Title: U-Th dating of lake sediments: Lessons from the 700 ka sediment record of Lake Junín, Peru — Datasets Version: 2.0 Date of Release: August 18, 2020 Last Update: September 16, 2020 Identifier: 10.5281/zenodo.4031644 Permalink: https://doi.org/10.5281/zenodo.4031644 Associated publication: Chen, C.Y., McGee, D., Woods, A., Pérez, L., Hatfield, R.G., Edwards, R.L., Cheng, H., Valero-Garcés, B.L., Lehmann, S.B., Stoner, J.S., Schwalb, A., Tal, I., Seltzer, G.O., Tapia, P.M., Abbott, M.B., and Rodbell, R.T. (2020) U-Th dating of lake sediments: Lessons from the 700 ka sediment record of Lake Junín, Peru. Quaternary Science Reviews. 244: 106422 doi: 10.1016/j.quascirev.2020.106422 Link to publication: https://doi.org/10.1016/j.quascirev.2020.106422 Suggested citation: Please reference the associated publication above when using any datasets or materials in this repository. Contact information: Christine Y. Chen, cychen@caltech.edu OR cychen.earth@gmail.com Dates of data collection and generation: February 2016 to July 2019 --------------- VERSION NOTES (v2.0) This online data repository was updated in order to correct errors in the U-Th data tables (Tables A1, A2, and A3) and subsequently update the age-depth model. The corrections impacted the files named 'CYChen et al. (2020) QSR - U-Th Data Tables for PLJ-1 and 1996 cores - v2.xlsx' and 'CYChen et al. (2020) QSR - Input Your Own Depths for Age-Depth Model - v2.xlsx'. No changes have been made to the original versions of other datasets. However, all components of this data repository have been given the "v2" identifier to be consistent with the versioning scheme applied to this data repository as a whole by Zenodo. --------------- DESCRIPTION OF DATA This data repository contains the following datasets and supplementary materials. Specific documentation and, where applicable, instructions of use are included in the readme section of each file. We refer the user to the original manuscript (see above) and the text of the Supplemental Materials published alongside this manuscript for additional general information regarding the collection and generation of these data. SUPPLEMENTARY TEXT AND FIGURES CYChen et al. (2020) QSR - Supplementary Materials.pdf DATA TABLES FOR PLJ-1 and 1996 CORE CYChen et al. (2020) QSR - U-Th Data Tables for PLJ-1 and 1996 cores - v2.xlsx: U-Th data tables for the PLJ-1 and 1996 core (with all digits preserved where possible, for reproducibility and data longevity) CYChen et al. (2020) QSR - Other Data for PLJ-1 core - v2.xlsx: sample locations and corresponding elemental concentrations, carbon coulometry, and color reflectance data CYChen et al. (2020) QSR - Ostracode Analysis for PLJ-1 core - v2.xlsx: ostracode paleoecological data from the PLJ-1 core MODELING URANIUM-THORIUM ISOTOPIC EVOLUTION OF LAKE SEDIMENTS CYChen et al. (2020) QSR - Modeling Spreadsheet for Uranium-Thorium Isotopic Evolution of Lake Sediments - v2.xlsx: spreadsheet for simulating the uranium-thorium isotopic evolution of a sample that has undergone open system behavior with respect to uranium and/or a sample of mixed composition (an impure carbonate) AGE-DEPTH MODEL CYChen et al. (2020) QSR - Input Your Own Depths for Age-Depth Model - v2.xlsx: spreadsheet for extracting the radiometric-based age-depth model generated by Bacon by splice depth

Millennial-scale climate variation during the Last Glacial period is evident in many locations worldwide, but it is unclear if such variation occurred in the interior of tropical South America, and, if so, how the low-latitude variation was related to its high-latitude counterpart. A high-resolution record, derived from the deep drilling of sediments on the floor of Lake Titicaca in the southern tropical Andes, is presented that shows clear evidence of millennial-scale climate variation between ∼60 and 20kaBP. This variation is manifested by alternations of two interbedded sedimentary units. The two units have distinctive sedimentary, geochemical, and paleobiotic properties that are controlled by the relative abundance of terrigenous or nearshore components versus pelagic components. The sediments of more terrigenous or nearshore nature likely were deposited during regionally wetter climates when river transport of water and sediment was higher, whereas the sediments of more pelagic character were deposited during somewhat drier climates regionally. The majority of the wet periods inferred from the Lake Titicaca sediment record are correlated with the cold events in the Greenland ice cores and North Atlantic sediment cores, indicating that increased intensity of the South American summer monsoon was part of near-global scale climate excursions.

