Larsen.indd 283Larsen et al. 2002: Polar Research 21(2), 283�290 Correlation of late Holocene terrestrial and marine tephra mark ers, north Iceland: implications for reservoir age changes Gudrún Larsen, Jón Eiríksson, Karen Luise Knudsen & Jan Heinemeier The tephrochronology of the last 3000 years has been investigated in soil sections in north Iceland and in a marine sediment core from the north Ice landic shelf, 50 km offshore. Tephra markers, identiÞ ed with major ele ment geochemical analysis of volcanic glass shards, serve to correlate the mar ine and terrestrial records. Hekla 3, the largest Holocene tephra marker from the volcano Hekla, in south Iceland, dated to 2980 years BP, is used as the basal unit in the tephra stratigraphy. AMS 14C dating of molluscs in the sediment core shows variable deviation from the tephro- chronological age model, indicating that the reservoir age of the sea water mass at the coring site has varied with time. A standard mar ine reservoir cor rection of 400 14C years appears to be reasonable at the present day in the coastal and shelf waters around Iceland, which are dom inated by the Irminger Current. However, values over 500 years are observed during the last 3000 years. We suggest that the intervals with increased and vari- able marine reservoir correction reß ect incursions of Arctic water masses derived from the East Greenland Current to the area north of Iceland. G. Larsen & J. Eiríksson, Science Institute, University of Iceland, IS-101 Reykjavík, Iceland; K. L. Knudsen, Dept. of Earth Sciences, University of Aarhus, DK-8000 Århus C, Denmark; J. Heinemeier, AMS 14C Dating Laboratory, Institute of Physics and Astronomy, University of Aarhus, DK-8000 Århus C, Denmark. Palaeoclimatic studies that extend beyond histor- ical and instrumental records must rely on dating tech niques which enable reconstruction of time series for climate variability related prox ies. The purpose of the present study is to obtain inde pen- dent control on 14C dating of a high res olution sediment core obtained from an oceano graphic boundary region between Atlantic and Arctic water masses. The study is a part of the Eur opean Union HOLSMEER project, which focuses on climatic variability in shal low marine and coastal settings during the last 2000 years. The study area on the north Icelandic shelf has the advantage of being close to the source vol- can oes of Holocene tephras, and it is located in an ocean og raphically sensitive boundary region between the relatively warm, saline Irminger Cur- rent and the East Icelandic Current, which forms a cold, low salinity tongue of surface water (Fig. 1). The shallow north Iceland shelf seabed is strong- ly affected by surface circulation in the area (Eir- íks son et al. 2000a). Tephra markers that can be traced from volcanic source regions into the mar- ine depositional environment provide inde pen- dent control on radiocarbon dates from that envir- on ment. The age of the tephra markers are based on historical records from Iceland for the last 900 years, correlation to the Greenland ice core chron ology, and radiocarbon dates of ter restrial material (see e.g. Thorarinsson 1958, 1967, 1974; Larsen & Thorarinsson 1977; Larsen 1982, 1984, 2000; Sæmundsson 1991; Grönvold et al. 1995; Zielinski et al. 1997; Haß idason et al. 2000). In this paper we demonstrate that a high res olution tephra stratigraphy can be used to link chronologies in the terrestrial and marine 284 Correlation of late Holocene terrestrial and marine tephra markers, north Iceland environ ments. The investigation represents the Þ rst detailed high resolution land�sea correlation of the regional terrestrial tephrochronology with the mar ine record, although several previous studies have demonstrated the importance of single tephra markers (e.g. Lacasse et al. 1995, 1996; Eiríks son et al. 2000a, 2000b; Jennings et al. 2000). This allows us to compare two age models for a 3000 year sediment record from a mar ine shelf core north of Iceland. One model is based on tephrochronological data, the other on accelerator mass spectrometry (AMS) radiocarbon dating of the marine