Zeeberg.indd 385Zeeberg et al. 2003: Polar Research 22(2), 385–394 Climate variability in the Eurasian Arctic is sig- nifi cantly modulated by North Atlantic oceanic heat fl ux. Weather stations on Novaya Zemlya since 1961 document summer temperatures 0.3 - 0.5 °C and winter temperatures 2.3 - 2.8 °C lower than in the fi rst half of the 20th centu- ry (Zeeberg & Forman 2001). This temperature decrease is associated with a prolonged neg- ative phase of the North Atlantic Oscillation (NAO), decreased advection of North Atlantic Water, and below-average southern Barents Sea sea surface temperature (SST) during the 1960s, ‘70s and ‘80s (Loeng 1991; Zeeberg & Forman 2001). Likewise, post-”Little Ice Age” warming of the Barents Sea and glacier retreat on north- ern Novaya Zemlya is associated with a persist- ent positive phase of the NAO. Meteorological observations at the polar station Russkaya Gavan’, north-west Novaya Zemlya (76° 11' N, 59° E, Fig. 1), between 1932 and 1995 are concurrent with a mass balance time series on the adjacent Shokal’ski Glacier from 1933 to 1969. Stabilization and advance of several tide- water glaciers at Novaya Zemlya in the second half of the 20th century refl ects decreased summer temperatures and/or increased precipi- tation (Koryakin 1986; Zeeberg & Forman 2001). Elevated winter precipitation (up to 20 mm above the 27 mm average) associated with increased cyclonic activity, together with slowing calv- ing rates, arrested the negative mass balance of the Shokal’ski Glacier between 1959 and 1966. Observations during the 20th century indicate, however, that continued regional warming and summer temperature anomalies less than < 1 °C compensate for the added precipitation, result- ing in negative mass balances and glacier retreat (Chizov et al. 1968; Zeeberg & Forman 2001). Here we evaluate fjord sediment records as proxy for decade- to century-scale fl uctuations of a tidewater glacier on north-west Novaya Zem- lya. This analysis is based on the inference that Glacier extent in a Novaya Zemlya fjord during the “Little Ice Age” inferred from glaciomarine sediment records JaapJan Zeeberg, Steven L. Forman & Leonid Polyak Glacier activity at Russkaya Gavan’, north-west Novaya Zemlya (Arctic Russia), is reconstructed by particle size analysis of three fjord sediment cores in combination with 14C and 210Pb dating. Down-core logging of particle size variation reveals at least two intervals with sediment coarsen- ing during the past eight centuries. By comparing them with reconstruc- tions of summer temperature and atmospheric circulation, these intervals are interpreted to represent two cycles of glacier advance and retreat sometime during ca. AD 1400–1700 and AD 1700–present. Sediment accumulation thus appears to be sensitive to century-scale fl uctuations of the Barents Sea climate. The identifi cation of two glacier cycles in the glaciomarine record from Russkaya Gavan’ demonstrates that during the “Little Ice Age” major glacier fl uctuations on Novaya Zemlya occurred in broad synchrony with those in other areas around the Barents Sea. J. J. Zeeberg, Netherlands Institute for Fisheries Research, Haringkade 1, Box 68, 1970 AB IJmuiden, The Netherlands, jzeebe1@uicalumni.org; S. L. Forman, Dept. of Earth and Environmental Sciences, University of Illinois at Chicago, 845 W. Taylor Street, Chicago, IL 60607-7059, USA; L. Polyak, Byrd Polar Research Center, Ohio State University, 1090 Carmack Rd., 108 Scott Hall, Columbus, OH 43210-1002, USA. 386 Glacier extent in a Novaya Zemlya fjord during the “Little Ice Age” the glacial-marine record predominantly refl ects meteorological and glacier-specifi c mod ulations of sediment input (see Elverhøi et al. 1983; Smith & Schafer 1987; Syvitski et al. 1987; Cowan et al. 1988; Gilbert 2000). To assess changes in sedi- mentation, particle size variation was measured for three 1.2 m long gravity cores from Russkaya Gavan’, a > 100 m deep, 10 km long fjord, dom- inated by the Shokal’ski Glacier (Fig. 1). Cores were retrieved at ca. 1 km (IP98-22), ca. 3.3 km (IP98-23) and ca. 3.8 km (IP98-24) from the present glacier terminus (Figs. 1, 2). A second glacier (Laktyonov