Data set on sedimentology, palaeoecology and chronology of Middle to Late Pleistocene deposits on the Taimyr Peninsula, Arctic Russia

This Data in Brief paper contains data (including images) from Quaternary sedimentary successions investigated along the Bol'shaya Balakhnya River and the Luktakh–Upper Taimyra–Logata river system on southern Taimyr Peninsula, NW Siberia (Russia). Marine foraminifera and mollusc fauna composition, extracted from sediment samples, is presented. The chronology (time of deposition) of the sediment successions is reconstructed from three dating methods; (i) radiocarbon dating of organic detritus (from lacustrine/fluvial sediment) and molluscs (marine sediment) as finite ages (usually <42 000 years) or as non-finite ages (>42 000–48 000 years) on samples/sediments beyond the radiocarbon dating limit; (ii) Electron Spin Resonance (ESR) dating on marine molluscs (up to ages >400 000 years); (iii) Optically Stimulated Luminescence (OSL) dating, usually effective up to 100–150 0000 years. Terrestrial Cosmogenic Nuclide (TCN) exposure dating has been applied to boulders resting on top of moraine ridges (Ice Marginal Zones). See (Möller et al., 2019) (doi.org/10.1016/j.earscirev.2019.04.004) for interpretation and discussion of all data.


Data
The data presented here and in M€ oller et al. [1] come from studies of sediment exposures along the Bol'shaya Balaknya and the Luktakh e Upper Taimyra e Logata river systems on the southern part of the Taimyr Peninsula, NW Siberia (Fig. 1), and from a complex of sites situated on the southern shore of Specifications

How data was acquired
The logging and photographing of excavated sedimentary successions (see logs in [1]), as well as sampling for palaeontological analyses and dating (all sampling points shown in sediment logs in ( [1]), took place during boat cruises along the Bol'shaya Balaknya River and the Luktakh eUpper TaimyraeLogata river systems on the Taimyr Peninsula, NW Siberia, in 2010 and 2012. Field sampling procedures are described in text, as well as laboratory procedures.

Data format
Raw and analysed Experimental factors Sediment successions in river-cut bluffs and solifluction scars were cleaned in vertical sections close to the permafrost table and logged to their lithofacies (Table 1), and sampled for palaeontological analysis (Tables 2e4) and dating ( 14 C, ESR, OSL; Tables 5e7). Erratic boulders on Ice Marginal Zones were sampled for TCN dating (Tables 8e10).

Experimental features
Sediment succession logging provide basis for palaeoenvironmental interpretation for discerned sediment units at the specific site and retrieved chronological data ( 14 C, ESR, OSL, TCN ages) form a base for temporal environmental reconstructions on a regional scale.

Data accessibility
Data is within this article

Value of the data
The comprehensive set of photographs of sediments and their structures provides a reference for interpretation of depositional settings/environments across the Arctic. The multi-disciplinary approach, combining a large chronometric database from radiocarbon, OSL, ESR, and terrestrial cosmogenic nuclide dating with "classical" palaeontological analyses of flora and fauna sets an example for deciphering the complex succession of glaciations and ice free periods. Presented data can be used to constrain palaeo-glaciological modelling of the Kara Sea Ice Sheet as part of the Eurasian Ice Sheet for described temporal phases.
The study adds new evidence to ongoing studies of the decisive roles both of this ocean and of the Arctic from a global change perspective.
the Khatanga River close to the small settlement of Novorybnoye (site 8, Fig. 1 36 Cl exposure ages on erratic boulders sampled from the top of mapped Ice Marginal Zones (IMZs) (see Fig. 12).

Sedimentology and stratigraphy
We focused on laterally extensive river bluff sections for sedimentological and lithostratigraphical descriptions, and targeted geochronological sampling. The sections were dug out in a stair-case manner (see Fig. 5B in [1]) in which sediment composition and structures were logged mostly at 1:10 scale (all site logs are in [1]). A number of images are presented below as examples of sediment composition and structures, and references to these are given in the site descriptions in [1]. Lithofacies codes in photographs are according to Table 1.

