81 Kr dating at the Guliya ice cap, Tibetan Plateau
Lide Tian, Florian Ritterbusch, Ji-Qiang Gu, Shui-Ming Hu, Wei Jiang, Zheng-Tian Lu, Di Wang, Guo-Min Yang
881
Kr dating at the Guliya ice cap, Tibetan Plateau
Lide Tian,
1, 2, 3, 4, ∗ Florian Ritterbusch, Ji-Qiang Gu, Shui-Ming Hu, Wei Jiang, Zheng-Tian Lu, † Di Wang, and Guo-Min Yang Institute of International Rivers and Eco-security, Yunnan University, Kunming 650500, China. CAS Center for Excellence in Tibetan Plateau Earth Sciences,Chinese Academy of Sciences, Beijing 100101, China. College of Resource and Environment, University of Chinese Academy of Sciences, Beijing, 100190, China. Yunnan Key Laboratory of International Rivers and TransboundaryEco–security, Yunnan University, Kunming 650091, China. Hefei National Laboratory for Physical Sciences at the Microscale,CAS Center for Excellence in Quantum Information and Quantum Physics,University of Science and Technology of China, Hefei 230026, China
We present radiometric Kr dating results for ice samples collected at the outlets of the Guliya icecap in the western Kunlun Mountains of the Tibetan Plateau. This first application of Kr datingon mid-latitude glacier ice was made possible by recent advances in Atom Trap Trace Analysis,particularly a reduction in the required sample size down to 1 µ L STP of krypton. Eight ice blockswere sampled from the bottom of the glacier at three different sites along the southern edges. The Kr data yield upper age limits in the range of 15 - 74 ka (90% confidence level). This is an orderof magnitude lower than the ages exceeding 500 ka which the previous Cl data suggest for thebottom of the Guliya ice core. It is also significantly lower than the widely used chronology up to110 ka established for the upper part of the core based on δ O in the ice.
I. INTRODUCTION
Alpine ice cores in the mid- and low-latitude regionsprovide high-resolution records of past climate and en-vironment. High rates of ice accumulation and melt-ing are responsible for the relatively short history ofice core records on the Tibetan Plateau as comparedto the polar regions. Longer ice cores and older iceare being sought on the Tibetan Plateau for the pur-pose of extending the climate history in this region. TheMalan and Puruogangri ice cores in the central TibetanPlateau [Thompson et al. 2006, Wang et al. 2003] andthe Dasuopu ice core in the middle of the Himalayas[Thompson et al. 2000, Yao et al. 2002] provide recordsof the past several thousand years. Samples from thebottom of the Dunde ice core in the northeastern Ti-betan Plateau were first interpreted to be glacial-stageice [Thompson et al. 1989], but later proved to be aHolocene deposit [Thompson et al. 2005]. The longest(308 . δ O signal with ∗ [email protected] † [email protected] the CH record from GISP2 in Greenland. Moreover,the Cl data suggest that the bottom ice may be olderthan 500 ka. Since then, the GIC1992 record has beenwidely used as a reference for correlating regional cli-mate signals [e.g. Cheng et al. 2012, Chevalier et al.2011, Cosford et al. 2008, Hayashi et al. 2009, Ma-howald et al. 2011]. However, the established Guliyachronology is difficult to reconcile with several recentfindings. [Cheng et al. 2012] encountered inconsistenciesbetween the δ O record of GIC1992 and the Kesang sta-lagmite record. Their work suggests that the relation-ship between δ O and CH may be inversed, leadingto a shortening of the GIC1992 age scale by a factor oftwo. Meanwhile, at the Chongce ice cap ( ∼
30 km awayfrom the GIC1992 drilling site), luminescence dating pro-vides an upper age limit of 42 ± Cl data suggests for the bottomice of GIC1992. Moreover, C dating in combinationwith ice flow modeling for ice cores from the Chongceice cap indicates Holocene deposition [Hou et al. 2018],which is consistent with all other Tibetan ice cores ex-cept GIC1992. Given the proximity between the Guliyaand the Chongce ice cap, these results make it difficultto argue that the large difference in age scale betweenGIC1992 and the other Tibetan ice cores is due to differ-ent local climate conditions in the western Kunlun Moun-tains [Thompson et al. 2005]. All the foregoing findingsraise the need for examining the GIC1992 chronologywith an independent dating method. Kr is a cosmogenic radionuclide with a half-life of229 ±
11 ka. The Kr concentration in the atmosphere(isotopic abundance Kr /Kr ∼ − ) is spatially homo-geneous with only small changes over the past 1.5 mil-lion years [Buizert et al. 2014]. These properties as well a r X i v : . [ phy s i c s . g e o - ph ] J un Figure 1. (a) Location of the Guliya ice cap on the Tibetan Plateau; (b) photograph showing the glacier cliff ( ∼
20 m tall) atsampling site GLY3; (c) Sampling sites GLY1, GLY2 and GLY3 for bottom ice of the Guliya ice cap during 2015 - 2017. Thered dot marks the summit (6710 m a.s.l.) and the red star the location of the Guliya ice core (GIC1992) drilling site (6200 ma.s.l.) from 1992 [Thompson et al. 1997]. as its chemical inertness make it a desirable tracer forgroundwater and ice over the age range of 40 ka to 1 . Kr (half-life 10 . ± .