record. It is sug gested that discrepancies between the two age models are related to palaeoceanographic changes in the region and resulting changes in the reserv oir age of the water masses on the north Icelandic shelf. Materials and methods The CALYPSO piston core MD992275 (66° 33.06� N, 17° 41.59� W; 440 m water depth) was collected from the north Icelandic shelf during the 1999 IMAGES cruise (Fig. 1). The research methods used for the core material presented here have been described by Eiríksson et al. (2000a) and by Knudsen & Eiríksson (2002). Tephra horizons, observed during visual and X- ray inspection of the cores, were subsampled and sieved on a 63 µm sieve for the preparation of polished thin sections. Major element geo chem- ical analysis of volcanic glass shards (Table 1) were carried out by standard wavelength dispersal technique on an ARL-SEMQ microprobe at the Geo logical Institute, University of Bergen, Nor- way, with an accelerating voltage of 15 kV, a beam current of 10 nA, and a defocused beam diameter of 6 - 12 µm. Natural and synthetic minerals and glasses were used as standards. Soil sections with tephra layers at six localities in north Iceland (Figs. 2, 3) provide detailed land- based tephrochronology for comparison with the mar ine records. Subsamples of the terrestrial teph ras were prepared in the same way as the mar- ine samples. Samples from three sections have been analysed for major elements on the ARL- SEMQ microprobe at the University of Bergen. The 14C datings of marine samples (Table 2) were carried out at the AMS 14C Dating Lab or- atory at the University of Aarhus, Denmark. The dates have been corrected for natural isotopic frac tion ation by normalization to δ13C = -25 � VPDB, and calibrated with CALIB4 (Stuiver et al. 1998a), using the marine model calibration curve (Stuiver et al. 1998b). A standard reservoir cor rection of about 400 14C years (∆R = 0) is built into this model (see also Andersen et al. 1989). In this paper we simply refer to calibrated years BP as years. Tephra markers in north Iceland and the north Icelandic shelf The number of late Holocene tephra layers from the last 3000 years in our soil sections in north Iceland ranges from about 50 in the eastern part to about 25 in the western part. Many of these tephra layers are probably too thin to form off- shore deposits, while others are regional markers with high potential as offshore markers. Tephra sections on land that are relevant for the present study are brieß y described below, and only the tephra layers detected so far in core MD992275 will be treated. The Svartárvatn lake lies about 90 km inland (1x on Fig. 2i), north of the volcanic areas where Fig. 1. Location map showing the study area and the present-day surface ocean circulation in the northern North Atlantic (based on Hurdle 1986). Depth contour interval 1000 m. Also shown are the central volcanoes of the most relevant volcanic sys tems. Loc ation of core MD992275 is 66° 33.06� N, 17° 41.59� W. 285Larsen et al. 2002: Polar Research 21(2), 283�290 Holo cene eruption frequency has been highest. Several soil sections were studied in the vicinity of the lake, and were merged to form a composite soil section showing the regional tephra stratig- raphy, in which nearly 50 tephra layers younger than Hekla 3 have been identiÞ ed (Fig. 3). Forty of these tephras have been analysed by electron micro probe. The tephra layers originate from at least six volcanic systems, but the great majority has the chemical characteristics of the Veidivötn� Bárd arb unga system. The other volcanic systems include the Gríms vötn, Hekla, Katla and Snæfells- jökull systems (Fig. 1). IdentiÞ cation of tephra lay ers from the last 11 centuries was partly based on previous work (Thorarinsson 1958, 1967, 1974; Larsen 1982, 1984; Haß idason et al. 2000), but work on the older tephra layers is still in progress. Analyses of nine tephra layers tephra from the terrestrial sections in north Iceland (Svartár vatn, Figs. 2, 3; and Eyjafjördur, Fig. 2) are presented in Table 1a. The V-1477, V-1410, Hekla 1300, Hekla 1104, Settlement (V-871/877) and Katla 9th century tephras