Foraminiferal analyses
Selected sites with marine or possibly marine strata were sampled for foraminiferal analyses. A total of 129 samples from eight sections (sections BBR 6,8,12,13,15,16,17, Nov 1 and LuR 6; Fig. 1) were collected. The samples were processed at the Dept. of Geoscience, Aarhus University, Denmark, using 40e160 g of dry sediment (most commonly 90e140 g). The samples were wet-sieved using tap-water and sieve sizes with mesh diameters of 63, 100 and 1000 mm, cf. [8], and dried in an oven at 40 C. The foraminifera in the 100e1000 mm fraction were subsequently concentrated using the heavy liquid C 2 Cl 4 (density of 1.6 g/cm 3 ), collected and taxonomically identified. Unfortunately, most samples proved barren; only very few foraminiferal specimens were found in only two of the sections and only benthic foraminifera were present (Table 2).

Marine mollusc faunas
Molluscs were collected during stratigraphic work, both for dating purposes ( 14 C, ESR) and, when encountered in larger numbers, for determination of the marine mollusc fauna for the relevant stratigraphic units (Table 3). The analyses were carried out at the Geological Museum, University of Copenhagen, Denmark. The biostratigraphy of Siberian raised marine sediments based on mollusc faunas has traditionally played an important role in the construction of a Pleistocene stratigraphy and reconstruction of palaeoenvironments, based on the species' present distribution, e.g. [9]. The species are classified according to their present distribution into Subarctic (SA), Arctic (A), and non-indicative (N/A). This is based on oceanographical parameters, notably the inflow of Atlantic water into the Arctic, a decisive factor in the distribution of near-shore marine ecosystems, and absence/duration of sea ice [10]. Subarctic species occur in the zone where Atlantic and Arctic water masses mix and seasonal sea ice occurs, such as today in the southern and eastern Barents Sea and western part of the Kara Sea, while Arctic species thrive in Arctic water masses with long lasting sea ice cover. A third biogeographical group, the Boreal species, is restricted to permanently ice free coasts. None of these species have been observed in the present material, although they occur in interglacial sediments in the Yenissei River basin to the south [9]. At present the eastern Kara Sea is dominated by Arctic water masses, but with a high inflow of fresh river water in the southern part [11].

Terrestrial and limnic macrofossil analyses
Organic debris in fluvial ripple-laminated successions was analysed from one site (LoR 3, Fig. 1), five samples in total, for their macrofossil content ( Table 4). The samples were wet-sieved (mesh !0.1 mm) and the residue left on the sieves was analysed using a Leica Wild dissecting microscope (analysed at Geological Survey of Denmark and Greenland (GEUS), Denmark (macrofossils)). The plant names are according to http://www.theplantlist.org/. Leaves, seeds and fruits were well preserved and come from local sources. The plant residue includes numerous remains of mosses; a few tentative identifications are included, but most moss remains were not identified. The remains of mosses usually preserve well and often dominate Quaternary macro-floras from the Arctic, reflecting that mosses are important constituents of Arctic plant communities. Some animal remains, especially Coleoptera fragments, were also identified to genera or species level (analysed at the Dept. of Biology and Environmental Science, Linnaueus University, Sweden (insects))