Glacier), which does not ter- minate in the fjord, may also contribute to fjord sedimentation through meltwater streams. Sed- iment accumulation in Russkaya Gavan’ may record glacier response to climate change during the “Little Ice Age” and other neoglacial events that have been widely recognized in the Barents Sea region and Scandinavia (e.g. Matthews 1991; Fig. 1. Russkaya Gavan’ with Shokal’ski and Laktyonov glaciers. The positions of cores IP98-22, 23 and 24 and the position of Core ASV-987 (indicated by a star) are shown. The long dashed line shows the inferred maximum glacier extent in the past 600 years. A shell obtained from the end moraine that indicates the “Little Ice Age” grounding line (short dashed line) of the Shokal’ski Glacier was dated to AD 1300–1400. Inset: pathways of North Atlantic Water along the Barents shelf and across the Barents Sea. The black block on Novaya Zemlya indicates the location of Russkaya Gavan’. 387Zeeberg et al. 2003: Polar Research 22(2), 385–394 Werner 1993; Lubinski et al. 1999). Methods Core collection at Russkaya Gavan’ Geological sampling and oceanographic meas- urements were performed in September 1998 from the RV Ivan Petrov. The glacial marine suc- cession sampled by the IP-98 gravity cores is indicated by an 8.8 kHz sonar profi le of the sea fl oor (Fig. 2). A thin (< 10 m) sedimentary cover (refl ector c) drapes the surface of another refl ec- tor (a, b), probably a glacigenic diamicton. Thick- er sedimentary sequences (ca. 30 m) can be seen in topographic depressions. At the location of one of the cores, IP98-23, an hermetically sealed box core was obtained to assure collection of the sedi- ment–water interface needed for 210Pb dating. A 6 m long gravity core (ASV-987) obtained and sam- pled in 1997 provides additional information on depositional variability within the fjord (Polyak et al. in press). Observations of water turbidity and conduc- tivity, temperature and depth (CTD) measure- ments in Russkaya Gavan’ in conjunction with the 1998 cruise indicate that meltwater is dis- charged from subglacial or englacial channels at the fjord head. The density difference between seawater and freshwater is 24 - 28 kg/m3, caus- ing rapid rise of the freshwater plume and settling of coarse-grained sediments in the resulting zone of deceleration, while fi ne fractions are dispersed throughout the fjord (Elverhøi et al. 1983; Syvit- ski et al. 1987; Gilbert 2000). A light transmis- sivity profi le and sampling of suspended matter demonstrates turbid layers along the sea surface and bottom of Russkaya Gavan’ (Fig. 2). Sus- pended sediment loads decrease ca. 2 km beyond cores 23, 24 and ASV-987. The spatial extent and coarseness of the plume principally refl ects gla- cier position, meltwater production, and fjord hydrodynamics (Syvitski et al. 1987: 111–174). Lithology Cores IP98-22, 23, and 24 consist of homoge- neous silty-mud with centimetre- to millime- tre-scale lamination and incidental clasts of ice- rafted debris (IRD; Fig. 3). Oxidation of organic carbon was indicated after core splitting by release of H2S and black colouration. Beds with diffuse lamination 10 - 20 cm thick alternate with beds that have fi ne lamination highlighted by reduced (black) monosulphides. Monosulphide beddings are most pronounced in IP98-23, sug- gesting periodic increase of clastic deposition alternating with settling of organic matter. The lower half of core 22 is more compact than the upper half and has two noticeable sandy layers. Down-core fl uctuations of grain size were established with a Malvern particle size ana- lyser (Mastersizer 2000, Hydro2000 MU), which measures grain diameters by laser dif- fraction while a subsample in liquid suspension is pumped through a recirculating cell. Cores 23 and 24 were sampled at 1 cm intervals, and core 22 at 2 cm intervals because of expect- ed increase of accumulation rates with proxim- Fig. 2. Sonar (8.8 kHz) profi le of Russkaya Gavan’, showing turbid layers measured in the fjord and the locations of the IP98 and ASV-987 gravity cores. The acoustic profi le demonstrates that a thin (< 10 m) sedimentary cover (refl ector c) is draping another refl ector (a, b) along the fjord, with accumulation up to 30 m thick in topographic depressions. Refl ector (a, b) is probably a glacigenic diamicton. 