have all been described earlier (Thor arinsson 1967; Larsen 1982, 1984; Ólafsson 1985). The Settlement tephra was dated to 871 ± 2 and 877 ± 4 AD in the GRIP and GISP 2 ice cores, respectively (Grönvold et al. 1995; Zielin ski et al. 1997). We report for the Þ rst time the presence in north Iceland of the white acid tephra from an eruption in Snæ fellsjökull vol- cano radiocarbon dated to 1750 ± 150 BP (Stein- thorsson 1967), previously reported only in west and north-west Iceland (Jóhannesson et al. 1981; Roth 2000). Peaty soil collected 0.6 - 2 cm below this tephra at Svartárvatn yielded a date of 1855 ± 25 BP. Analyses of the Snæfellsjökull 1 tephra in the Svartárvatn section are presented in Table 1a as S-Svart and those of a sample from a section 10 km west�north-west of the volcano Fig. 2. (a�h) Dispersal of eight of the tephra layers relevant for this study. The lines show the outermost isopach of each layer as known at present. This isopach is 0.5 cm for the V-1477 and V-871/877 tephra layers, 0.2 cm for Hekla 1104 and 0.1 cm for the others. The central volcanoes of the relevant volcanic systems are indicated by capitals letters in 2i (full names on Fig. 1). Note that V-1477 and V-871/877 were erupted on long volcanic Þ ssures belonging to the Veidivötn�Bárdarbunga system. Key sections in north Iceland are indicated by an X and a number. Relevant for this study are 1: Svartárvatn section; 2: Breidavík section; 3: Eyjafjördur section. 286 Correlation of late Holocene terrestrial and marine tephra markers, north Iceland are included for comparison as Sn-1. Analyses of a basaltic tephra a few centimetres above the Snæfellsjökull 1 tephra are presented in Table 1a as G-Svart. This tephra is characterized by TiO2 in the range 2.1 - 2.4 weight % and is the older of two tephra layers in the key sections having these TiO2 values (the other is 12th century AD). Finally we present analyses of Hekla 3 from the Svartárvatn section. Eight tephra horizons of homogeneous glass comp osition have been found in the uppermost 7 m of core MD992275 (Fig. 3, Table 1b). Several hori zons with abundant glass grains have also been ana lysed, but as the glass composition was hetero geneous the analyses are not included here except for a few grains with Snæ fells jökull 1 composition from two such horizons (Table 1c). The four youngest horizons in core MD992275, at depths of 179, 209, 239 and 271 cm (Table 1b), are correlated to the V-1477, V-1410, Hekla 1300 and Hekla 1104 AD tephra layers in the key sec- tions (Fig. 3). The 1477, 1410 and 1300 AD tephras are reported in an offshore deposit for the Þ rst time, but Hekla 1104 has also been found in other cores in the area (Eiríksson et al. 2000a, 2000b). The two tephra horizons at 322 and 324 cm form a pair where the upper tephra has the chemical Table 1 (opposite page). (a) Representative microprobe glass analyses of tephra lay ers relevant to this study from key sec- tions in north Ice land. Eight of the tephra samples are from the Svart árvatn key section; the exceptions are the Katla 9th cen tury tephra (sampled in the Eyjafjördur key section) and the Sn-1 sample (from a section on the Snæfellsnes peninsula). The analyses do not show the complete composition range of each tephra layer. (b) Representative microprobe glass an alyses of tephra horizons in the core MD992275. Five hori- zons with basaltic glass grains of homogeneous com position occur at depths of 179, 209, 322, 324 and 438 cm. Three hori- zons with silicic glass grains of limited com positional range occur at depths of 239, 271 and 687 cm. (c) Snæ fells jö kull 1 composition shards in MD992275: selected analyses from two horizons (depths 431 cm and 455 cm) of glass grains with a variety of compositions. Low totals are due to the small shard size of the tephra and the few analyses avail able. They are nevertheless strong indications of the grains� origin. Fig. 3. Tephra stratigraphy of core MD992275 and the key sections Breidavík and Svartárvatn. All the tephra layers are shown as black lines without speciÞ c thickness. Asterisks indic ate that glass from the tephra layer or horizon has been ana lysed for major element composition. Grey line shows infer- red position of Snæfellsjökull 1 in core MD992275. Stip pled line indicates tephra enriched horizons of mixed ori gin but containing Snæ fellsjökull grains. Tephra layers rele vant to this study are labelled with source (volcanic syst em) and eruption year. The volcanic systems are: Veid ivötn�Bárdarbunga (V), Grímsvötn (G), Hekla, Katla and Snæfellsjökull. 