The improvements and validation of several parameterization schemes in the second version of the Beijing Climate Center Atmosphere–Vegetation Interaction Model(BCC_AVIM2.0) are introduced in this study. The main upda...The improvements and validation of several parameterization schemes in the second version of the Beijing Climate Center Atmosphere–Vegetation Interaction Model(BCC_AVIM2.0) are introduced in this study. The main updates include a replacement of the water-only lake module by the common land model lake module(Co LM-lake) with a more realistic snow–ice–water–soil framework, a parameterization scheme for rice paddies added in the vegetation module, renewed parameterizations of snow cover fraction and snow surface albedo to accommodate the varied snow aging effect during different stages of a snow season, a revised parameterization to calculate the threshold temperature to initiate freeze(thaw) of soil water(ice) rather than being fixed at 0°C in BCC_AVIM1.0, a prognostic phenology scheme for vegetation growth instead of empirically prescribed dates for leaf onset/fall, and a renewed scheme to depict solar radiation transfer through the vegetation canopy. The above updates have been implemented in BCC_AVIM2.0 to serve as the land component of the BCC Climate System Model(BCC_CSM). Preliminary results of BCC_AVIM in the ongoing Land Surface, Snow, and Soil Moisture Model Intercomparison Project(LS3 MIP) of the Coupled Model Intercomparison Project Phase 6(CMIP6) show that the overall performance of BCC_AVIM2.0 is better than that of BCC_AVIM1.0 in the simulation of surface energy budgets at the seasonal timescale. Comparing the simulations of annual global land average before and after the updates in BCC_AVIM2.0 reveals that the bias of net surface radiation is reduced from-12.0 to-11.7 W m-2 and the root mean square error(RMSE) is reduced from 20.6 to 19.0 W m-2;the bias and RMSE of latent heat flux are reduced from 2.3 to-0.1 W m-2 and from 15.4 to14.3 W m-2, respectively;the bias of sensible heat flux is increased from 2.5 to 5.1 W m-2 but the RMSE is reduced from 18.4 to 17.0 W m-2.展开更多
The growing demand for semiconductor devices simulation poses a big challenge for large-scale electronic structure calculations.Among various methods,the linearly scaling three-dimensional fragment(LS3DF)method exhibi...The growing demand for semiconductor devices simulation poses a big challenge for large-scale electronic structure calculations.Among various methods,the linearly scaling three-dimensional fragment(LS3DF)method exhibits excellent scalability in large-scale simulations.Based on algorithmic and system-level optimizations,we propose a highly scalable and highly efficient implementation of LS3DF on a domestic heterogeneous supercomputer equipped with acceler-ators.In terms of algorithmic optimizations,the original all-band conjugate gradient algorithm is refined to achieve faster convergence,and mixed precision computing is adopted to increase overall efficiency.In terms of system-level optimiza-tions,the original two-layer parallel structure is replaced by a coarse-grained parallel method.Optimization strategies such as multi-stream,kernel fusion,and redundant computation removal are proposed to increase further utilization of the com-putational power provided by the heterogeneous machines.As a result,our optimized LS3DF can scale to a 10-million sili-con atoms system,attaining a peak performance of 34.8 PFLOPS(21.2% of the peak).All the improvements can be adapt-ed to the next-generation supercomputers for larger simulations.展开更多
基金Supported by the National Key Research and Development Program of China(2017YFA0604300,2016YFA0602100,and2016YFA0602602)National Natural Science Foundation of China(41275075 and 91437219)
文摘The improvements and validation of several parameterization schemes in the second version of the Beijing Climate Center Atmosphere–Vegetation Interaction Model(BCC_AVIM2.0) are introduced in this study. The main updates include a replacement of the water-only lake module by the common land model lake module(Co LM-lake) with a more realistic snow–ice–water–soil framework, a parameterization scheme for rice paddies added in the vegetation module, renewed parameterizations of snow cover fraction and snow surface albedo to accommodate the varied snow aging effect during different stages of a snow season, a revised parameterization to calculate the threshold temperature to initiate freeze(thaw) of soil water(ice) rather than being fixed at 0°C in BCC_AVIM1.0, a prognostic phenology scheme for vegetation growth instead of empirically prescribed dates for leaf onset/fall, and a renewed scheme to depict solar radiation transfer through the vegetation canopy. The above updates have been implemented in BCC_AVIM2.0 to serve as the land component of the BCC Climate System Model(BCC_CSM). Preliminary results of BCC_AVIM in the ongoing Land Surface, Snow, and Soil Moisture Model Intercomparison Project(LS3 MIP) of the Coupled Model Intercomparison Project Phase 6(CMIP6) show that the overall performance of BCC_AVIM2.0 is better than that of BCC_AVIM1.0 in the simulation of surface energy budgets at the seasonal timescale. Comparing the simulations of annual global land average before and after the updates in BCC_AVIM2.0 reveals that the bias of net surface radiation is reduced from-12.0 to-11.7 W m-2 and the root mean square error(RMSE) is reduced from 20.6 to 19.0 W m-2;the bias and RMSE of latent heat flux are reduced from 2.3 to-0.1 W m-2 and from 15.4 to14.3 W m-2, respectively;the bias of sensible heat flux is increased from 2.5 to 5.1 W m-2 but the RMSE is reduced from 18.4 to 17.0 W m-2.
基金This work was supported by the National Key Research and Development Program of China under Grant No.2021YFB0300600the National Natural Science Foundation of China under Grant Nos.92270206,T2125013,62032023,61972377,T2293702,and 12274360+2 种基金the Chinese Academy of Sciences Project for Young Scientists in Basic Research under Grant No.YSBR-005the Network Information Project of Chinese Academy of Sciences under Grant No.CASWX2021SF-0103the Key Research Program of Chinese Academy of Sciences under Grant No.ZDBSSSW-WHC002.
文摘The growing demand for semiconductor devices simulation poses a big challenge for large-scale electronic structure calculations.Among various methods,the linearly scaling three-dimensional fragment(LS3DF)method exhibits excellent scalability in large-scale simulations.Based on algorithmic and system-level optimizations,we propose a highly scalable and highly efficient implementation of LS3DF on a domestic heterogeneous supercomputer equipped with acceler-ators.In terms of algorithmic optimizations,the original all-band conjugate gradient algorithm is refined to achieve faster convergence,and mixed precision computing is adopted to increase overall efficiency.In terms of system-level optimiza-tions,the original two-layer parallel structure is replaced by a coarse-grained parallel method.Optimization strategies such as multi-stream,kernel fusion,and redundant computation removal are proposed to increase further utilization of the com-putational power provided by the heterogeneous machines.As a result,our optimized LS3DF can scale to a 10-million sili-con atoms system,attaining a peak performance of 34.8 PFLOPS(21.2% of the peak).All the improvements can be adapt-ed to the next-generation supercomputers for larger simulations.
