Simulation Contrast And Assessment Of Global Ice Sheet Changes During The Last Glacial Maximum

The Paleoclimate Modelling Intercomparison Project Phase III(PMIP3)developed a standard experiment of the Last Glacial Maximum(LGM),and 18 research teams around the world used the different climate system models to run the experiment.Four of these models simulated the global surface temperature,precipitation/snowfall,evaporation rate,snowmelt,and snow depth during the LGM.In this paper,the surface mass balance during the LGM simulated by these four models was calculated using the degree-day model and the melting model,compared with the snow depth and melting rate simulated by each model,and the reconstructed ice sheet during the LGM.This can judge whether the correct surface material balance can be simulated by the optimized model.At the same time,parameter inversion is performed for the degree-day model and the melting model to infer the influencing factors behind the errors generated by each model simulation.In order to compare the advantages and disadvantages of the global ice sheet simulation during the LGM,the model's ability to simulate the climate of the LGM was compared to provide a reference for the future selection of suitable models for relevant research.This study has important scientific significance for improving the accuracy of climate simulation during the LGM and better understanding the formation of glaciers during the LGM and even global climate change.In this paper,the global surface temperature obtained by simulation of each model is used,and then the positive degree days(PDD)of each model during the LGM is calculated by the degree-day model.Then,the PDD was introduced into the melting model,and the glacial surface mass balance of each model during the LGM was obtained in combination with the simulated precipitation/snowfall.Compare the results of each model and combine the reconstruction data,it was found that all the models had excessive snow accumulation,and formed glaciers in the Qinghai-Tibet Plateau.This is inconsistent with the results obtained by reconstructing the data.In addition,some models have errors in the simulation of other regions.The above results indicate that the climate model still needs many improvements to accurately simulate the climate rules of the ice age.Secondly,in order to display and compare the simulation results intuitively,,this paper creatively designs a grading rule based on the relevant conditions for the formation of the ice sheet,combined with the reconstruction data of the LGM,and compare surface mass balance results and simulated snow depth results.It is found that the CCSM4 model is superior to the simulations of snow depth changes over the Qinghai-Tibet Plateau during the LGM.CNRM-CM3 model underestimate the surface mass balance in actual ice sheet areas in North America,Northern Europe,etc.At the same time,there are still errors in the simulation of surface mass balance on the Qinghai-Tibetan Plateau.The above results also show that the current climate model still has limitations.Thirdly,because the parameter settings for calculating the surface mass balance for each model are all default values,or they are estimated based on modern climate,there may be some errors.Therefore,in order to obtain the best melt rate value,we performed parameter inversion on the melting model.Then,the surface mass balance obtained under each parameter was fitted to the simulated snow depth trend to obtain the best parameter value of the melting model.At the same time,the surface temperature of the LGM was simulated using the CESM model to obtain the degree-day model parameter values that are more in line with the actual conditions in each study area.Fourth,according to the third step,after getting the best melting rate value and surface mass balance for each model,compared with the simulated average annual precipitation/snowfall and melting rate from models at the same time,in order to determine if model optimization is effective,and analyze whether the root cause of the error is caused by the surface temperature or snowfall.The results showed that the optimization of the model is effective.The simple model can simulate the material balance changes in the ice sheet area simulated by the complex model.Moreover,all models had errors in the simulation of temperature or snowfall in the Qinghai-Tibet Plateau except CNRM-CM5.In the CNRM-CM5 model,the simulated surface mass balance is too small due to the high simulated surface temperature,cause larger errors.In the North American ice sheet area,there are different degrees of abnormality in each model.Among them,the CNRM-CM5 and FGOALS-g2 models is caused by high surface temperature,and the CCSM4 is caused by a big amount of snowfall.In the mountains,such as the Rocky Mountains,there are also significant differences in the climatic characteristics of the simulated models.