The purpose of this paper is to analyze the numerous holocene subfossil trees (Pinus silvestris) buried in alluvial deposits in the Southern French Alps. These trees lived between the Allerod and Subboreal periods, according to 14C dates. Our dendochronological studies explain the trees' sudden death as due to morphological crisis brought on by climatic oscillations. Tree-ring series could be used to identify the variability of early Holocene atmospheric 14C levels.

During the last decade, several radiocarbon-dated varve chronologies have been produced. The main goal at first was the extension of the 14C calibration curve beyond 10,000 BP. This paper aims to discuss varve chronologies of Soppensee and Holzmaar Lakes. Although both chronologies encountered problems, high-resolution 14C dating and relative varve time have been obtained for events during the Late Glacial.

The radiocarbon timescale has been calibrated by dendrochronology back to 11.8 ka cal BP, and extended to 14.8 ka cal BP using laminated marine sediments from the Cariaco Basin. Extension to nearly 23.5 ka cal BP is based on comparison between 14C and U-Th ages of corals. Recently, attempts to further extend the calibration curve to >40 kyr are based on laminated sediments from Lake Suigetsu, Japan, foraminifera in North Atlantic sediments, South African cave deposits, tufa from Spain, and stalagmites from the Bahamas. Here we compare these records with a new comparison curve obtained by 234U(super 230) Th ages of aragonite deposited at Lake Lisan (the last Glacial Dead Sea). This comparison reveals broad agreement for the time interval of 20-32 ka cal BP, but the data diverge over other intervals. All records agree that Delta-14C values range between approximately 250-450 per mil at 20-32 ka cal BP. For ages >32 ka cal BP, the Lake Suigetsu data indicate low Delta-14C values of less than 200 per mil and small shifts. The other records broadly agree that Delta-14C values range between approximately 250 and 600 per mil at 32-39 ka cal BP. At approximately 42 ka cal BP, the North Atlantic calibration shows low Delta-14C values, while the corals, Lisan aragonites, and the Spanish tufa indicate a large deviations of 700-900 per mil. This age is slightly younger than recent estimates of the timing of the Laschamp Geomagnetic Event, and are consistent with increased 14C production during this event.

This paper presents an updated atmospheric radiocarbon calibration from annually laminated (varved) sediments from Lake Suigetsu (LS), central Japan. As presented earlier, the LS varved sediments can be used to extend the radio-carbon time scale beyond the tree ring calibration range that reaches 11,900 cal BP. We have increased the density of 14C measurements for terrestrial macrofossils from the same core analyzed previously. The combined data set now consists of 333 measurements, and is compared with other calibration data.

The Holocene varve chronology of annually laminated sediment sequences from Lake Meerfelder Maar agree for most of the record with dendro-calibrated accelerator mass spectronomy radiocarbon dates from the same site. Only between 9710 and 9950 cal BP does an offset of 240 yr appear between both data sets. At this position, a micro-disturbance in the varve succession has been detected by thin section analyses and was quantified in terms of missing varves. A comparison with the nearby record from Lake Holzmaar, as well providing high resolution AMS 14C and varve chronologies, revealed that such gaps (ca. 2% in time for the entire Holocene) are exceptional for these long-varved maar lake records. Moreover, since sections of missing years appear for both profiles at different stratigraphic positions, a combination of both the Meerfelder Maar and Holzmaar records enables us to bridge erroneous zones in varve chronologies. This confirms the high potential of two long-varved records in close vicinity to each other for the elimination of dating errors and for increasing chronological precision at a time resolution that is normally regarded as within the counting errors. Late Glacial varve and 14C data beyond the dendro-calibration from Meerfelder Maar and their tentative tele-connections to other high resolution data sets reveal unexplained age discrepancies in the calendar year time scale of about 200 years.