Spatiotemporal Variations Of Near-surface Air Temperature Over The Antarctic Ice Sheet And Its Response To Tropical Sea Surface Temperature Anomalies

As one of the most basic elements in the climate system,near-surface air temperature plays an important role in the Antarctic sea-ice-air interaction.Its variability has a profound impact on the mass balance of the Antarctic ice sheet,the surrounding ecosystem and the rise of global sea level.Constrained by the extremely harsh natural environmental conditions in Antarctica,meteorological stations are relatively sparse and the observation time is often discontinuous,which seriously restricts the study of the spatio-temporal variability process and driving mechanism of near-surface air temperature over the Antarctic ice sheet.Therefore,it is a basic premise to establish spatio-temporal continuous high-resolution near-surface temperature dataset,using limited station observations combined with remote sensing data or climate model results.On this basis,it is of great importance to explore the near-surface air temperature variability over the Antarctic ice sheet,especially the interdecadal transition and its connection with tropical sea surface temperature anomaly,in order to reveal the driving mechanism of the spatio-temporal variability of Antarctic temperature in the context of global warming.In this thesis,based on the Neural network and Kriging interpolation method,four sets of air temperature data with different spatio-temporal resolutions are reconstructed,using the meteorological station temperature observation data,combined with MODIS（Moderate Resolution Imaging Spectroradiometer）land surface temperature measurement and global reanalysis data.And the accuracy of these products is verified.Using high-precision long-term reconstructed temperature datasets,we explore the characteristics of near-surface air temperature variability,especially the interdecadal transition over the Antarctic ice sheet,and analyze the response to tropical Pacific sea surface temperature anomalies.The main conclusions are as follows:（1）In this thesis,four sets of air temperature data with different spatio-temporal resolutions are reconstructed.First,the MODIS land surface temperature（LST）measurements are combined with the near-surface air temperature recorded by 119 meteorological stations.The neural network fitting model is used to reconstruct a high spatial resolution（0.05°×0.05°）monthly near-surface air temperature products(RECON_MODIS)over the Antarctic ice sheet during 2001-2018.Secondly,in order to extend the time scale of the reconstructed product,the measured near-surface air temperature data of 85 meteorological stations over the Antarctic ice sheet are connected with three reanalysis data（CFSR,ERA5 and MERRA-2）of 2m near-surface air temperature.The spatial interpolation is based on Kriging technique to reconstruct three long-scale monthly near-surface air temperature products(RECON_CFSR,RECON_ERA5 and RECON_MERRA-2)over the Antarctic ice sheet during 1960-2019,with spatial resolution of 0.5°×0.5°,0.25°×0.25°and 0.5°×0.625°.（2）The accuracy verification results show that the MODIS land surface temperature data combined with the neural network model is suitable for the reconstruction of near-surface air temperature over the Antarctic ice sheet.The reconstruction accuracy is high,with the mean bias of 0.09℃,the mean absolute error of 2.23℃,the coefficient of determination of 0.97,and the performance is better than that of ERA5.The accuracy of RECON_CFSR,RECON_ERA5 and RECON_MERRA-2 is also relatively reasonable,with correlation coefficients exceeding 0.65 and root-mean-square error within 0.6.And RECON_ERA5 have the highest accuracy,with correlation coefficient reaching 0.70 and root-mean-square error of 0.49.The results of RECON_MODIS and RECON_ERA5 in the overlapping period are in better agreement,with high correlation and small overall deviation,which further confirms the reliability of the reconstruction results.（3）The interdecadal variability of near-surface air temperature over the Antarctic ice sheet show obvious regional differences,especially the interdecadal variation transition occurred in the late 1990s,and is characterized by obvious seasonal dependence.During 1979-1997,the warming trend is mainly concentrated in the Antarctic Peninsula and the West Antarctic ice sheet,while the East Antarctic ice sheet has shown an obvious cooling trend.During 1999-2019,cooling trends extend from the Antarctic Peninsula to most of the West Antarctic ice sheet,a much larger area than previously reported.This seems to indicate that the global warming hiatus event also occurs in these two regions,espicially in summer and spring.The East Antarctic ice sheet shows a warming trend,particularly in the area around the South Pole during 1999-2019.（4）The tropical Pacific sea surface temperature anomaly is teleconnected with the Antarctic ice sheet and its surrounding region by driving the atmospheric Rossby wave train,and its influence on the near-surface air temperature of the Antarctic ice sheet exhibits obvious interdecadal variability and seasonal dependence.From 1979-1997,the significant warming of the Antarctic Peninsula in summer and the significant cooling of the East Antarctic ice sheet and West Antarctic ice sheet in autumn can be attributed to the enhancement of the Southern Annular Mode（SAM）toward positive phase.El Ni?o-Southern Oscillation（ENSO）forcing from the tropical Pacific weakens the Amundsen Sea Low,promoting the cooling trend of the West Antarctic ice sheet.In the early 21st century,the Interdecadal Pacific oscillation（IPO）shifted into negative phase.In summer,the significant cyclonic anomaly in the Weddell Sea have a significant cooling effect on the Antarctic Peninsula.Moreover,the strengthening of the blocking high pressure over the Drake Passage and the northern Antarctic Peninsula also contribute to the cooling trend of the Antarctic Peninsula.In spring,the negative phase of IPO lead to a deepening of the Amundsen Sea Low,which have a significant cooling effect on the West Antarctic ice sheet.And enhanced low-pressure system in the Weddell Sea contribute to a significant warming trend in the region from Coates Land to the South Pole.