02 a), whichis mainly produced by nuclear fuel reprocessing, canbe used to identify any young ( <
60 a) componentsor contamination of an old sample with modern air[Winger et al. 2005]. Development of the analyticalmethod of Atom Trap Trace Analysis (ATTA) has maderadiokrypton dating available to the earth science com-munity at large [Jiang et al. 2012]. Due to the large re-quired sample size (5 - 10 µ L STP of krypton), so far Krhas been used mainly for dating groundwater while forglacier ice only a demonstration study was conducted onlarge blue ice samples ( ∼
350 kg) from Taylor Glacier,Antarctica [Buizert et al. 2014]. Recently, the requiredsample size for Kr- and Kr-analysis has been re-duced down to 1 µ L STP of krypton, which can be ex-tracted from about 10 kg of Antarctic ice (containing ∼
100 mL STP air per kg ice) or 20 - 40 kg of Tibetan glacierice (25 - 50 mL STP air/kg) [Li et al. 2011]. This samplesize is still too large to re-assess the historic GIC1992 di-rectly, but is sufficient for Kr dating of samples fromthe margin sites of the Guliya ice cap, as presented inthis work.
II. METHODSA. Site description and ice sampling
Guliya is a large ice cap in the western Kunlun Moun-tains on the Tibetan Plateau with a total area of about376 km [Thompson et al. 1997, Yao et al. 1997]. Itssouthern part is of nonsurge type with stationary termi-nus positions [Yasuda and Furuya. 2015]. Remote sens-ing data show that the glaciers in this region have ex-perienced less change in recent decades compared toother glaciated mountainous regions in western China[Shangguan et al. 2017]. The Guliya ice cap even gainedmass from 2000 - 2015 [Kutuzov et al. 2018] primarilydue to increasing precipitation in the westerly regime[Yao et al. 2012]. Ice core drilling and ground pen-etrating radar show that the glacier thickness variesfrom about 50 m at the summit to a maximum thick-ness of 371 m at a location 1 . < ◦ [Kutuzov et al. 2018]. Limited field ob-servation indicates increasing negative surface mass bal-ance going from the equilibrium line altitude of around6000 m to lower elevation sites [Li et al. 2019]. The ab-lation of the ice cap is also characterized by cliff meltingat the end of the glacier outlets so that the bottom icelayers become accessible over large sections of the glacieredge. B. Air extraction from the ice samples
For Kr and Kr analysis, the air trapped in the icehas to be extracted. Prior to extraction, the surface ofthe ice samples is cleaned to remove any layers or flakydebris that may contain modern air. The ice is thenbrought out of the cold room and placed in a stainlesssteel chamber which is thereafter sealed and evacuatedfor about 30 min by scroll pumps through a water trap(stainless steel bellow immersed in ethanol at − ◦ C).Since during pumping the chamber is constantly beingflushed by the water vapor from the sublimating ice, theremaining atmospheric gas in the container is renderednegligible. After evacuation, the chamber is heated bya stove for 60 - 90 min (depending on the ice mass) un-til the ice has completely melted. The gas released fromthe ice passes through the water trap and is compressedinto a sample cylinder. The air content of the ice sampleis determined based on the final pressure in the sam-ple cylinder (Table I). Extraction efficiencies higher than95% and contamination with modern air below 1% aretypically achieved with this degassing method at a pro-cessing time of about 2 - 3 hours per sample. More detailson the extraction system and procedure are provided inthe supporting material.