388 Glacier extent in a Novaya Zemlya fjord during the “Little Ice Age” (a) (b) 389Zeeberg et al. 2003: Polar Research 22(2), 385–394 ity to the glacier. Marine carbonates and organ- ics were removed during sample preparation by adding HCl (10 %) and H2O2 (30 %), respective- ly. When reactions ceased, the mixture was dilut- ed with distilled water before addition of a dis- persant (ca. 0.3 g Na4P2O7.10H2O). A subsample was then taken with a pipette while stirring and released in the sample tank of the Mastersizer. To separate remaining particle agglomerations, ultrasonics were applied for 10 s. Instrument set- tings were held constant with laser obscuration at approximately 15 % and a pump speed of 2500 rpm. Each sample yielded an apparent Gaussian distribution of particle sizes. Plotted in Fig. 3 are particle size ranges as volumetric percentages of the total sample and the particle size of the 90th percentile (d90). The d90 shows fl uctuation of the coarse tail of the distribution (Fig. 5), registering particle size fl uctuations more sensitively than, for example, the median (d50). Fine silt and clay (Wentworth scale: fractions < 16 µm) predominate in each core, comprising 80 - 90 % of core 24 (the most distal core), and 70 - 90 % in cores 22 and 23 (Fig. 3). Median par- ticle size is 7.6 µm for core 22, 7.1 µm for core 23 and 6.7 µm for core 24, refl ecting a decrease of coarse fractions away from the glacier. This trend is also observed in other studies of glaci- omarine sediments (e.g. Gilbert 2000; Desloges et al. 2002). The particle size analysis demon- strates that coarse silt and sand (fractions 63 - 125 µm and > 125 µm) comprise up to 12 % of core 22, compared to < 6 % in core 23 and < 2 % in core 24. In cores 23 and 24 fi nely laminated beds are somewhat (respectively 10 and 2 %) enriched in coarse sediments (medium to coarse silt and fi ne sand). IRD occurs throughout each core as inci- dental, < 10 mm long angular to subangular rock fragments. IRD is notably abundant in the lower half of core 24. The largest IRD fragments are 20 mm in core 23 (depth 12 cm) and core 24 (depth 61 cm). Fig. 3. Plots of grain size percentages with depth for cores (a, opposite page) IP98-22, (b, opposite page) 23, and (c) 24. From left to right are shown contents of sand, silt, and clay grouped in ranges (Wentworth scale): 1) coarser than fi ne sand, 2) coarse silt and fi ne sand, 3) medium to coarse silt, 4) fi ne silt and clay. Core photographs were taken immediately after core splitting to preserve black colouration of layering caused by reduced monosulphides. The core “shadows” in this fi gure (images to the right of the photographs) are composite X-radiographs used to identify molluscs and IRD. (c) 390 Glacier extent in a Novaya Zemlya fjord during the “Little Ice Age” Results Chronology: 14C and 210Pb dating Radiocarbon dating of in situ molluscs from the top and base of the cores provides broad chrono- logical control (Figs. 3, 4, Table 1), with calibrat- ed age ranges spanning 300 to 500 years (Stuiv- er et al. 1998). An additional challenge for 14C dating is the large and variable local marine reservoir correction. A study of bivalves from Novaya Zemlya fjords indicates a reservoir cor- rection of 775 ± 200 yr for Portlandia arctica, a detrital-feeding bivalve that can burrow up to 20 cm into sediments and feeds on old organ- ic matter within (Forman & Polyak 1997). This mollusc has a broad salinity tolerance and may live close to glaciers. Hence, the large reservoir correction indicated by this shell may result from assimilation of old carbon from glacier meltwater and/or detrital organics from sediments. Suspen- sion feeding bivalves (e.g. Astarte sp., Macoma sp.) are also common in glaciomarine sediments from Russkaya Gavan’, and previously yielded a reservoir age of ca. 400 ± 100 yr (Forman & Polyak 1997; see also Heier-Nielsen et al. 1995). Excess 210Pb activity measurements on uncon- solidated near-surface sediments provide a more precise age model for the past century than 14C (e.g. Smith & Schafer 1987; Hughen et al. 2000). Linear regression of 210Pb on depth in a 30 cm sealed core retrieved with IP98-23 gives a sed- imentation rate of 8 mm/yr, assuming a 226Ra background of 1.5 dpm/g (Fig. 4). 