Corrigendum to "Quaternary glaciation and hydrologic variation in the South American tropics as reconstructed from the Lake Titicaca drilling project" [Quaternary Research 68 (2007) 410–420] - Volume 69 Issue 2

A 136-m-long drill core of sediments was recovered from tropical high-attitude Lake Titicaca, Bolivia-Peru, enabling a reconstruction of past climate that spans four cycles of regional glacial advance and retreat and that is estimated to extend continuously over the last 370,000 yr. Within the errors of the age model, the periods of regional glacial advance and retreat are concordant respectively with global glacial and interglacial stages. Periods of ice advance in the southern tropical Andes generally were periods of positive water balance, as evidenced by deeper and fresher conditions in Lake Titicaca. Conversely, reduced glaciation occurred during periods of negative water balance and shallow closed-basin conditions in the lake. The apparent coincidence of positive water balance of Lake Titicaca and glacial growth in the adjacent Andes with Northern Hemisphere ice sheet expansion implies that regional water balance and glacial mass balance are strongly influenced by global-scale temperature changes, as well as by precessional forcing of the South American summer monsoon. (c) 2007 University of Washington. All rights reserved.

We developed records of clastic sediment flux to 13 alpine lakes in Peru, Ecuador, and Bolivia, and compared these with independently dated records of regional glaciation. Our objectives are to determine whether a strong relationship exists between the extent of ice cover in the region and the rate of clastic sediment delivery to alpine lakes, and thus whether clastic sediment records serve as reliable proxies for glaciation during the late Pleistocene. We isolated the clastic component in lake sediment cores by removing the majority of the biogenic and authigenic components from the bulk sediment record, and we dated cores by a combination of radiocarbon and tephrochronology. In order to partially account for intra-basin differences in sediment focusing, bedrock erosivity, and sediment availability, we normalized each record to the weighted mean value of clastic sediment flux for each respective core. This enabled the stacking of all 13 lake records to produce a composite record that is generally representative of the tropical Andes. There is a striking similarity between the composite record of clastic sediment flux and the distribution of ∼100 cosmogenic radionuclide (CRN) exposure ages for erratics on moraine crests in the central Peruvian and northern Bolivian Andes. The extent of ice cover thus appears to be the primary variable controlling the delivery of clastic sediment to alpine lakes in the region, which bolsters the increasing use of clastic sediment flux as a proxy for the extent of ice cover in the region. The CRN moraine record and the stacked lake core composite record together indicate that the expansion of ice cover and concomitant increase in clastic sediment flux began at least 40ka, and the local last glacial maximum (LLGM) culminated between 30 and 20ka. A decline in clastic sediment flux that began ∼20ka appears to mark the onset of deglaciation from the LLGM, at least one millennium prior to significant warming in high latitude regions. The interval between 20 and 18ka was marked by near-Holocene levels of clastic sediment flux, and appears to have been an interval of much reduced ice extent. An abrupt increase in clastic sediment flux 18ka heralded the onset of an interval of expanded ice cover that lasted until ∼14ka. Clastic sediment flux declined thereafter to reach the lowest levels of the entire length of record during the early–middle Holocene. A middle Holocene climatic transition is apparent in nearly all records and likely reflects the onset of Neoglaciation and/or enhanced soil erosion in the tropical Andes.