Geochronology
Four dating methods were employed: Accelerator Mass Spectrometer radiocarbon dating (AMS 14 C; molluscs, terrestrial organic material), Electron Spin Resonance (ESR; molluscs), Optically Stimulated ¼ North Taimyr ice marginal zone according to Alexanderson et al. [5]. Lines marked P south and west of the Urdakh IMZ are piedmont glacier moraines, deposited by ice from the Putorana Plateau. Yellow circles, numbered 1e15, mark the position of sites/ site areas described stratigraphically in [1] and below in this paper. Small circles color-coded in green, red, purple, yellow and white (chronostratigraphic division) mark positions of stratigraphic sites described in [2]. The base map is from the International Bathymetric Chart of the Arctic Ocean (IBCAO) [6]. Luminescence (OSL; sediment) and in situ Terrestrial Cosmogenic Nuclide surface exposure dating (TCN; boulders).
Radiocarbon dating. e A total of 66 AMS 14 C ages were determined at the AMS Radiocarbon Dating Laboratory, Department of Geology at Lund University, Sweden (Table 5). Pre-treatment of mollusc shells included leaching to~70% of their original mass. Finite ages from terrestrial material (wood, organic detritus, plant macrofossils, bone) are given as conventional radiocarbon years ( 14 C age BP) with 1s age deviation, as well as calibrated calendar years (cal yr BP or cal ka BP), calculated with the software package Oxcal 4.3.2 [12] and with use of IntCal 13 (mean age ±1s).
ESR dating. e A total of 39 marine mollusc samples were dated by Electron Spin Resonance (ESR) at the Research Laboratory for Quaternary Geochronology at Tallinn Technical University, Estonia (Anatoly Molodkov) ( Table 6). Unexposed shells were retrieved from within cleaned sections, followed by sampling of sediments enclosing the sampled shell for later measurements of background dose rates. The method is based on direct measurements of the amount of radiation-induced paramagnetic centres, trapped in the fossil shell substance and created by the natural radiation resulting from radioactivity in the shell itself and from the enclosing sediment. Standard analytical procedures were used according to Molodkov [13] and Molodkov et al. [14] and ESR age were calculated from the measured total radiation dose that the shell received during its burial versus dose rate [15]. In some sediment sections where sediment logs indicate the presence of molluscs it was unfortunately not possible to retrieve molluscs for ESR dating, either because they were too low in concentration, very friable and/or partly dissolved in situ. Although their presence was confirmed by weathered-out and hardened shells lying on exposed sediment surfaces, such shells are un-suitable for ESR dating because of prolonged daylight exposure and the difficulty of unambiguous identification of samples of the relevant burial sediment.
OSL dating. e A total of 76 sediment samples were dated by Optically Stimulated Luminescence (OSL) ( Table 7). Sediment samples were taken by means of hammering 20 cm long PVC tubes into cleaned pit walls of suitable sediment (see Fig. 5C in [1]). Samples marked with an OSL laboratory code R-xxxxxx (Table 7) were processed at Aarhus University's Nordic Laboratory for Luminescence (NLL) Dating located at the Risø Campus, Roskilde, Denmark, while samples marked S-xxxxx were handled at SCIDR Luminescence Laboratory, Sheffield University, UK. After conventional grain-size  and density separation and subsequent chemical purification, the single aliquot regenerative (SAR) dose protocol was applied to multi-grain (180e250 mm) quartz aliquots (8 mm diameter, typically >18 per sample) to estimate the equivalent dose, D e [16,17]), using blue (470 ± 30 nm) light stimulation, 260 C preheating for 10 s, and a cut heat of 220 C. Photon detection was through a U-340 glass filter, and the signal used for D e determination was based on the first 0.8 s of OSL, less a background based on the signal detected between 1.6 and 2.4 s of stimulation. To test the applicability of this chosen protocol to the measurement of the dose recoded by the quartz OSL signal, we applied a dose recovery test ( [18]) to at least 3 aliquots from each sample dated at the NLL, after initial bleaching with blue light for 100s, followed by a 10 ks pause and a further 100s bleach. The average measured/given dose ratio is 0.999 ± 0.011 (n ¼ 168) demonstrating that our protocol is able to accurately measure a dose given to a sample prior to any laboratory heating. The equivalent doses (D e ), measured for each sample are given in Table 7.
