Analysis Of Climate Feedbacks Over East Asia Using The Radiative Kernel Method

East Asia is one of the key regions most sensitive to climate change due to its high population density and rapid economic development.Under global warming,East Asia has experienced rapid warming,and it will warm with a stronger degree than the global mean.Previous studies have suggested that climate feedbacks are of vital importance to accurately simulate and predict the surface air temperature changes over East Asia,but few quantitative research has focused on East Asian climate feedbacks.The radiative kernel method is an efficient method for quantifying climate feedbacks,which decomposes individual climate feedback into the radiative kernel and the climate response term.In this study,we develop a new set of radiative kernels based on multi-year satellite observations,then we systematically and quantitatively analyze the climate feedbacks over East Asia by using the new kernels combined with observations and the Coupled Model Intercomparison Project Phase 6（CMIP6）simulations.The main conclusions are as follows:（1）A new set of surface temperature,atmospheric temperature,water vapor,surface albedo,and cloud radiative kernels is constructed based on BCC＿RAD radiative transfer model and 16-year satellite observations,and these new kernels can well reproduce the clear-sky and all-sky radiation budget at the top of the atmosphere.The short-term global mean Planck,lapse rate,tropospheric water vapor,stratospheric temperature and water vapor,and surface albedo feedbacks in response to the observed short-term climate variations during2000–2019 derived from the new kernels are-3.27,-0.20,1.57,0.12,and 0.86 W m^-2 K^-1,respectively.The short-term global mean cloud feedback derived from the adjusted cloud radiative forcing method based on noncloud radiative kernels(0.24 W m^-2 K^-1)is larger than that derived from cloud radiative kernels(0.19 W m^-2 K^-1).The choice of radiative kernels influences the feedback estimation,especially the land Planck feedback in the Northern Hemisphere as well as the surface albedo feedback and cloud feedback in the Arctic and the Southern Ocean.（2）Short-term local cloud feedback over East Asia is estimated by using the newly developed cloud radiative kernels applied to observations during 2000–2019.The magnitude of short-term cloud feedback over East Asia is 0.72 W m^-2 K^-1,mainly due to the contributions of nimbostratus and stratus in the East Asian monsoon region（EAMR）.The positive cloud feedback over East Asia in each season is mainly located in EAMR.The cloud feedback in EAMR is dominated by marine stratus in spring,land deep convective cloud in summer,land nimbostratus in autumn,and nimbostratus and stratus in winter.The cloud feedback over East Asia is chiefly due to the decrease in middle and low cloud fraction in EAMR,which is caused by the reduced relative humidity（RH）and the weakened ascending motion.（3）Short-term local climate feedbacks over East Asia are analyzed by using the newly developed noncloud kernels applied to observations and CMIP6 Atmospheric Model Intercomparison Project（AMIP）simulations during 2000–2014.Over East Asia,observations suggest that lapse rate feedback makes the largest contribution to local warming,while multi-model means suggest that cloud feedback is the largest contributor to local warming.Compared to the observations,Planck and lapse rate feedbacks are underestimated by 0.03and 0.24 W m^-2 K^-1,respectively,while tropospheric water vapor,stratospheric temperature,and cloud feedbacks are overestimated by 0.17,0.07,and 0.11 W m^-2 K^-1,respectively.The inter-model spread of cloud feedback is the largest for East Asian local short-term climate feedback processes.The simulation biases and inter-model spread of cloud feedback are probably due to the simulated cloud fraction responses of cirrostratus,deep convective cloud,nimbostratus,and stratus.（4）The relationship between long-term and short-term local cloud feedbacks over eastern China is analyzed by using the newly developed noncloud kernels applied to CMIP6simulations.The long-term cloud feedback derived from the quadrupling of CO₂ runs ranges from 0.40 to 3.31 W m^-2 K^-1,while the short-term counterpart derived from the historical runs ranges from 0.15 to 4.77 W m^-2 K^-1.The inter-model spread of cloud feedback at both timescales is mainly due to the inter-model spread of the responses of mid-low cloud cover and liquid water path（LWP_a）within ascent regions.The inter-model correlation between long-term and short-term LWP_a responses,likely due to the model-specific sensitivities of LWP_a to lower-tropospheric RH,is primarily responsible for the strong inter-model correlation of cloud feedbacks across timescales.