C. Krypton purification and Kr measurement
The extracted gas from the ice samples was sent to theUniversity of Science and Technology of China (USTC)for krypton purification and for ATTA analysis of both Kr and Kr. Krypton is separated from the extractedgas using a purification system based on titanium getter-ing and gas chromatography [Tu et al. 2014], typicallyyielding krypton purities and recoveries both higher than90%. The Kr and Kr measurements are performedwith the latest ATTA instrument at USTC, where indi-vidual Kr and Kr atoms are selectively laser-cooledand then detected in a magneto-optical trap. The stableand abundant Kr is also measured for normalization.The resulting Kr/ Kr and Kr/ Kr ratios for thesample are compared to the corresponding ratios of areference krypton gas to derive the Kr abundance asa percentage of the atmospheric value (pMKr) and the Kr abundance given in the units of dpm/cc (decay perminutes per cc STP krypton), a convention originatingfrom decay counting. More details on Kr and Kranalysis with ATTA can be found in [Jiang et al. 2012].
Figure 2. Vertical δ O profiles along a 5 . .
27 m column at GLY3. The boxes show the positionsof the Kr-dated glacier ice samples along the vertical pro-files. The zero in height corresponds to the visible bottomof the glacier cliff, but is not necessarily the bedrock as de-bris may cover the lowermost part of the glacier. The sizeand the vertical position of the samples are roughly to scale.For GLY2, the δ O data of the lower 3 . − . (cid:104) and a standard deviation (std) of 1 . (cid:104) whereasin the upper 2 m the average is − . (cid:104) (std=2 . (cid:104) ). ForGLY3, the average is − . (cid:104) (std=2 . (cid:104) ). III. RESULTS AND DISCUSSIONA. δ O results
Figure 2 shows the oxygen isotope variation along the5 . .
27 m profile at GLY3.It is difficult to match these short δ O profiles fromGLY2 and GLY3 with the δ O record from GIC1992[Thompson et al. 1997]. However, the δ O fluctuationsalong the profile provide hints whether the ice origi-nates from the bottom or not. The accumulation layersof the glacier rapidly become thinner towards the bot-tom. The fast fluctuations of the δ O signal is averagedout when the thickness of the layers become less thanthe 2 cm cutting interval. The large fluctuations in the δ O profile at GLY3 as well as in the upper part ofGLY2 are comparable to those at the top of GIC1992[Thompson et al. 2018], suggesting that these samplesare not derived from the bottom of the glacier. The sam-ples were collected from the visible lowest part of theglacier cliff, which is not necessarily the lowest part ofthe ice as the bottom may be covered by debris from theglacier. This explanation is supported by the observationof a large amount of pebbles being deposited in front ofthe glacier cliff at GLY3. In contrast, the fluctuationsin the δ O profile at GLY2 exhibit reduced fluctuationstowards the bottom of the glacier. This indicates thatGLY2-4, collected at the bottom of the GLY2 profile, islikely close to the very bottom of the glacier ice. The re-duced fluctuations in the lower 3 . δ O profileat GLY2 may also result from mixing of ice of differentages due to complex flow leading to averaging of the δ Ovalues. The same mechanism may be responsible for thehigher fluctuations in the δ O profile at GLY3 and atthe top of GLY2 (e.g. if layers with higher δ O valuesare transported next to layers with lower δ O values)although no stratigraphic disturbance has been observedat the three sampling sites.It is difficult to match the δ O records from this studyto the one from GIC1992 because of the high ambiguityin matching the excursions and because the ice at GLY2and GLY3 originates from a different accumulation zonethan the ice at the GIC1992 drilling site. At the heightof 3 . δ O record at GLY2 exhibits a shift in themean from − . (cid:104) to − (cid:104) and below that the stan-dard deviation is reduced from 2 (cid:104) to 1 . (cid:104) (Figure 2).This behavior is similar for the δ O signal of GIC1992with the difference that the mean δ O value at the bot-tom 40 m is higher than the bottom 3 . (cid:104) . This is likely due to the altitude difference ofthe accumulation zones of the ice at GLY2 and GIC1992. B. Air content
The measured air contents in the ice samples arelisted in Table I. They vary from 32 mL STP / kg to59 mL STP / kg, which is typical for Himalayan ice cores[Hou et al. 2007, Li et al. 2011] and significantly lowerthan the air content of Antarctic ice, typically rang-ing between 100 - 120 mL STP / kg [Buizert et al. 2014,Raynaud and Lebel. 1979], or that of Greenland iceat 80 - 100 mL STP / kg [Raynaud et al. 1997]. This isdue to the lower air pressure at high elevation (5500-6700 m) of the deposition site and the higher tem-perature compared to Antarctica [Eicher et al. 2016,Martinerie et al. 1992]. We deliberately collected thesamples from ice layers with visibly high bubble contentand avoided those with transparent ice which are likelylayers of re-frozen meltwater. C. Kr and Kr results