Because feldspar infra-red stimulated luminescence (IRSL) signals are more difficult to reset by daylight than the OSL signals from quartz [19,20], the apparent quartz and feldspar deposition ages of a particular sediment give information on the probability that the most light sensitive signal (quartz OSL) was fully reset prior to deposition. Accordingly, multi-grain (180e250 mm) feldspar aliquots (3 mm diameter, at least 3 aliquots per sample) extracted from the samples processed by NLL were measured using a post IR-IR SAR protocol, with a preheat temperature of 250 C for 1 minute, and stimulation with IR (870 nm) for 100 s while the aliquot was held at 50 C (IR 50 ), followed by a further 100 s with the sample held at 225 C (pIRIR 225 ) [21] ( [22]. Detection was through BG-39 and 7e59 filters. Signals used for dose estimation were based on the first 4 s of stimulation, less a background based on the signal between 95 and 100 s of stimulation. Multi-grain quartz and feldspar aliquots were employed because this study aims to identify well-bleached samples; the average dose is then the most appropriate dose estimate [23], and for a given number of measurements, this is most precisely measured using large aliquots. The samples were analysed for natural radionuclide concentrations in the laboratory, using highresolution gamma spectrometry [24,25]. These concentrations were converted into dose rates using conversion factors listed by Olley [26]; a cosmic ray contribution was calculated according to [27], assuming the modern burial depth has applied throughout the lifetime of the site. Both field and laboratory saturated water contents were measured. The resulting total dose rates to quartz are summarised in Table 7; the dose rates to feldspar can be derived by adding 0.81 Gy/ka to these values (based on an assumed concentration of 12 %K in feldspar extracts [28].  The quartz ages resulting from the measurements described above are summarised in Table 7, together with the ratios of the feldspar IR 50 and pIRIR 225 ages to quartz OSL ages (for the NLL-measured samples). The quartz ages are then characterised as 'probably well bleached', 'well bleached' or unknown based on these age ratios, following M€ oller and Murray [29].
Terrestrial Cosmogenic Nuclide (TCN) ( 36 Cl) exposure dating. e Erratic boulders on top of the major ice-marginal zone ridges were scouted by means of Mi8 helicopter transport, with flights over the ridges at 150 km/hr at 100 m height. We flew for a total of 2 days and covered~1500 km in total distance, but large boulders suitable for 36 Cl exposure dating proved difficult to find. Unfortunately, the Urdakh IMZ ('U' on Fig. 1) is covered with a sparse larch forest, and this prevented landing at potentially suitable boulders. Sampling was, however, possible at 11 sites along the Sampesa, the Syntabul e Severokokorsky and the Upper Taimyra e Baikuronyora ice marginal zones (Fig. 1), and with double sampling at a few sites, 16 boulders were sampled in total.
Samples were collected from the top surface of the largest available boulders in the vicinity, using an angle grinder and sawing the boulder in a cross-hatched pattern(see Fig 5D and E in [1]), enabling an exact estimate of the sample thickness. All sampled boulders were basalt and rested on flat surfaces on the crest of the IMZs. Sample coordinates and altitudes were obtained in the field using a handheld GPS. Topographic shielding was negligible for all sampled boulders. The dry bulk density was measured before crushing and sieving to the 250-125 mm fraction at Lund University, and averaged 3.0 g/cm 3 (Table 8). From each sample, c. 10 g was retained for whole rock elemental analyses at SGS Minerals Services, Canada, where major and trace elements were measured using X-ray fluorescence (XRF) and inductively coupled plasma e optical emission spectrometry (ICP-OES), respectively (Tables 9 and 10).   