210Pb dating is often calibrated by location of the 137Cs “bomb” spike. The 137Cs fallout peak was ca. 1964, but there is often a delay of 5 to 10 years before cesium settles from the water column with sedi- ments. Analysis of the reference core did not yield the cesium spike, which may be due to sediment mixing or low isotope concentration, because deposition from the atmosphere decreases above ca. 60° N (Hughen et al. 2000; J. Smith, pers. comm. 2001). The sedimentation rate derived by 210Pb analysis demonstrates that the reference core covers the 1970s. The 137Cs isotope slight- ly increases downcore, consistent with elevated ambient 137Cs during this period (Fig. 4). In con- trast to the 210Pb-derived decadal rate, 14C dating of consolidated sediments provides a maximum limiting chronology because of the effect of time averaging, potentially yielding sedimentation rates that are too low. Basal ages of at least 600 years were derived by 14C dating of Macoma sp. bivalves for cores IP98- 23 (1.12 m) and IP98-24 (1.24 m). The resulting linear sedimentation rate of ca. 2 mm/yr is com- parable to rates calculated for subpolar fjords of north Spitsbergen (Elverhøi et al. 1983) and west Greenland (Desloges et al. 2002). Comparison with core ASV-987, which has a mean accumu- lation rate of ca. 6 mm/yr (Polyak et al. in press) indicates substantial variability of sediment dis- tribution within the fjord. Higher accumulation Fig. 4. (a) Age model for the IP98 and ASV-987 cores based on calibrated 14C ages obtained from molluscs. (b) Modern sedi- mentation rates in Russkaya Gavan’ are derived from analysis of the 210Pb decay series in a 0.3 m core that collected the sedi- ment–water interface (including modern 210Pb) at the location of IP98-23. The 137Cs isotope increases down-core, consistent with elevated 137Cs levels during the 1960s and 1970s. (a) (b) 391Zeeberg et al. 2003: Polar Research 22(2), 385–394 rates in ASV-987 shows that thicker sediment was sampled here, possibly refl ecting additional sedimentation from the Laktyonov Glacier’s fl u- vial delta (Fig. 1). A mollusc at 56 cm in core IP98-22 yielded a submodern age, indicating a higher accumulation rate (ca. 5 - 10 mm/yr) for core 22 than for cores 23 - 24, consistent with its glacier proximal posi- tion. With accumulation rates 5 (14C) to 8 (210Pb) mm/yr, core 22 is probably continuous through the 19th century. There is no 14C age control for the past three centuries for IP98-23 and 24. Cross- correlation between cores based on shell ages and peaks in the coarse fraction suggests that of IP98- 23 and 24 the core tops, comprising the past ca. 50 - 100 years, were lost during coring (Fig. 5). Glacier proximity in the sedimentary record The sediment record in Russkaya Gavan’ is dom- inated during the past ca. 800 years by fi ne- grained glacifl uvial deposition with some input from IRD and sediment gravity fl ows. The exclu- sively clay to medium silt-sized particle ranges measured in cores 22 - 24 suggest that these cores were retrieved from glacier-distal environments, dominated by settling of silt and agglomerated clay particles from suspension. Coarse silts and sands (63 - 125 and > 125 µm, Fig. 3) constitute < 2 % in core 24 (3.8 km from the glacier), < 6 % in core 23 (3.3 km from the glacier), and < 12 % in core 22 (1 km away from the glacier). Medium-grained sand is deposited within ca. 200 m from ice fronts in fjords of north-west Spitsbergen (Elverhøi et al. 1993). Turbulent, high energy meltwater discharge by temperate, high precipitation glaciers of south-east Alaska deposits sand within 1 km of the glacier termi- nus (Cowan et al. 1988). Hence, for the subpo- lar Shokal’ski Glacier we infer that the absence of medium sands (250 µm) in IP98-22, obtained ca. 1 km from the present glacier terminus, sug- gests that this core was > 0.2 km from the ground- ing line at all times (Fig. 1). The increase in frac- tions > 125 µm between 40 and 80 cm indicates increased meltwater discharge and/or increased glacier proximity in the mid-19th century (Figs. 3, 5). Limited glacier advance (< 1 km) is con- sistent with an apparent absence of “Little Ice Fig. 5. Correlation of IP cores based on calibrated 14C age ranges. Shown here is the fl uc- tuation with depth of the grain diameter of the 90th percentile (d90, see text for explanation). 