The measured Kr and Kr abundances for the eightglacier ice samples as well as two air samples of Lhasa arelisted in Table I. As described above, the Kr in the at-mosphere has almost exclusively been produced anthro-pogenically in the past 60 years. Therefore, any sampleolder than that should have a vanishing Kr abundance.Five of the eight samples have Kr activity levels below3% of the Lhasa air value (Table I), whereas GLY2-1,GLY2-2 and GLY3-2 have Kr values corresponding toabout 8%, 3% and 5%, respectively. Air leaks are thor-oughly investigated on instruments used in the degassing,purification and ATTA measurement, leading to the con- clusion that contamination of modern air during theseprocesses is below 1%. It thus seems more likely thatmodern air had already entered the ice prior to sampling,e.g. by cracking/melting and refreezing, as has been ob-served in earlier works on glacier ice close to the surfaceof margin sites [Craig et al. 1990, Buizert et al. 2014].Since there is no obvious correlation between Kr andwhether the sample is from the surface or from a cave,potential contamination processes at the very front ofthe glacier ice cliff do not seem to be responsible forthat. Since the measured Kr abundances are close tothe modern value of 100 pMKr, contamination of mod-ern air at these low concentrations does not affect thereported Kr abundances within the given precisions.For all samples they are consistent with modern atmo-spheric Kr abundance within 1 σ , except for GLY3-1which still lies within a 2 σ error. We translate the mea-sured relative Kr abundances into Kr-ages using theFeldman-Cousins method [Feldman and Cousins. 1998].As the Kr abundances are close to modern, this methodyields upper age limits (90% confidence level) for the in-dividual samples that range between 15 - 74 ka.
D. Implication for the Guliya ice core chronology
The obtained results for Kr and δ O of the Guliyamargin samples allow for a discussion in the context ofthe results from GIC1992 [Thompson et al. 1997] (see in-troduction). The Kr measurements do not show evi-dence for ice older than 74 ka at the bottom of the sam-pled margin sites of the Guliya ice cap. For the samplesfrom GLY1, where the ice from GIC1992 is expected tooutcrop [Kutuzov et al. 2018], the upper limits for the Kr age do not exceed 52 ka. For GLY2-4, whose δ Oprofile exhibits bottom ice characteristics, the Kr re-sults provide an upper age limit of only 25 ka. The ob-tained upper age limits do not necessarily rule out the ex-istence of older ice somewhere else in the Guliya ice cap.It is possible that the old ice at the bottom of GIC1992 isfrozen to the bedrock and does not flow out to the mar-gin sites. However, radar measurements indicate thatthe ice at the bottom of GIC1992 does flow and is nottrapped at the bedrock [Kutuzov et al. 2018]. A furtherexplanation is that the stratigraphy of the glacier ice isfolded when travelling from the GIC1992 drilling site tothe margin, such that the old ice may not be at the bot-tom. No evidence for folding was observed at the glacierterminals, which exhibit clear horizontal layer structures,but folding on intermediate distance scales may have oc-curred. Yet another possibility is that the bottom 100 mof GIC1992, which are supposedly older than 50 ka, arerapidly thinning towards the outlet of the glacier, andtherefore may be contained in a much smaller verticalextent at the very bottom of the glacier cliff. Since thesamples at GLY1 were taken in about 2 m height abovebedrock, they may not reach into this old bottom section.However, measurements of the mass balance and the
Table I. Compilation of the Kr and Kr results. The Kr abundance is reported in units of pMKr (percent ModernKrypton). The atmospheric level is 100 pMKr. The Kr abundance is reported in the units of dpm/cc (decays per minute percc STP of krypton). The errors are 1 σ standard deviations whereas upper limits are reported for a 90% confidence level.sample note weight Air content Krypton Kr Kr Kr-agekg mL STP/kg µ L STP dpm/cc pMKr kaGLY1-1 Cave 34 52 1.4 < . ± < . ± . ± < . ± . ± < . ± . ± < . ± . ± < < . ± < . ± . ± < . ± . ± < ± ± glacier surface velocity [Li et al. 2019, Chadwell 2017]indicate that a large fraction of the upper glacier layersis lost when flowing from the equilibrium line altitudeto the edge of the glacier at GLY1 where the remainingglacier cliff is about 10 m in height. Therefore, it doesnot seem likely that the bottom 100 m at the GIC1992drilling site are thinning to below our sampling heightabout 2 m above bedrock at GLY1. IV. CONCLUSIONS AND OUTLOOK
Radiometric Kr dating has been used to determinethe age of bottom ice samples at the Guliya ice cap.Eight ice blocks, each weighing 28 - 69 kg, were collectedat three different outlets of the glacier, and analyzed for Kr using the Atom Trap Trace Analysis method. The Kr results yield upper limits in the range of 15 - 74 ka,which is an order of magnitude lower than previouslysuggested by Cl dating of the Guliya ice core and alsosignificantly lower than the Guliya chronology reachingup to 110 ka based on δ O measurements. After resultsfrom the Kesang stalagmite cave ( ∼
860 km distance tothe Guliya ice cap) and the Chongce ice cap ( ∼
30 km dis-tance), the Kr data in this work (obtained directly frombottom samples of the Guliya ice cap) represent yet an-other result that calls for further dating measurements tocheck the established Guliya chronology. Measurementsof C, Cl, Be, δ O atm and argon isotope ratios areplanned for a new Guliya ice core that has been drilled in2015 close to the location of the 1992 Guliya core drillingsite [Thompson et al. 2018]. Meanwhile, at the USTClaboratory, work is in progress to further reduce the sam-ple size required for Kr analysis so that bottom samplesfrom a Guliya ice core can be measured directly.