Fig. 17 in [1]). Four sediment units (AeD) were identified from shallow test pits in the~15 m high slope above the river. (B) Boulder and cobble armour of the river beach below the high-water mark at site LoR 5; the clasts result from erosion into the unit B diamict. (C) Close-up of the glacio-tectonically laminated diamict (unit B) at site LoR 6 ( Fig. 1; sediment log is Fig. 18 in Fig. 17 in [1]). Note lenticular sand intraclast (boudin) and the more angular, finely intra-laminated clay intraclasts (marked by small white arrows). (D) Sand intraclast (boudin) with internal primary lamination conforming to its outer shape; unit B diamict at site LoR 6. (E) Close-up of one of the clay intraclasts with preserved intra-lamination (2e5 mm) found in the unit B diamict at site LoR 6.    Table 1 Lithofacies codes (1st, 2nd and 3rd order code system) and their description as used in this work (basic system according to Eyles et al. [7] . Sensitivity analyses were conducted using the CRONUScalc calculator [33,34] to evaluate the potential impact of a rock surface erosion rate of 1 mm/kyr on the apparent exposure ages (Table 8). Table 2 Foraminiferal counts provided as raw count data in the actual sample. Only samples from the parts of the sections, where foraminifera are present, are included. Author names of taxa are also given. Of seven sections along the Bol'shaya Balaknya River, sampled for foraminiferal analyses (sections BBR 6,8,12,13,15,16,17), and the Novorybnoye 1 section (Fig. 1), all but two were found barren. Section LuR 6 along the Luktakh River (Fig. 1) was only analysed for foraminifera in it lowermost unit A, but not in marine sediments further up (unit C) in the sediment succession. Section logs are found in Figs. 7,8,9,12,13,14 Table 4 Plants and animals remains from fluvial sediments at site Logata River 3 (LoR 3b and 3d), sediment unit D. Section logs for sites LoR 3 are found in Fig. 19   Radiocarbon ages (n ¼ 69) from stratigraphic sections at sites along the Bol'shaya Balaknya River and the Luktakh e Upper Taimyra e Logata river system (Fig. 1). More exact site locations are seen on Fig. 6 and Fig. 15 in M€ oller et al. [1], and stratigraphic positions of samples are indicated in sediment logs in M€ oller et al. [1], Figs. 8, 10 11, 13, 14, 16, 18 and 19. Sites with sediment units marked with (*) are not described in [1], but will be used in a forthcoming paper. Finite radiocarbon ages on terrestrial material have been recalculated to calibrated 14 Table 6 Electron Spin Resonance (ESR) ages on molluscs from stratigraphic sections at sites along the Bol'shaya Balaknya River, the Luktakh e Upper Taimyra e Logata river system and the Novorybnoye site (Fig. 1). More exact site locations are seen in Fig. 6 and Fig. 15 in M€ oller et al. [1], and stratigraphic positions of samples are indicated in sediment logs in Figs Notes: U in is the uranium content in shells; U, Th, K are the uranium, thorium and potassium content in sediments; D S is the total dose rate; P s is the palaeodose. 1) Two shells of different species from the same sample were analyzed, and mean age taken.
2) The sample was dated by the ESR open system (ESR-OS) method (Molodkov, 1988).  Table 7 Optically Stimulated Luminescence (OSL) ages from stratigraphic sections at sites along the Bol'shaya Balaknya River, the Luktakh e Upper Taimyra e Logata river system and the Novorybnoye site (Fig. 1). More exact site locations are seen on Fig. 6 and Fig. 15 in [1], and stratigraphic positions of samples are indicated in sediment logs in [1] Table 8 Properties and analytical data for boulders on the Sampesa (SA), Syntabul eSeverokokorsky (NK) and Upper Taimyra e Baikuronyora (UT_B) Ice Marginal Zones (IMZ) analysed for cosmogenic 36 Cl (TCN exposure dating). Altitudes, latitudes, and longitudes were determined with GPS. For all samples, measured bulk rock density is 3.0 g/cm 3 , thickness is 5.0 cm, and topographic shielding is negligible. The rock dissolved indicates the amount processed for AgCl extraction chemistry. The Cl carrier is from PRIME Lab and has a 35 Cl/ 37 Cl ratio of 273. Uncertainties on 35 Cl/ 37 Cl and 36 Cl/Cl ratios and exposure ages represent propagated 1s analytical/internal uncertainties only. Sample 36 Cl concentrations are corrected for 36 Cl contributed by procedural blanks. Exposure age uncertainties in parentheses incorporate external uncertainties, including production rate uncertainties; comparisons of the 36 Cl ages with those derived from independent chronometers (e.g., radiocarbon, OSL) must account for these external uncertainties. Ages "w/erosion" are calculated with a prescribed rock surface erosion rate of 1 mm/ kyr. See Fig. 21