287Larsen et al. 2002: Polar Research 21(2), 283�290 SiO2 TiO2 Al2O3 FeO MgO CaO Na2O K2O P2O5 V-1477 50.84 1.84 13.23 12.94 6.80 11.46 2.60 0.28 0.07 50.66 1.84 13.32 12.72 6.75 11.60 2.32 0.22 0.25 50.06 1.90 13.26 12.78 7.07 11.78 2.31 0.29 0.22 51.06 1.96 13.05 13.38 6.55 11.42 2.01 0.21 0.17 49.70 2.02 13.88 12.85 6.87 12.12 2.31 0.18 0.22 50.17 2.06 13.19 13.15 6.62 11.51 2.53 0.23 0.24 V-1410 49.54 1.91 13.86 12.31 7.07 11.82 2.53 0.28 0.06 49.81 1.96 13.99 12.35 7.21 12.04 2.35 0.25 0.02 48.65 1.85 14.00 12.48 6.81 11.78 2.33 0.27 0.20 49.47 1.92 13.72 12.40 7.04 11.85 2.35 0.25 0.11 49.09 1.84 14.13 12.26 6.92 11.64 2.33 0.18 0.28 49.33 1.83 14.19 12.44 6.95 11.76 2.29 0.25 0.25 Hekla 1300 62.43 0.84 15.50 7.69 0.79 4.28 4.59 2.00 0.36 62.23 0.92 14.85 8.15 0.92 4.13 4.51 2.04 0.48 61.61 1.10 15.22 8.74 1.19 4.71 4.11 1.77 0.37 61.30 1.02 14.99 9.59 1.51 4.68 4.28 2.03 0.26 60.29 1.34 14.66 9.97 1.68 4.94 4.04 1.81 0.57 58.47 1.26 15.90 10.10 2.08 5.57 4.23 1.56 0.48 Hekla 1104 73.76 0.26 13.67 2.62 0.03 1.75 4.41 2.89 0.06 72.82 0.21 14.21 3.00 0.05 1.88 4.53 2.70 0.00 72.69 0.25 14.08 3.26 0.11 1.92 4.23 2.58 0.00 72.41 0.27 14.20 3.31 0.12 2.01 4.33 2.66 0.00 71.69 0.29 14.47 3.24 0.01 1.90 4.52 2.84 0.00 70.68 0.32 14.40 3.34 0.09 1.91 4.36 2.89 0.13 V-871 / 877 48.47 1.94 13.58 12.65 6.74 11.45 2.55 0.22 0.22 49.38 1.92 13.68 13.05 6.36 11.03 2.47 0.29 0.09 49.33 1.86 13.71 12.70 6.30 11.35 2.43 0.21 0.27 48.85 1.86 13.70 12.85 6.79 11.39 2.34 0.17 0.20 49.93 1.82 13.56 12.81 6.51 11.19 2.20 0.26 0.22 49.59 1.77 13.58 12.94 6.39 11.22 2.36 0.30 0.19 K 9th century 49.53 4.61 13.17 15.13 4.45 9.37 2.42 0.82 0.55 48.83 4.43 13.46 15.75 4.44 9.40 2.96 0.77 0.46 48.55 4.83 13.28 14.88 5.27 10.42 2.59 0.86 0.47 48.04 4.74 13.12 15.55 4.87 9.80 3.00 0.86 0.64 47.75 4.56 12.94 15.55 4.90 9.46 2.59 0.90 0.60 47.52 4.35 13.09 15.02 4.91 9.74 2.64 0.81 0.49 G-Svart 49.68 2.42 13.92 12.31 6.32 11.02 2.39 0.43 0.20 49.36 2.34 14.60 12.12 6.70 11.35 2.64 0.38 0.31 49.43 2.29 14.12 12.04 6.63 11.45 2.38 0.44 0.31 48.95 2.27 14.24 12.26 6.40 11.20 2.51 0.42 0.16 49.57 2.26 14.38 12.12 7.03 11.52 2.58 0.35 0.23 50.12 2.17 14.16 11.94 6.76 11.37 2.53 0.40 0.15 S-Svart 67.48 0.38 15.63 4.11 0.21 1.81 4.77 4.19 0.05 67.42 0.48 15.80 4.25 0.35 1.72 5.62 4.49 0.09 66.08 0.41 15.98 4.00 0.28 2.00 5.61 4.01 0.00 65.46 0.51 15.99 4.62 0.33 2.08 5.39 4.04 0.00 65.42 0.45 15.96 4.39 0.34 2.03 5.64 4.02 0.02 64.96 0.47 15.74 4.40 0.30 1.94 5.21 4.05 0.01 Sn-1 69.32 0.33 15.54 3.37 0.16 1.30 5.12 4.59 0.02 67.63 0.41 15.61 3.83 0.30 1.61 5.14 4.51 0.08 67.32 0.42 15.70 4.29 0.49 1.94 5.06 4.38 0.04 67.16 0.36 15.22 4.36 0.24 1.57 4.70 4.41 0.00 66.36 0.40 15.71 4.40 0.32 1.61 5.25 4.34 0.25 65.81 0.36 16.34 4.20 0.44 2.31 5.17 3.93 0.02 Hekla 3 71.76 0.22 14.48 3.04 0.12 1.98 4.43 2.66 0.01 71.45 0.25 14.20 3.13 0.14 2.07 4.59 2.52 0.06 71.12 0.28 14.26 3.04 0.10 1.96 4.69 2.66 0.11 71.11 0.26 14.07 3.05 0.16 1.97 4.63 2.47 0.00 70.43 0.19 14.31 3.16 0.14 2.04 4.48 2.67 0.02 69.30 0.14 14.31 3.12 0.03 2.05 4.62 2.54 0.03 SiO2 TiO2 Al2O3 FeO MgO CaO Na2O K2O P2O5 MD992275 / 179 49.66 1.90 13.94 13.36 6.65 11.49 2.44 0.30 0.12 49.47 1.94 13.32 13.46 6.30 11.50 2.51 0.23 0.24 48.86 1.94 14.21 12.88 6.49 11.65 2.23 0.23 0.31 48.15 2.01 13.41 12.66 7.56 12.48 2.14 0.25 0.11 49.25 2.02 13.18 13.67 6.26 11.56 2.40 0.31 0.16 50.65 2.09 12.96 13.71 6.60 11.43 2.21 0.23 0.15 MD992275 / 209 51.53 1.83 13.25 13.10 6.10 11.81 2.21 0.25 0.12 51.05 1.93 13.26 12.68 6.57 12.04 2.15 0.23 0.15 50.94 1.96 14.61 12.53 7.32 11.05 2.39 0.19 0.24 50.62 1.76 14.13 12.85 6.87 11.49 2.26 0.21 0.19 50.31 1.91 14.38 13.11 6.55 10.98 2.38 0.24 0.21 49.36 1.83 14.19 12.68 7.01 11.59 2.30 0.23 0.17 MD992275 / 239 61.73 1.22 