ACKNOWLEDGMENTS
This work is funded by National Natural Science Foun-dation of China (41530748), the National Key Researchand Development Program of China (2016YFA0302200)and the Chinese Academy of Sciences (XDB21010200).We thank Lili Shao and Cheng Wang from the Instituteof Tibetan Plateau Research for their assistance in theice degassing and Lei Zhao from USTC for purifying thekrypton samples.
An edited version of this paper was published by AGU.Copyright 2019 American Geophysical Union.Tian, L., Ritterbusch, F., Gu, J.-Q., Hu, S.-M., Jiang,W., Lu, Z.-T., et al. (2019). Kr dating at the Guliyaice cap, Tibetan Plateau. Geophysical Research Letters,46. https://doi.org/10.1029/ 2019GL082464.
SUPPORTING MATERIALPhotographs showing the sampling sites
Photographs of the glacier cliffs at the three differentsampling sites are shown in Figure S1.
Figure S1. Photographs showing the sampling sites GLY1 (a),GLY2 (b) and GLY3 (c) at the Guliya ice cap. In (a) and (b)the caves that have been dug for sampling are shown. Theyare about 1 . Extraction of air from the ice samples
A system for degassing the air from large (up to 90 kg)glacier ice samples has been set up in the course of thisstudy (Figure S2). The ice tank has a volume of 140 Lwith 50 cm inner diameter and 60 cm in height. The lid isO-ring sealed and and has a window on top that allowsvisual monitoring of the state of the ice. Pump1 andpump2 are Edwards nXDS10i dry scroll vacuum pumps.They are both used to evacuate the system while pump2is also used to compress the sample gas from the ice tankinto the sample container.V1 and V2 are O-ring sealed bellow valves. V3 and V4are stainless steel welded bellow valves with 6mm tubepress fittings. The opening range of V1 is used to controlthe flux of gas out of the tank. P1 and P3 are capacitancepressure gauges (100 Pa – 200 kPa range). P1 is used tomeasure the pressure in the tank and P3 the pressurein the sample cylinder. P2 is a Baratron pressure gaugewith an upper limit of 1 . . . − ◦ C. Theair is extracted from the ice in three steps:(1) The ice is placed in the tank and the lid closed off.The residual atmospheric air is removed from the tankby pump1 and pump2 via the water trap which protectsthe pumps from the water vapor sublimating from theice. After about 15 minutes a steady pressure of about100 Pa at P1 in the tank is established, which correspondsto the sublimation pressure of ice at around − ◦ C (Thisis the storage temperature of the ice in the freezer). Thetank is then pumped for another 15 minutes. Since dur-ing pumping the chamber is constantly being flushed bythe water vapor from the sublimating ice, the remain-ing atmospheric gas in the container is decreased to anegligible level.(2) After removal of the atmospheric air, valve V1 isclosed. The tank is heated with a gas stove until the iceis fully melted, a process that can be visually observedthrough a window in the lid of the tank. The meltingprocess typically takes 60 - 90 minutes depending on themass of the ice. While the ice is being melted, the samplecylinder and the rest of the system are evacuated withV2, V3 and V4 open.(3) After the ice has fully melted the gas released fromthe ice is mostly in the headspace above the meltwater,since the volume of the tank is typically more than twicethe volume of the water. Then, V2 and V3 are closed,V1 is opened (V4 remains open), and pump2 is used to compress the gas released from ice via the water trapto remove the water vapor. In order to keep the watervapor load low, dose valve V1 is regulated such that thepressure in the tank does not go below 25 mbar which isclose to the water vapor pressure at 20 ◦ C. After about10 minutes of compression, the increase of the pressurein the sample cylinder (measured by P3) slows down.After a further 5 minutes, the sample cylinder is closedwith valve V4. As the volume of the sample cylinderis known, the extracted amount of air can be obtainedfrom the final pressure reading of P4 with an accuracyof about 5%. The extraction efficiency of the degassingsystem has been measured by mixing degassed water witha known amount of air in the ice tank and compressingthe air into the sample cylinder following Step (3). Themeasured extraction efficiency is > [Buizert et al. 2014] C. Buizert, D. Baggenstos, W. Jiang,R. Purtschert, V. V. Petrenko, Z. T. Lu, P. M¨uller,T. Kuhl, J. Lee, J. P. Severinghaus, E. J. Brook(2014), Radiometric Kr dating identifies 120,000-year-old ice at Taylor Glacier, Antarctica.