392 Glacier extent in a Novaya Zemlya fjord during the “Little Ice Age” Age” moraines on the acoustic profi le. Possi- ble moraine ridges (see Elverhøi et al. 1983) ca. 3 km down the fjord were probably produced by an earlier grounding line (Fig. 2). Historical observations (Jermolaev 1934 cited in Chizov et al. 1968; also Zeeberg & Forman 2001) indicate that during the 19th century, the fl oating (calving) glacier margin extended further than the ground- ing line and over the location of IP98-22. The gla- cier’s limited response to climate fl uctuations in the past ca. 800 years may refl ect its steep area- elevation gradient (Fig. 1), making it less suscep- tible to changes in equilibrium line altitude (Zee- berg & Forman 2001). Discussion Interpretation of the grain size signal Suspended sediment loads in subpolar, glacier- dominated fjords commonly range between 50 and 200 mg/L during summer, but in winter these loads are usually < 2 mg/L (Syvitski et al. 1987: 123, 134). During winter, glacier melt and output of sediment with meltwater is minimal. Climate change primarily affects the length of the summer (melt) season, which for Russkaya Gavan’, at sea level, has been less than two months (July and August) during the exceptionally warm 20th cen- tury (> 1 °C anomaly; Briffa et al. 1995). The net annual mass balance may be determined during these summer months by a few melt days, and thus by cloudiness and sea ice, which limits heat transport to the region. Cold periods can be characterized by a low number of “melt years”, reduced meltwater delivery, and low net sedi- ment accumulation. Increased melt would result in increased sediment output and an increase of coarse fractions. The 14C-constrained glaciomarine record from Russkaya Gavan’ reveals two intervals with sig- nifi cant sediment coarsening and a concomitant drop of fi ne fractions (Fig. 3). Correlation of the cores based on 14C ages reveals that these coars- ening episodes were probably produced by the same cycles of glacier advance–stabilization– retreat (Fig. 5). Because the core sites character- ize different sedimentation zones in the fjord, the expression of coarsening varies between cores. Thus, a narrow peak in core 23 (40 cm, coarse fraction 6 %) appears to correlate with a broader peak in core 24 (35 cm, coarse fraction 2 %). The age of the lower coarsening event can be broadly estimated as between the 14th and 17th century, whereas the upper coarse interval may be about two centuries younger based on sedimentation rates of 2+ mm/yr. The younger coarsening event may be correlative to the coarse part of core 22, although age control of this core is insuffi cient for a defi nitive correlation. Table 1. Pelecypod species and ages for the IP98 cores. Depth (cm) Species andres. correction 14C age (yr BP) (lab. number) Res. corr. 14C age Cal age (yr AD ) 1σ (2σ) and probability Midpoint plotted (1σ) Core IP98-22 56 Portlandia biv 290 ± 40 modern modern 1950 Core IP98-23 55 Portlandia biv. 400 ± 100 1025 ± 45 (AA-37278) 625 ± 145 1270–1434 (1066–1625) 1243–1442 (0.997) 775 ± 200 250 ± 245 1428–mod (1280–mod) 1478–1698 (0.573) 1588 106 Macoma biv. 1225 ± 55 (AA-37279) 825 ± 155 1024–1297 (895–1419) 1147–1291 (0.610) 1160 Core IP98-24 102 Portlandia? fragments 775 ± 200 1205 ± 45 (AA-37281) 805 ± 145 1037–1376 (977–1419) 1150–1298 (0.663) 430 ± 245 1297–1945 (1060–mod) 1293–1675 (0.946) 1484 110 Macoma biv. 400 ± 100 925 ± 55 (AA-37280) 525 ± 155 1296–1486 (1216–1786) 1288–1517 (0.937) 1402 393Zeeberg et al. 2003: Polar Research 22(2), 385–394 Inferred glacial history Glaciers on Novaya Zemlya during the 20th cen- tury demonstrate variable response to regional warming, because increased Barents Sea SST enhance both summer ice melt and winter