15.64 9.13 1.63 4.95 4.29 1.76 0.41 61.20 1.02 15.21 8.95 1.55 4.95 4.08 1.80 0.33 61.11 1.15 15.73 9.34 1.49 5.00 4.33 1.76 0.58 60.05 1.32 15.31 8.76 1.58 4.49 4.32 1.67 0.39 58.74 1.12 15.53 9.30 1.38 5.06 4.08 1.77 0.52 57.59 1.15 15.72 9.44 1.62 5.34 4.38 1.67 0.23 MD992275 / 271 74.19 0.20 14.39 3.13 0.08 2.13 4.11 2.81 0.01 72.69 0.25 13.38 3.24 0.11 2.05 4.50 2.65 0.03 72.65 0.22 14.46 3.17 0.00 2.08 4.67 2.84 0.00 72.04 0.30 14.24 3.24 0.14 2.06 4.61 2.77 0.11 70.74 0.31 13.68 3.29 0.08 1.99 4.30 2.62 0.07 69.36 0.24 13.94 4.73 0.06 4.40 5.16 1.62 0.04 MD992275 / 322 48.61 1.75 14.26 12.51 6.76 11.27 2.39 0.26 0.15 50.54 1.88 13.89 12.85 6.88 11.59 2.11 0.23 0.31 48.82 1.97 13.61 13.22 6.37 10.83 1.77 0.20 0.11 49.89 2.03 13.67 12.92 6.69 11.27 2.46 0.18 0.07 49.49 2.06 14.04 11.85 7.37 12.32 2.42 0.22 0.18 49.49 2.09 13.67 12.53 6.91 11.26 2.46 0.23 0.07 MD992275 / 324 48.12 4.48 13.59 15.32 4.89 10.02 2.91 0.94 0.55 47.66 4.40 12.89 15.04 4.83 10.02 2.99 0.72 0.62 47.59 4.96 13.15 15.17 4.59 9.92 3.06 0.88 0.43 47.45 4.70 13.38 14.71 4.78 9.86 3.14 0.92 0.53 47.40 4.72 12.74 14.71 4.55 9.94 2.98 0.86 0.67 46.32 4.50 13.45 14.81 4.77 9.99 3.12 0.76 0.43 MD992275 / 438 49.68 2.35 14.18 12.34 6.32 11.31 2.66 0.29 0.34 49.68 2.30 14.38 12.41 7.17 11.34 2.46 0.32 0.14 49.58 2.18 14.67 12.24 6.74 11.04 2.35 0.41 0.28 49.16 2.20 14.52 12.16 6.77 11.25 2.49 0.43 0.20 49.13 2.26 14.35 12.26 6.88 11.18 2.40 0.34 0.23 49.12 2.33 14.47 12.15 6.54 10.95 2.56 0.30 0.26 MD992275 / 687 71.45 0.18 14.37 3.19 0.00 2.00 5.02 2.46 0.00 71.16 0.26 14.29 3.18 0.04 1.98 4.92 2.38 0.00 70.21 0.03 14.21 3.20 0.09 1.97 4.39 2.36 0.00 69.66 0.19 14.05 3.18 0.22 2.15 4.48 2.39 0.11 67.28 0.33 15.06 5.34 0.30 3.08 4.59 2.17 0.07 65.42 0.45 15.18 6.12 0.38 3.51 4.64 2.04 0.12 SiO2 TiO2 Al2O3 FeO MgO CaO Na2O K2O P2O5 431 65.44 0.49 15.60 3.91 0.29 1.75 5.16 4.46 0.10 63.70 0.38 15.19 3.95 0.32 1.66 5.20 4.00 0.00 455 64.46 0.47 15.93 4.86 0.52 2.31 5.05 3.95 0.03 60.82 0.59 15.81 4.74 0.44 2.08 4.49 3.80 0.12 (a) (b) (c) 288 Correlation of late Holocene terrestrial and marine tephra markers, north Iceland signature of the Veidivötn�Bárdarbunga system and the lower one has that of the Katla system. Three such pairs have been identiÞ ed in the key sections Svartárvatn and Eyjafjördur. The youngest pair is the Settlement tephra and the 9th century Katla tephra, the second youngest pair is from the early 9th / late 8th century AD and the third is much older. The chemical signature Þ ts best with the late 9th century tephras. The Settlement tephra has not yet been traced to the north coast of Iceland (Fig. 2e shows the 0.5 cm isopach) despite its volume of 5 km3 as freshly fallen tephra (recalculated from Larsen 1984), but offshore deposition is likely considering the magnitude of the eruption. Below 3.3 m the core has only partly been searched for tephra horizons. One distinct tephra hor izon has been detected at 438 cm in the core (Fig. 3). It consists of homogeneous basaltic glass and has TiO2 values in the 2.1 - 2.4 weight % range that allow it to be correlated to the G-Svart teph ra in the Svartárvatn section. Chemically identi cal grains to those of Snæfellsjökull 1 have been found in two of the glass-rich horizons at depths of 431 and 455 cm in core MD992275 (shown as stip pled lines in Fig. 3), together with a var i ety of acid, intermediate and basaltic glass grains. Examples of glass grains with Snæ fells jö- kull 1 compo sition are presented in Table 1c. A tephra hori zon consisting entirely of acid glass with the chem ical characteristics of the Hekla system occurs at 687 cm. The composition and range is con sis tent with the Hekla 3 tephra in the Svart ár vatn section as well as the Hekla 3 tephra iden ti Þ ed in core HM107-03 by Eiríksson et al. (2000a). Radiocarbon dates of molluscs from core MD 992275 are listed in Table 2 as well as the age of dated tephra markers. The results are plotted against core depth in Fig. 4, which shows an age model based on tephrochronology as well as the cal ibrated radiocarbon dates. Above the Hekla 3 teph ra marker, over 100 years need to be sub- tracted from the radiocarbon