Proceedings ofthe National Academy of Sciences , 111(19):6876, doi:10.1073/pnas.1320329111.[Chadwell 2017] C. Chadwell (2017), Reliability analysis fordesign of stake networks to measure glacier surface ve-locity,
Journal of Glaciology , 45(149), 154-164, doi:10.1017/s0022143000003130.[Cheng et al. 2012] H. Cheng, P. Z. Zhang, C. Sp¨otl, R. L.Edwards, Y. J. Cai, D. Z. Zhang, W. C. Sang, M. Tan,and Z. S. An (2012), The climatic cyclicity in semiarid-arid central Asia over the past 500,000 years,
GeophysicalResearch Letters , 39(1), doi: 10.1029/2011gl050202.[Chevalier et al. 2011] M.-L. Chevalier, G. Hilley, P. Tappon-nier, J. Van Der Woerd, J. Liu-Zeng, R. C. Finkel,F. J. Ryerson, H. Li, and X. Liu (2011), Con-straints on the late Quaternary glaciations in Ti-bet from cosmogenic exposure ages of moraine sur-faces,
Quaternary Science Reviews , 30(5), 528-554, doi:https://doi.org/10.1016/j.quascirev.2010.11.005.[Cosford et al. 2008] J. Cosford, H. Qing, D. Yuan, M.Zhang, C. Holmden, W. Patterson, and C. Hai(2008), Millennial-scale variability in the Asian mon-soon: Evidence from oxygen isotope records fromstalagmites in southeastern China,
Palaeogeography,Palaeoclimatology, Palaeoecology, , 266(1), 3-12, doi:https://doi.org/10.1016/j.palaeo.2008.03.029.[Craig et al. 1990] H. Craig, T. E. Cerling , R. D. Willis, W.A. Davis, C. Joyner, N. Thonnard (1990), Krypton-81in Antarctic ice: first measurement of a Krypton age onancient ice,
EOS 71:1825. , 71:1825.[Eicher et al. 2016] O. Eicher, M. Baumgartner, A. Schilt, J.Schmitt, J. Schwander, T. F. Stocker, and H. Fischer(2016), Climatic and insolation control on the high-resolution total air content in the NGRIP ice core,
Cli-mate of the Past , 12(10), 1979-1993, doi: 10.5194/cp-12-1979-2016.[Feldman and Cousins. 1998] G. J. Feldman and R. D.Cousins (1998), Unified approach to the classical sta- tistical analysis of small signals,
Physical Review D , 57,3873–3889, doi: 10.1103/PhysRevD.57.3873.[Hayashi et al. 2009] T. Hayashi, Y. Tanimura, Y. Kuwa-hara, M. Ohno, M. Mampuku, R. Fujii, H. Sakai,T. Yamanaka, T. Maki, M. Uchida, W. Yahagi, H.Sakai (2009), Ecological variations in diatom as-semblages in the Paleo-Kathmandu Lake linked withglobal and Indian monsoon climate changes for the last600,000 years,
Quaternary Research , 72(3), 377-387, doi:10.1016/j.yqres.2009.07.003.[Hou et al. 2004] S. Hou, D. Qin, J. Jouzel, V. Masson-Delmotte, U. Von Grafenstein, A. Landais, N. Caillon,and J. Chappellaz (2004), Age of Himalayan bottomice cores,
Journal of Glaciology , 50(170), 467-468, doi:10.3189/172756504781829981.[Hou et al. 2007] S. Hou, J. Chappellaz, J. Jouzel, P. C.Chu, V. Masson-Delmotte, D. Qin, D. Raynaud, P. A.Mayewski, V. Y. Lipenkov, and S. Kang (2007), Sum-mer temperature trend over the past two millennia usingair content in Himalayan ice,
Climate of the Past , 3(1),89-95, doi: 10.5194/cp-3-89-2007.[Hou et al. 2018] S. Hou, T. M. Jenk, W. Zhang, C. Wang,S. Wu, Y. Wang, H. Pang, and M. Schwikowski (2018),Age ranges of the Tibetan ice cores with emphasis onthe Chongce ice cores, western Kunlun Mountains,
