained from the calculation results of the same model in the Radiative Forcing Model Intercomparison Project (RFMIP). We estimate the climate feedback parameter by applying the global energy budget equation to each member, while the forced component of feedback parameter is obtained by applying the energy balance equation to the ensemble mean. Feedback due to variability is obtained as their difference. λ from the atmospheric model experiment amip-piForcing driven by the observed sea surface temperature (SST) during the historical period is also used.
From 1970 to 2014, the effective climate sensitivity (EffCS) from amip-piForcing is 1.92 ℃, which is lower than the value (2.57 ℃) from the MIROC6 abrupt CO₂ quadrupling experiment (4xCO2). It is consistent with past studies in which the pattern effect intensifies negative feedback in this period. Historical EffCS ranges wide and both values are within the ensemble range (Fig. 1). It should be noted that the EffCS from amip-piForcing is placed at the edge of the ensemble, but this means that changes in feedback based on observed SST can occur within one realization of ensemble. Since the 1980s, the forced feedback is almost equal to λ from 4xCO2. This suggests that it is more reliable to use recent data when estimating feedback.
We investigate the GMT anomaly in 2014 against the 1860-1879 average. The observed value of HadCRUT5 is +0.99 ℃, but in MIROC6 it is +0.83±0.17 ℃ (range is one standard deviation). The contributions of forcing, ocean heat uptake (OHU), and feedback to the temperature anomalies of the ensemble mean are +0.52, −0.18, and +0.43 ℃, respectively. However, the contributions of OHU and feedback to the spread among members are positively correlated (r = 0.58), and thus the spread of the temperature anomaly increases. This means that the unforced pattern effect due to variability occurs closely related to OHU.