precip- itation (Zeeberg & Forman 2001). However, pro- longed warming and summer temperature anom- alies > 0.5 °C result in a declining Shokal’ski Glacier mass balance (Chizov et al. 1968; Mikha- lov & Chizov 1970). Summer temperature anom- alies are well documented by tree-ring derived temperature time series for northern Russia (Briffa et al. 1995) and temperature composites (including ice cores) for the Northern Hemisphere (Mann et al. 1999). Strongly negative summer temperature anomalies may herald the advance of glaciers on Novaya Zemlya in the early 14th cen- tury (and also in the 19th century). This is con- sistent with the dating of an Astarte bivalve from a subfossil moraine ridge at a distance of ca. 500 m from Shokal’ski Glacier’s present terminus to AD 1300–1400 (Zeeberg 2001). The abundance of IRD in the bottom half of core 24 may indicate stabilization of the advanced glacier with iceberg calving at some time between ca. AD 1300 and 1700. Glacier melt that would have ended this cycle was probably caused by increased advection of North Atlan- tic Water into the Barents Sea during a persistent positive phase of the NAO. Between AD 1450 and 1650 there were at least four episodes with per- sistent positive phase NAOs (Cook et al. 2002). A drop of North Atlantic sea level pressure is indi- cated by ion contents in the GISP-2 ice core after AD 1400 (Meeker & Mayewski 2002). After ca. AD 1650 the NAO remains comparatively weak until the 20th century. Strong negative summer temperature anomalies (Briffa et al. 1995; Mann et al. 1999) in combination with increased precip- itation, as suggested by strengthening of the Ice- landic Low (Meeker & Mayewski 2002), proba- bly caused signifi cant glacier advance on Novaya Zemlya in the 19th century. We suggest that the second glacier cycle inferred from the glaciomar- ine sequence from Russkaya Gavan’ corresponds to the 19th century glacier maximum and subse- quent decay. The coarse layers in the bottom half of core 22 may indicate turbidity currents trig- gered by the advancing glacier (Figs. 3, 5). The sediment cores from Russkaya Gavan suggest two “Little Ice Age” glacier advanc- es sometime during ca. AD 1300–1700 and AD 1700–present (Fig. 5). These events appear to be generally consistent with glacial geologic obser- vations from other areas around the Barents Sea. Glacier advances in Franz Josef Land are con- strained by 14C dating of in situ mosses from gla- cier margins to ca. AD 1400–1600 and post-1650 (Lubinski et al. 1999). Two stages of moraine sta- bilization, in the 14th century and post-1700, have been described for Svalbard on the basis of liche- nometry (Werner 1993). In southern Scandinavia, “Little Ice Age” maxima were reached between AD 1400 and AD 1600 (Matthews 1991). Gla- ciers on Svalbard, Franz Josef Land, and Novaya Zemlya attained their greatest “Little Ice Age” extent in the second half of the 19th century (Werner 1993; Lubinski et al. 1999; Zeeberg & Forman 2001). The identifi cation of two glacier cycles in the glaciomarine record demonstrates that sediment accumulation in Russkaya Gavan’ is sensitive to century scale fl uctuations of the Barents Sea climate. During the “Little Ice Age”, major glacier fl uctuations on Novaya Zemlya appear to have occurred in broad synchrony with those in other areas around the Barents Sea. Acknowledgements.—210Pb and 137Cs analyses were performed by John N. Smith, Bedford Institute of Oceanography (Dart- mouth, Canada). Particle size analyses were done in coopera- tion with Torbjörn Törnqvist (University of Illinois at Chica- go). Sonar profi le is reproduced courtesy of P. I. Krinitski, V. A. Gladysh, and Y. P. Goremykin (Okeangeologia, St. Peters- burg, Russia). Lawrence Febo produced the X-ray images and Jeanne Jaros (both of the Byrd Polar Research Center) the core photographs. This paper benefi ted from reviews by J. A. Mat- thews and an anonymous referee. National Science Founda- tion award no. OPP-9796024 supported the research. References Briffa, K. R., Jones, P. D., Schweingruber, F. H., Shiyat- ov, S. G., & Cook, E. R. 1995: Unusual twentieth-century summer warmth in a 1,000-year temperature record from