| The Mid-Pleistocene Transition(MPT,1.2 Myr to 800 kyr BP)is an important period in the Earth’s climate history,when the global temperature dropped,the ice volume gained,and the sea level fell,the glacial cycle shifted from the previous regular 41 kyr to current 100 kyr.The causes of the MPT is a crucial paleoclimate open question,the studies on it can complete the Pleistocene global climate and ice-sheet record,thereby reducing the uncertainty of predicting the future climate and ice-sheet evolution by modeling,and improving the understanding of the Earth’s climate system.Antarctic deep ice cores are likely to hold the direct record of greenhouse gases covering the MPT,which can provide the scientific basis for the causes of the MPT.Therefore,the International Parternerships in Ice Core Sciences(IPICS)identified the acquisition of1.5-Myr ice cores covering MPT in Antarctica as one major goal of Antarctic exploration.Drilling deep ice cores is time-consuming and technically difficult,thus locating ice cores is very important.Potential locations for old ice cores should meet two conditions.Firstly,the maximum age of the ice should be older than 1.5 Myr.Secondly,cold base is expected for avoiding melting and destroyed climate record in the basal ice.Therefore,the research on age of ice and ice temperature can provide important information for locating old ice cores.The Dome Fuji(Dome F)region in the East Antarctic is potential for holding old ice cores,there is a deep ice core dated back to 720 kyr BP.The understanding of the age and temperature of the ice sheet is also limited here.Ice-flow modelling is a common way to model the ice sheet parameters and reconstruct age and temperature field of ice.Ice penetrating radargram reveals the inner structure of the ice sheet,which can provide a strong constraint for the ice-flow modeling and thus improve the reliability of the model results.In this research,we used ice-penetrating radar observations as the constraints of a 1-D ice-flow model and a 3-D ice-flow model,to investigate the age and temperature of the ice sheet in the Dome F region,which support locating old ice cores.The workflow is as listed:(1)We processed the airborne ice-penetrating radar dataset in the Dome F region collected by Alfred Wegener Institute Helmholtz Centre for Polar and Marine Research(AWI),and tracked the internal reflection horizons(IRHs),which reveal the inner stratigraphy of ice sheet.Then we dated the IRHs by the timescale of the Dome F ice core,to provide important age constraints for ice-flow models.In addition,we extracted the bed return power(BRP)from radar data and deduced the quantitative relation between the radar attenuation rate and ice temperature based on impurity concentration in the ice,which can support the research on basal reflectivity and basel thermal state.(2)We used the 1-D steady-state mechanical inverse model Iso Inv1 D constrained by the age of isochrones to invert accumulation rate and shape factor with a resolution of1 km,and deduce the age of ice and basal thermal state along the radar profiles.Modeled age of basal ice shows the distribution and age density of the 1.5 Myr-old ice.Modeled basal thermal state shows that melting widely distributes in the Dome F region,while the stagnant ice is mainly near New Dome Fuji(NDF,located on the subglacial mountain).(3)The 1-D inverse model is computational efficient and has high resolution,but ignores advection in the ice sheet and does not take into account the thermal conditions.Thus we futher adapted a 3-D ice flow model to simulate the age and thermal structure(temperature field and basal thermal state)of ice in the study area.The model is coupled by an ice flux and thermal model,a basal hydrology model,and a rheology and shape function model,model results have a resolution of 2.5 km.The sensitivity study of the model shows that the modeled ice temperature and age of basal ice are significantly modulated by the input geothermal heat flux(GHF),the modeled ice flow is not effected by GHF.There is a clear ice divide centered on the Dome F drill site and passing NDF according to the ice flow results,where the ice is slower,more rigid and viscous,less susceptible to deformation and thus holds more continuous climate record.(4)In order to diminish the effect of initial model input and improve the reliability of the model results,a 3-D ice-flow inverse model is constructed by constraining the 3-D ice-flow model with radar internal stratigraphy and the subglacial lakes.A genetic algorithm inverse workflow improved by simulated annealing algorithm is used to invert the accumulation rate and GHF,then obtained the best-fit thermal structure and age of ice.The modeled basal temperature reaches the pressure melting point and melting exists in most of the study area,while there is lower basal temperature and ice older than 1.5 Myr around NDF and southeast of the Dome F drill site.The relative basal reflectivity calculated from BRP and modeled 3-D ice temperature is relatively low around the NDF and southeast of the Dome F drill site,which indicates the lower possibility of melting and thus more complete climate record in the ice.We used ice penetrating radar observations as constrints for a 1-D ice-flow inverse model and a 3-D ice-flow inverse model,and deduced the basal thermal state and age of ice with high agreement.Both models derive 1.5 Myr ice with cold base around the NDF,which indicates more complete climate record.In addition,NDF is located on the ice divide,holding more continuous climate record.Therefore,the area around NDF is desirable for holding old ice in the Dome F region.In conclusion,the age of ice,ice temperature and ice flow give ideas for locating old ice cores.Meanwhile,the other parameters derived by the model,such as the shape factor,accumulation rate,and GHF help with understanding the ice dynamics and thermal structure in the Dome F region and supporting further studies in the future. |