TheCryosphere , 12(7), 2341-2348, doi: 10.5194/tc-12-2341-2018.[Jiang et al. 2012] W. Jiang, K. Bailey, Z.-T. Lu, P. Mueller,T.P. O’Connor, C.-F. Cheng, S.-M. Hu, R. Purtschert,N.C. Sturchio, Y.R. Sun, W.D. Williams, G.-M. Yang(2012), An atom counter for measuring Kr and Krin environmental samples,
Geochimica et CosmochimicaActa , 91, 1-6, doi: 10.1016/j.gca.2012.05.019.[Kutuzov et al. 2018] S. Kutuzov, L. G. Thompson, I.Lavrentiev, and L. Tian (2018), Ice thickness measure-ments of Guliya ice cap, western Kunlun Mountains (Ti-betan Plateau), China,
Journal of Glaciology , 1-13, doi:10.1017/jog.2018.91.[Li et al. 2011] J. Li, B. Xu, and J. Chappellaz (2011),Variations of air content in Dasuopu ice corefrom AD 1570–1927 and implications form climatechange,
Quaternary International , 236(1-2), 91-95, doi:
Cold Regions Science and Technology , 159, 71-85, doi: 10.1016/j.coldregions.2018.12.001.[Loosli and Oeschger. 1969] H. H. Loosli and H. Oeschger(1969), Ar and Kr in the atmosphere,
Earthand Planetary Science Letters , 7(1), 67-71, doi:10.1016/0012-821X(69)90014-4.[Lu et al. 2014] Z.T. Lu, P. Schlosser, W.M. Smethie Jr.,N.C. Sturchio, T.P. Fischer, B.M. Kennedy, R.Purtschert, J.P. Severinghaus, D.K. Solomon, T. Tan-hua, R. Yokochi (2014), Tracer applications of noble gasradionuclides in the geosciences,
Earth-Science Reviews ,138, 196-214, doi: 10.1016/j.earscirev.2013.09.002.[Mahowald et al. 2011] N. Mahowald, S. Albani, S. En-gelstaedter, G. Winckler, and M. Goman (2011),Model insight into glacial–interglacial paleodust records,
Quaternary Science Reviews , 30(7-8), 832-854, doi:10.1016/j.quascirev.2010.09.007.[Martinerie et al. 1992] P. Martinerie, D. Raynaud, D. M.Etheridge, J.-M. Barnola, and D. Mazaudier (1992),Physical and climatic parameters which influence the aircontent in polar ice,
Earth and Planetary Science Letters ,112(1), 1-13, doi: 10.1016/0012-821X(92)90002-D.[Raynaud and Lebel. 1979] D. Raynaud and B. Lebel (1979),Total gas content and surface elevation of polar ice sheets,
Nature , 281(5729), 289-291, doi: 10.1038/281289a0.[Raynaud et al. 1997] D. Raynaud, J. Chappellaz, C. Ritz,and P. Martinerie (1997), Air content along the Green-land Ice Core Project core: A record of surface climaticparameters and elevation in central Greenland,
Jour-nal of Geophysical Research: Oceans , 102(C12), 26607-26613, doi:10.1029/97JC01908.[Shangguan et al. 2017] D. Shangguan, S. Liu, Y. Ding, J.Li, Y. Zhang, L. Ding, X. Wang, C. Xie, and G. Li(2017), Glacier changes in the west Kunlun Shan from1970 to 2001 derived from Landsat TM/ETM+ and Chi-nese glacier inventory data,
Annals of Glaciology , 46(1),204-208, doi: 10.3189/172756407782871693.[Thompson et al. 1989] L. G. Thompson, E. Mosley-Thompson, M. E. Davis, J. F. Bolzan, J. Dai, L. Klein,T. Yao, X. Wu, Z. Xie, and N. Gundestrup (1989),Holocene–late Pleistocene climatic ice core records fromQinghai-Tibetan plateau,
Science , 246(4929), 474-477,doi: 10.1126/science.246.4929.474.[Thompson et al. 1997] L. G. Thompson, T. Yao, M. E.Davis, K. A. Henderson, E. Mosley-Thompson, P. N. Lin,J. Beer, H. A. Synal, J. Cole-Dai, and J. F. Bolzan (1997),Tropical climate instability: The last glacial cycle from aQinghai-Tibetan ice core,