The Pole-Equator-Pole (PEP) projects of the PANASH (Paleoclimates of the Northern and Southern Hemisphere) programme have significantly advanced our understanding of past climate change on a global basis and helped to integrate paleo-science across regions and research disciplines. PANASH science allows us to constrain predictions for future climate change and to contribute to the management of consequent environmental changes. We identify three broad areas where PEP science makes key contributions. 1. The pattern of global changes. Knowing the exact timing of glacial advances (synchronous or otherwise) during the last glaciation is critical to understanding inter-hemispheric links in climate. Work in PEPI demonstrated that the tropical Andes in South America were deglaciated earlier than the Northern Hemisphere (NH) and that an extended warming began there ca. 21 000 cal years BP. The general pattern is consistent with Antarctica and has now been replicated from studies in Southern Hemisphere (SH) regions of the PEPII transect. That significant deglaciation of SH alpine systems and Antarctica led deglaciation of NH ice sheets may reflect either i) faster response times in alpine systems and Antarctica, ii) regional moisture patterns that influenced glacier mass balance, or iii) a SH temperature forcing that led changes in the NH. This highlights the limitations of current understanding and the need for further fundamental paleoclimate research. 2. Changes in modes of operation of oscillatory climate systems. Work across all the PEP transects has led to the recognition that the El Niño Southern Oscillation (ENSO) phenomenon has changed markedly through time. It now appears that ENSO operated during the last glacial termination and during the early Holocene, but that precipitation teleconnections even within the Pacific Basin were turned down, or off. In the modern ENSO phenomenon both inter-annual and seven year periodicities are present, with the inter-annual signal dominant. Paleo-data demonstrate that the relative importance of the two periodicities changes through time, with longer periodicities dominant in the early Holocene. 3. The recognition of climate modulation of oscillatory systems by climate events. We examine the relationship of ENSO to a SH climate event, the Antarctic cold reversal (ACR), in the New Zealand region. We demonstrate that the onset of the ACR was associated with the apparent switching on of an ENSO signal in New Zealand. We infer that this related to enhanced zonal SW winds with the amplification of the pressure fields allowing an existing but weak ENSO signal to manifest itself. Teleconnections of this nature would be difficult to predict for future abrupt change as boundary conditions cannot readily be specified. Paleo-data are critical to predicting the teleconnections of future changes.

The local last glacial maximum in the tropical Andes was earlier and less extensive than previously thought, based on 106 cosmogenic ages (from beryllium-10 dating) from moraines in Peru and Bolivia. Glaciers reached their greatest extent in the last glacial cycle ∼34,000 years before the present and were retreating by ∼21,000 years before the present, implying that tropical controls on ice volumes were asynchronous with those in the Northern Hemisphere. Our estimates of snowline depression reflect about half the temperature change indicated by previous widely cited figures, which helps resolve the discrepancy between estimates of terrestrial and marine temperature depression during the last glacial cycle.

Past fluctuations of tropical and sub-tropical glaciers provide important palaeoclimate proxies for regions where other forms of evidence are rare. However, published equilibrium-line altitude (ELA) estimates for tropical and sub-tropical glaciers at the LGM vary widely, reflecting the diversity of methods and approaches employed by different research groups. This complicates regional and global comparisons of ELA estimates, and emphasises the need for standardised methods. The distinctive character of tropical and sub-tropical glaciers, however, means that standard methods for reconstructing former glacier limits, ELAs, and palaeoclimate need to be adapted for local conditions. Many methods of ELA reconstruction explicitly or implicitly make assumptions about glacier mass balance gradients, and care needs to be taken that the choice of accumulation area ratios (AARs), balance ratios (BRs) and terminus-to-head ratios (THARs) is appropriate, as such indices are influenced by climatic regime, debris cover and other factors. ELA reconstructions should employ multiple methods, and should be cross-checked and fully reported, to allow assessment of the accuracy of ELA estimates. Reliable glacier chronologies are equally important. Dating should be based on multiple radiometric techniques wherever possible, and method of dating, the type of material dated, and the context of the date must all be reported.