dates to bring them to the tephra age model level. Discussion The tephrochronological age model for core MD992275 (Fig. 4) is at present based on Þ ve well-constrained dates of tephra layers younger than 1130 years (the time span of Iceland�s record- ed history) and on terrestrial radiocarbon dates on two older tephra layers: Snæfellsjökull 1 and Hek la 3. A straight-line Þ t age model between the tephra layers is presented, and slight adjust- ments can be expected as more terrestrial dates become available. The date of the Hekla 3 marker tephra is well-constrained (Dugmore et al. 1995). The two dates on the Snæfellsjökull 1 tephra are in reason able agreement, considering that the 1855 ± 25 date is on a soil slice separated from the tephra by 0.6 cm and the 1750 ± 150 date is on soil im mediately below. The presence of the Snæfells- jökull 1 tephra in the core is manifested by the occur rence of glass grains in the sediment below and above the G-Svart tephra. The lowest level of iden tiÞ ed glass grains with Snæfellsjökull 1 com- position is at 455 cm in core MD992275 and such grains are also found above the G-Svart tephra. These grains are considered redeposited, and the level of the primary Snæfellsjökull 1 tephra is tentatively estimated at 460 cm depth (Fig. 3), slightly below the level of their Þ rst appear ance. Previous age model investigations on the north Icelandic shelf have shown discrepancies between 400 year reservoir age corrected radio- Fig. 4. Age�depth diagram for core MD992275. Two possible age models are shown. The solid line indicates an age model based only on tephra markers (tephrochronological age model). The AMS 14C dates are shown with ± one standard deviation (see also Table 2). 289Larsen et al. 2002: Polar Research 21(2), 283�290 carb on age models and tephrochronological age models, both for the upper Holocene and the Late glac ial (Eiríksson et al. 2000a, 2000b; Knud sen & Eir íks son 2002). They concluded that a south- ward shift of the Polar Front ca. 3000 years ago, indi cated by benthic and planktonic foramin- ifera, coincided with an increase of about 130 years in the reservoir age correction needed to Þ t their radio carbon and tephrochronological age models. This value corresponds approximately to the reserv oir age of ca. 550 14C years reported by Taub er & Funder (1975). Before 3000 years BP, a 400 year correction was sufÞ cient back to ca. 4500 years BP. The results presented here show that core site MD992275 has been extensively affected by Arctic water for the past 3000 years. At present, the site is closer to the Polar Front than site HM107-03 studied by Eiríksson et al (2000a) and Knud sen & Eiríksson (2002). The time lag between the tephra age model and the calibrated radiocarbon dates in core MD992275 increases sharply to at least 400 years after the deposition of Hekla 3, decreasing to nearly zero at ca. 1900 years BP. Above that there is a lag of 100 - 200 years with broad maxima at around 1400, 600 and 300 years BP. More radio- carbon dates and a tighter control on the tephro- chronology above Hekla 3 is necessary for the quanti Þ cation of these age deviations and their frequency, as well as high resolution studies of palaeoceanographic prox ies throughout the core. This work is cur rently in progress within the HOLSMEER project. Conclusions 1) Nearly 50 tephra layers from the last 3000 years have been documented in key sections in north Ice land, providing a basis for high res- olution terrestrial tephra stratigraphy to which marine records can be correlated. 