Siberia. Nature 376, 156–159. Chizov, O. P., Koryakin, V. S., Davidovich, N. V., Kanevsky, Z. M., Singer, E. M., Bazheva, V. Y., Bazhev, A. B. & Khmelevskoy, I. F. 1968: Results of researches on the pro- gram of the International Geophysical Year. Glaciation of the Novaya Zemlya. Glaciology IX section of IGY program. No. 18. (In Russian, with English summary.) Moscow: Nauka Publishing House. Cook, E. R., d’Arrigo, R. D. & Mann, M. E. 2002. A well veri- fi ed, multiproxy reconstruction of the winter North Atlantic Oscillation Index since AD 1400. J. Clim. 15, 1754–1764. Cowan, E. A., Powell, R. D. & Smith, N. D. 1988: Rainstorm- 394 Glacier extent in a Novaya Zemlya fjord during the “Little Ice Age” induced event sedimentation at the tidewater front of a tem- perate glacier. Geology 16, 409–412. Desloges, J. R., Gilbert, R., Nielsen, N., Christiansen, C., Rasch, M. & Øhlenschläger, R. 2002: Holocene glacima- rine sedimentary environments in fi ords of Disko Bugt, west Greenland. Quat. Sci. Rev. 21, 947–963. Elverhøi, A., Lønne, O. & Seland, R. 1983: Glaciomarine sed- imentation in a modern fjord environment, Spitsbergen. Polar Res. 1, 127–149. Forman, S. L. & Polyak, L. 1997: Radiocarbon content of pre- bomb marine mollusks and variations in the 14C reservoir age or coastal areas of the Barents and Kara seas, Russia. Geophys. Res. Lett. 24, 885–888. Gilbert, R. 2000: Environmental assessment from the sed- imentary record of high-latitude fi ords. Geomorphology 32, 295–314. Heier-Nielsen, S., Heinemeier, J., Nielsen, H. L. & Rud, N. 1995: Recent reservoir ages for Danish fjords and mar ine waters. Radiocarbon 37, 875–882. Hughen, K. A., Overpeck, J. & Anderson, R. F. 2000: Recent warming in a 500-year palaeotemperature record from varved sediments, Baffi n Island, Canada. The Holocene 10, 9–19. Koryakin, V. S. 1986: Sokrashcheniye oledeneniya na ostrov- akh Yevraziyskoy Arktiki v XX veke. (Decrease in glacier cover on the islands of the Eurasian Arctic during the 20th century.) Pol. Geogr. Geol. 10, 157–165. Loeng, H. 1991: Features of the physical oceanographic con- ditions of the Barents Sea. Polar Res. 10, 5–18. Lubinski, D. J., Forman, S. L. & Miller, G. H. 1999: Holocene glacier and climate fl uctuations on Franz Josef Land, Arctic Russia, 80° N. Quat. Sci. Rev. 18, 85–108. Mann, M. E., Bradley, R. S. & Hughes, M. K. 1999. North- ern Hemisphere temperatures during the past millennium: inferences, uncertainties, and limitations. Geophys. Res. Lett. 26, 759–762. Matthews. J. A. 1991: The late Neoglacial (‘Little Ice Age’) glacier maximum in southern Norway: new 14C-dating evi- dence and climatic implications. The Holocene 1, 219–233. Meeker, L. D. & Mayewski, P. A. 2002: A 1400-year high- resolution record of atmospheric circulation over the North Atlantic and Asia. The Holocene 12, 257–266. Mikhalov, V. I. & Chizov, O. P. 1970: Rezul’taty gliatsiolog- icheskikh issledovanii na Novoi Zemle v 1969 g. (Scien- tifi c results of the glaciological investigations on Novaya Zemlya in 1969.) Data Glaciolog. Stud. 17, 186–199. Polyak, L., Murdmaa, I. & Ivanova, E. in press: A high- resolution, 800-year glaciomarine record from Russkaya Gavan’, a Novaya Zemlya fjord, eastern Barents Sea. The Holocene. Smith, J. N. & Schafer, C. T. 1987: A twentieth century record of climatologically-modulated sediment accumulation rates in a Canadian fjord. Quat. Res. 27, 232–247. Stuiver, M., Reimer, P. J., Bard, E., Beck, J. W., Burr, G. S., Hughen, K. A., Kromer, B., McCormac, F. G., Van der Plicht, J. & Spurk, M. 1998: Intcal98 radiocarbon age cali- bration, 24,000-0 cal BP. Radiocarbon 40, 1041–1083. Syvitski, J. P. M., Burrell, D. C. & Skei, J. M. 1987: Fjords. Processes and products. Berlin: Springer Verlag. Werner, A. 1993. Holocene moraine chronology, Spitsbergen, Svalbard: lichonometric evidence for multiple Neoglacial advances in the Arctic. The Holocene 3, 128–137. Zeeberg, J. J. 2001: Climate and glacial history of the Novaya Zemlya Archipelago, Russian Arctic. Amsterdam: Rozen- berg Publishers. Zeeberg, J. J. & Forman, S. L. 2001: Changes in glacier extent on north Novaya Zemlya in the twentieth century. The Holocene 11, 161–175.