The manifestation of major climatic events such as the timing of deglaciation and whether, or not, the Younger Dryas affected Andean systems has garnered considerable recent attention. Even the Holocene is rapidly emerging as a time of considerable interest in Neotropical palaeoclimatology and palaeoecology. The Holocene of the Neotropics is now revealed as a time of some temperature change with precipitation:evaporation ratios fluctuating markedly. Major changes in lake level, ice-accumulation, and vegetation are indicative of changes both in precipitation and temperature regimes, Although global-scale forcing mechanisms may Underlie some of these changes, e.g. the precessional rhythm, other variability appears to be localised. In a record from near the Upper forest limit of the eastern Peruvian Andes, pollen, charcoal, and sedimentary data suggest that the deglaciational period from ca. 17 000 to ca. 11 500 cal. yr BP was a period of rapid climatic oscillations, set against an overall trend of warming. A warm-dry event is evident between ca. 9500 and ca. 7300 cal. yr BP, and comparisons with other regional archives suggest that it was regional in scale. A ca. 1500-yr periodicity in the magnetic susceptibility data is evident between 12000 and 6000 cal.yr BP, reaching a peak intensity during the city event. A weaker oscillation with a 500-600-yr periodicity is present throughout much of the Holocene. The Uppermost sample of the pollen analysis reveals deforestation as modern human land Use simplified the landscape. Copyright (C) 2005 John Wiley & Sons, Ltd.

AbstractThe manifestation of major climatic events such as the timing of deglaciation and whether, or not, the Younger Dryas affected Andean systems has garnered considerable recent attention. Even the Holocene is rapidly emerging as a time of considerable interest in Neotropical palaeoclimatology and palaeoecology. The Holocene of the Neotropics is now revealed as a time of some temperature change with precipitation:evaporation ratios fluctuating markedly. Major changes in lake level, ice‐accumulation, and vegetation are indicative of changes both in precipitation and temperature regimes. Although global‐scale forcing mechanisms may underlie some of these changes, e.g. the precessional rhythm, other variability appears to be localised. In a record from near the upper forest limit of the eastern Peruvian Andes, pollen, charcoal, and sedimentary data suggest that the deglaciational period from ca. 17 000 to ca. 11 500 cal. yr BP was a period of rapid climatic oscillations, set against an overall trend of warming. A warm‐dry event is evident between ca. 9500 and ca. 7300 cal. yr BP, and comparisons with other regional archives suggest that it was regional in scale. A ca. 1500‐yr periodicity in the magnetic susceptibility data is evident between 12 000 and 6000 cal. yr BP, reaching a peak intensity during the dry event. A weaker oscillation with a 500–600‐yr periodicity is present throughout much of the Holocene. The uppermost sample of the pollen analysis reveals deforestation as modern human land use simplified the landscape. Copyright © 2005 John Wiley & Sons, Ltd.

Cosmogenic dating provides a long-awaited means of directly dating glacial deposits that pre-date the last glacial cycle. Although the potential benefits of longer chronologies are obvious, the greater uncertainty associated with older cosmogenic ages may be less readily apparent. We illustrate the challenges of developing and interpreting a long chronology using Our data from the Peruvian Andes. We used surface exposure dating with cosmogenic radionuclides (CRNs; Be-10 and Al-26) to date 140 boulders on moraines in valleys bordering the Junin Plain (11 degrees S, 76 degrees W) in central Peru. Our chronology spans Multiple glacial cycles and includes exposure ages greater than 1 million years, which indicate that long-term rates of boulder erosion have been very low. Interpreting the chronology of moraines for glaciations that predate the last glacial cycle is complicated by the need to consider boulder erosion and exhumation, Surface uplift, and inheritance of CRNs front previous exposure intervals. As an example, we recalculate exposure ages using our boulder erosion rates (0.3-0.5 metres per million years) and estimated surface uplift rates to emphasise both the challenges involved in interpreting old Surface exposure ages and the value of chronological data, even with large uncertainties, when reconstructing the palaeoclimate of a region. Copyright (C) 2005 John Wiley & Sons, Ltd.