2) Reliable correlations between terrestrial and marine tephra layers can be based on geochem- ical signatures and mapping of the distribution of these layers on land and on the sea ß oor. 3) Correlation of eight tephra horizons in mar- ine core MD992275 to dated tephra layers in key sections on land allows the construction of a teph- ro chronological age model for core MD992275 Table 2. Radiocarbon dates and ages of dated tephra markers in core MD992275. All radiocarbon samples were calibrated with CALIB4 (Stuiver et al. 1998a, 1998b). The age of historical tephra layers is reported as calendar years BP (before 1950), rounded off to the nearest decade. Radio carbon dates of the pre-settlement tephra layers are on ter restrial material below the tephras; for details on Hekla 3 see Dugmore et al. (1995). Tephra marker depths correspond to the base level of each unit. A standard reservoir correction of about 400 years for marine samples is built into the radiocarbon model. ° = assumed standard δ13C value. Depth cm. Lab. no. Material; 14C age (BP) Cal. age(s) BP Cal. ± 1 σ δ13C MD992275 dated tephra layer ± 1 σ R = 400 (BP) 63 - 64 AAR-7116 Thyasira equalis 695 ± 45 310 410 - 290 -7° 100 - 101 AAR-7117 Thyasira cf. equalis 785 ± 40 440 470 - 410 -6.85 122 - 123 AAR-6089 Siphonodentalium lobatum 895 ± 45 510 530 - 480 +0.70 133 - 135 AAR-7118 Thyasira equalis 815 ± 45 460 490 - 430 -8.98 162 - 163 AAR-7119 Nuculana sp. 945 ± 35 530 550 - 510 +0.51 179 V-1477 470 209 V-1410 540 220 - 224 AAR-7120 Thyasira equalis, Thyasira sp. 1265 ± 45 790 880 - 740 -8.69 239 Hekla 1300 650 259 - 260 AAR-7121 Thyasira equalis 1420 ± 50 950 1000 - 920 -8.00 272 Hekla 1104 850 287 - 291 AAR-6931 Thyasira equalis 1555 ± 35 1110 1160 - 1060 -6.84 321 Settlement layer 1080 332 - 333 AAR-7122 Siphonodentalium lobatum 1710 ± 45 1270 1290 - 1230 +0.44 373 - 374 AAR-6932 Siphonodentalium lobatum 1905 ± 40 1450 1510 - 1400 +0.69 381 - 382 AAR-6933 Bathyarca glacialis 2020 ± 40 1570 1620 - 1530 +1.65 ca. 460 LL-1169A Snæfellsjökull 1 (peat) 1750 ± 150 1690 - 1630 1860 - 1520 ca. 460 KIA17232 Snæfellsjökull 1 (peaty soil) 1855 ± 25 1820 1820 - 1730 -24.6 472 - 473 AAR-6934 Cf. Dentalium entalis 2245 ± 40 1850 1890 - 1810 +0.72 512 - 513 AAR-6935 Cf. Siphonodentalium lobatum 2530 ± 45 2180 2290 - 2120 +0.91 537 - 538 AAR-7123 Thyasira equalis 2680 ± 65 2340 2440 - 2310 -6.80 582 - 584 AAR-6936 Thyasira sp. 3110 ± 70 2860 2960 - 2780 1° 632 - 633 AAR-7124 Thyasira cf. equalis 3345 ± 45 3200 3260 - 3150 -8.05 687 Hekla 3 (peat) 2879 ± 34 2990 - 2970 3080 - 2950 696 - 697 AAR-6937 Siphonodentalium lobatum 3265 ± 50 3080 3160 - 3010 +0.96 796 - 797 AAR-6938 Bathyarca glacialis 3795 ± 50 3710 3810 - 3660 +2.11 290 Correlation of late Holocene terrestrial and marine tephra markers, north Iceland and comparison between marine and ter restrial dates. 4) Discrepancies between reservoir corrected radiocarbon dates of molluscs and the tephro- chronological age model vary from tens to hun- dreds of years. 5) The discrepancies between the radiocarbon and tephrochronological age models may be related to incursions of Arctic water masses into the northern North Atlantic to the north and east of Iceland. Acknowledgements.�This paper is a contribution to the Eur- opean Union 5th Framework HOLSMEER project (contract no. EVK2-CT-2000-00060), which supported Þ eld work, dat- ing and geochemical analysis. Core material was obtained from the MD9922 IMAGES Cruise in 1999. We are grate ful to the Institut Français pour la Recherche et la Tech nologie Pol- aires for the IMAGES coring operations on board the Mari on Dufresne. Participation in the cruise and the sci entiÞ c work was sup ported by grants from the Icelandic Re search Council and the Danish Natural Science Research Council. We thank Prof. L. A. Simonarson, University of Ice land, for identifying mol luscs for the radiocarbon dating. We thank O. Tumyr and H. Haß idason at the University of Berg en for assistance and discussions at various stages of this project. References Andersen, G. J., Heinemeier, J., Nielsen, H. L., Rud, N., John- sen, S., Sveinbjörnsdóttir, Á. E. and Hjartarson, Á. 1989: AMS 14C dating on the Fossvogur sediments, Iceland. Radio carbon 31, 592�600. Dugmore, A. J., Cook, G. T., Shore, J. S., Newton, A. J., Edwards, K. J. & Larsen, G. 1995: Radiocarbon dating teph- ra layers in Britain and Iceland. Radiocarbon 37, 379�388. Eiríksson, J., Knudsen, K. L., Haß idason, H. & Heinemeier, J. 2000a: Chronology of late Holocene climatic events in the north ern North Atlantic based on AMS 14C dates and tephra markers from the volcano Hekla, Iceland. J. Quat. Sci. 15, 573�580. Eiríksson, J., Knudsen, K. L., Haß idason, H. & Henriksen, P. 2000b: Late-glacial and