Science , 276(5320), 1821-1825,doi: 10.1126/science.276.5320.1821.[Thompson et al. 2000] L. G. Thompson, T. Yao, E. Mosley-Thompson, M. E. Davis, K. A. Henderson, andP. Lin (2000), A high-resolution millennial recordof the south Asian monsoon from Himalayan icecores,
Science , 289(5486), 1916-1920, doi: 10.1126/sci-ence.289.5486.1916.[Thompson et al. 2005] L. G. Thompson, M. E. Davis, E. Mosley-Thompson, P. N. Lin, K. A. Henderson, and T.A. Mashiotta (2005), Tropical ice core records: evidencefor asynchronous glaciation on Milankovitch timescales,
Journal of Quaternary Science , 20(7-8), 723-733, doi:10.1002/jqs.972.[Thompson et al. 2006] L. G. Thompson, E. Mosley-Thompson, M. E. Davis, T. A. Mashiotta, K. A.Henderson, P. N. Lin, and T. Yao (2006), Ice coreevidence for asynchronous glaciation on the TibetanPlateau,
Quaternary International , 154(0), 3-10, doi:10.1016/j.quaint.2006.02.001.[Thompson et al. 2018] L. G. Thompson, T. Yao, M. E.Davis, E. Mosley-Thompson, G. Wu, S. E. Porter,B. Xu, P. Lin, N. Wang, E. Beaudon, K. Duan, M.R. Sierra-Hernandez, D. V. Kenny (2018), Ice corerecords of climate variability on the Third Pole withemphasis on the Guliya ice cap, western KunlunMountains, Quaternary Science Reviews , 188, 1-14, doi:10.1016/j.quascirev.2018.03.003.[Tu et al. 2014] L.-Y. Tu, G.-M. Yang, C.-F. Cheng, G.-L. Liu, X.-Y. Zhang, and S.-M. Hu (2014), Anal-ysis of Krypton-85 and Krypton-81 in a Few Litersof Air,
Analytical Chemistry , 86(8), 4002-4007, doi:10.1021/ac500415a.[Wang et al. 2003] N. Wang, L. G. Thompson, M. E. Davis,E. Mosley-Thompson, T. Yao, and J. Pu (2003), In-fluence of variations in NAO and SO on air temperatureover the northern Tibetan Plateau as recorded by δ O inthe Malan ice core,
Geophysical Research Letters , 30(22),CLM 1-5, doi: 10.1029/2003gl018188.[Winger et al. 2005] K. Winger, J. Feichter, M. B. Kali-nowski, H. Sartorius, and C. Schlosser (2005), A newcompilation of the atmospheric Krypton-85 inventoriesfrom 1945 to 2000 and its evaluation in a global transportmodel,
Journal of Environmental Radioactivity , 80(2),183-215, doi: 10.1016/j.jenvrad.2004.09.005.[Yao et al. 1997] T. Yao, L. G. Thompson, and Y. Shi (1997),Climatic variation since the Last Interglacial recorded inthe Guliya ice core,
Science in China (D) , 40(6), 662-668,doi: 10.1007/BF02877697.[Yao et al. 2002] T. Yao, K. Duan, B. Xu, N. Wang, J. Pu, S.Kang, X. Qin, and L. G. Thompson (2002), Temperatureand methane changes over the past 1000 years recordedin Dasuopu glacier (central Himalaya) ice core,
Annalsof Glaciology , Vol 35, 35, 379-383.[Yao et al. 2012] T. Yao, L. G. Thompson, W. Yang, W. Yu,Y. Gao, X. Guo, X. Yang, K. Duan, H. Zhao, B. Xu,J. Pu, A. Lu, Y. Xiang, D. B. Kattel and D. Joswiak(2012), Different glacier status with atmospheric circula-tions in Tibetan Plateau and surroundings,
Nature Cli-mate Change , 2(9), 663-667, doi: 10.1038/Nclimate1580.[Yasuda and Furuya. 2015] T. Yasuda and M. Furuya (2015),Dynamics of surge-type glaciers in West KunlunShan,
Northwestern Tibet, Journal of GeophysicalResearch: Earth Surface , 120(11), 2393-2405, doi:10.1002/2015jf003511.[Zhang et al. 2018] Z. Zhang, S. Hou, and S. Yi (2018), Thefirst luminescence dating of Tibetan glacier basal sedi-ment,