The history of Holocene glaciation serves as an important record of glacier mass balance and, therefore, of climatic change. The moraine record of Holocene glaciation in the tropical Andes, however, is fragmentary and poorly dated. In contrast, increases in the rate of accumulation of inorganic sediment in glacier-fed lakes have been linked to periods of Neoglaciation in many mountain regions. The interpretation of such a record of Neoglaciation from sediment cores in glacier-fed lakes in the tropical Andes can provide the continuity and chronologic control that is lacking in the existing moraine record. Unusual exposures of glacial lacustrine sediment in the Cordillera Blanca, Few, provide a rare opportunity 10 assess the link between climatic change, glaciation, and lacustrine sedimentation.Intentional lowering of water levels in Laguna Faron (9 degrees S, 77 degrees 44'W, 4200m a.s.l.) in 1985 resulted in the incision and exposure of at least 20 m of deltaic deposits at the eastern end of the lake. Three deltaic units can be identified: horizontal topset beds, steeply clipping and deformed foreset beds, and horizontally laminated fine-grained sediment. Six radiocarbon ages ranging from 1800 +/- 210 to 465 +/- 95 (14)Cyr Bp on wood indicate that the average rate of delta progradation in the late Holocene has been approximately 290 m per 1000 yr. The lake formed during deglaciation at least 10 000 yr ago and if such a rate of prograclation of the delta had prevailed over the entire Holocene, then the delta would be at least three times as extensive as it is today. Thus the rate of delta prograclation has varied significantly over the Holocene. We suggest that the rate of delta prograclation was at least three times greater when glaciers were in advanced positions. These positions are clearly delimited by Neoglacial moraines, which are within 1-2 km of the exposures studied and within 1 km of modern ice limits. The most recent increase in the rate of delta prograclation is evidenced by an increase in sedimentation rate ca. 575 +/- 90 C-14 yr BP, coincident with the onset of the Little Ice Age. Copyright (C) 2005 John Wiley & Sons, Ltd.

The modern glaciers of the tropical Andes are a small remnant of the ice that occupied the mountain chain during past glacial periods. Estimates of local Last Glacial Maximum (LGM) snowline depression range from low (e.g., 200–300m in the Junin region, Peru), through intermediate (600m at Laguna Kollpa Kkota in Bolivia), to high (e.g., 1100–1350m in the Cordillera Oriental, Peru). Although a considerable body of work on paleosnowlines exists for the tropical Andes, absolute dating is lacking for most sites. Moraines that have been reliably dated to ∼21calkyr BP have been identified at few locations in the tropical Andes. More commonly, but still rarely, moraines can be bracketed between about 1014Ckyr (∼11.5calkyr BP) and 3014Ckyr BP. Typically, only minimum-limiting ages for glacial retreat are available. Cosmogenic dating of erratics on moraines may be able to provide absolute dating with sufficient accuracy to identify deposits of the local LGM. Ongoing work using cosmogenic 10Be and 26Al in Peru and Bolivia suggests that the local LGM may have occurred prior to 21calkyr BP.

This paper presents the results of a hydrogeochemical study of the Lake Junin watershed, Peru, to evaluate the controls over chemical composition of water in this remote Andean site. Stream, spring, and lake waters in the Lake Junin basin are all Ca2+ and HCO3− type. Evaporation has enriched the δ18O of the lake water by about 6‰ compared to that of input water. The 87Sr/86Sr ratio of dissolved strontium varies from 0.70777 to 0.72242 in the watershed because of water interactions with limestone and silicate rocks, respectively, in the watershed. The 87Sr/86Sr ratios of strontium in water from streams that flow through clastic sedimentary rocks east of the lake is more radiogenic (>0.7100) than elsewhere in the watershed, where the 87Sr/86Sr ratios of surface waters range between 0.7076 and 0.7090. These lower ratios are consistent with marine limestone sources found in these watersheds. Synoptic discharge measurements, combined with the strontium isotopic data, show that that three rivers provide over 90% of the surface water to the lake: the Río Chacachimpa, the Río San Juan, and the Río Palcamayo and about 60% of the total water input to the lake, the rest coming from direct precipitation.