Holocene palaeoceanography of the north Icelandic shelf. J. Quat. Sci. 15, 23�42. Grönvold, K., Óskarsson, N., Johnsen, S. J., Clausen, H. B., Ham mer, C. U., Bond, G. & Bard, E. 1995: Ash layers from Ice land in the Greenland GRIP ice core correlated with oce an ic and land sediments. Earth Planet. Sci. Lett. 135, 149�155. Haß idason, H., Eiriksson, J. & Kreveld, S. V. 2000: The tephro chronology of Iceland and the North Atlantic region during the Middle and Late Quaternary: a review. J. Quat. Sci. 15, 3�22. Hurdle, B. G. 1986: The Nordic seas. New York: Springer. Jennings, A., Syvitski, J., Gerson, L., Grönvold, K., Geirs- dót tir, Á., Hardardóttir, J., Andrews, J. & Hagen, S. 2000: Chronology and paleoenvironments during the late Weich- selian deglaciation of the southwest Iceland shelf. Boreas 29, 167�183. Jóhannesson, H., Flores, R. M. & Jónsson, J. 1981: A short account of the Holocene tephrochronology of the Snæfells- jökull central volcano, western Iceland. Jökull 31, 23�30. Knudsen, K. L. & Eiríksson, J. 2002: Application of tephro- chronology to the timing and correlation of palae ocean- ographic events recorded in Holocene and Lateglacial shelf sediments off north Iceland. Mar. Geol. 191, 165�188. Lacasse, C., Sigurdsson, H., Carey, S., Paterne, M. & Guich- ard, F. 1996: North Atlantic deep-sea sedimentation of Late Quat ernary tephra from the Iceland hotspot. Mar. Geol. 129, 207�235. Lacasse, C., Sigurdsson, H., Jóhannesson, H., Paterne, M. & Carey, S. 1995: Source of Ash Zone 1 in the North Atlantic. Bull. Volcanol. 57, 18�32. Larsen, G. 1982: Gjóskutímatal Jökuldals og nágrennis. (Teph ro chronology of the Jökuldalur area.) In H. Thórarins- dót tir et al. (eds.): Eldur er í Nordri. (Fire in the north.) Pp. 51�65. Reykjavík: Sögufélag. Larsen G. 1984: Recent volcanic history of the Veidivötn Þ s- sure swarm, southern Iceland�an approach to volcanic risk assessment. J. Volcanol. Geotherm. Res. 22, 33�58. Larsen, G. 2000: Holocene eruptions within the Katla vol- canic system, south Iceland: characteristics and environ- mental impact. Jökull 49, 1�28. Larsen, G. & Thorarinsson, S. 1977: H4 and other acid Hekla tephra layers. Jökull 27, 28�46. Ólafsson, G. 1985: Gjóskulög í Austurdal og Vesturdal, Skaga Þ rdi. (Tephra layers in Austurdalur and Vesturdalur, Skagafjördur.) BSc thesis, University of Iceland. Roth, V. 2000: Um nákudungslagakambinn í landi Bæjar í HrútaÞ rdi. (The Zirphaea crispata deposit at Bær in Hrúta- fjördur.) BSc thesis, University of Iceland. Sæmundsson, K. 1991. Jardfrædi Kröß ukerÞ sins. (Geology of the Kraß a volcanic system.) In A. Gardarson & Á. Einars son (eds.): Náttúra Mývatns. (The natural history of lake Mývatn.) Pp. 25�95. Reykjavík: Hid Íslenska Nát- túrufrædifélag. Steinthorsson, S. 1967: Tvær nýjar C14 aldurákvardanir á ös ku- lögum úr Snæfellsjökli. (Two new C14 dates on tephra lay ers from Snæfellsjökull.) Náttúrufrædingurinn 37, 236�238. Stuiver, M., Reimer, P. J., Bard, E., Beck, J. W., Burr, G. S., Hughen, K. A., Kromer, B., McCormack, G., van der Plicht, J. & Spurk, M. 1998a: INTCAL98 radiocarbon age calibration, 24,000�0 cal BP. Radiocarbon 40, 1041�1083. Stuiver, M., Reimer, P. J. & Braziunas, T. F. 1998b: High- precision radiocarbon age calibration for terrestrial and marine samples. Radiocarbon 40, 1127�1151. Tauber, H. & Funder, S. 1975: 14C content of recent molluscs from Scoresby Sund, central East Greenland. Grønl. Geol. Unders. Rapp. 75, 95�99. Thorarinsson, S. 1958: The Öræfajökull eruption of 1362. Acta Nat. Isl. II 2. Thorarinsson, S. 1967: The eruptions of Hekla in historical times. In T. Einarsson (ed.): The eruption of Hekla 1947�1948 I. Pp. 1�170. Reykjavík: Societas Scientiarum Islandica. Thorarinsson, S. 1974: Vötnin stríd. Saga Grímsvatnagosa og Skeidarárhlaupa. (The swift-ß owing rivers. Grímsvötn eruptions and glacial ß oods in Skeidará.) Reykjavík: Bók a- út gáfa Menningarsjóds. Zielinski, G. A., Mayewski, P. A., Meeker, L. D., Grönvold, K., Germani, M. S., Whitlow, S., Twickler, M. S. & Taylor, K. 1997: Vol canic aerosol records and tephrochronology of the Sum mit, Greenland, ice cores. J. Geophys. Res. 102, 26 625�26 640.