The long 14C chronologies currently used as calibration curves combine results from wood that grew in the western United States, the British Isles and Germany. Although these results show few significant differences in the 14C content of contemporaneous wood when averaged over the length of the chronology (i.e., the means of overlapping sections of chronology are the same), closer examination shows considerable variability. Separating the sections of chronology according to the provenance of the wood used for calibration reveals patterns that suggest small but finite differences in the 14C content of wood from different locations. We conclude that there is some evidence that German and American wood give dates older by between 20 and 40 yr from those of Irish oak for some periods. Additionally we suggest that the shift of the Belfast 1986 calibration data by ca. 18 yr toward older dates may not be valid and that the resultant offset between the Belfast 1986 and Seattle 1993 data shows a small but real difference in the 14C content of contemporaneous American, German and Irish wood. Intralaboratory measurements made in Belfast on contemporaneous German and Irish oak, and bristlecone pine and Irish oak, give offsets of 39 and 41 yr, respectively, with the Irish oak dating younger. Previous studies, in which sample pairs of American and English and French wood were processed in the same laboratory, also showed American wood to be slightly depleted in 14C. None of the findings of this study would significantly alter calibrated 14C dates.

We calculated a new ∆R for New Zealand waters using 14C dates of historic shell previously presented by McFadgen and Manning (1990), and fish otoliths dated by Kalish (1993). We obtained a new estimate of −25 ± 15 14C yr. To test the accuracy of this value for correcting conventional 14C marine ages, we dated shell of a variety of different species excavated at the prehistoric site of Shag Mouth, North Otago. We compared the results with a pooled mean date for terrestrial samples and calculated a local ∆R value that we found statistically indistinguishable from the new ∆R estimate for New Zealand.

We investigated relations among solar activity, climate and cosmogenic radiocarbon in a time series of various astrophysical, geophysical, archaeological and historical data. We studied records of tree-ring thickness, aurora borealis, the catalog of visible sunspots, sedimentary deposits from lakes and oceans, global glacial advance and retreat chronology, polar ice cores and human migrations. In these data, we searched for evidence of medium- and long-term solar cycles. Application of different spectral techniques to the atmospheric 14C concentration time series indicates the existence of spectral lines at a few dominant periodicities ranging from 11 yr to ca. 2 ka. Different laboratories have confirmed the presence of the ca. 210-and 2000-yr spectral features in long 14C series in tree rings. The ca. 210-yr 14C cycle is probably caused by heliomagnetic modulation of the cosmic-ray flux. The extrema of both the ca. 210-yr 14C period and solar activity correlate with the cold and warm epochs of global climate, at least for the past millennium, and this correlation has the correct sign. The periods of low solar activity are well correlated with the Little Ice Ages. The cause of the ca. 2 ka 14C period is, as yet, uncertain, but evidence from the analyses of various natural records shows that it could have a solar origin. In this study, we obtained powerful manifestations of solar activity and climate warming epochs at ca. 1500, 3800, 6100, 8200, 10,500 and 12,600 bp. A similar feature occurs in epochs of minimum amplitude in the 14C content in tree rings. Thus, solar activity may affect both the 14C content in the Earth's atmosphere and climate.

People usually study the chronologies of archaeological sites and geological sequences using many different kinds of evidence, taking into account calibrated radiocarbon dates, other dating methods and stratigraphic information. Many individual case studies demonstrate the value of using statistical methods to combine these different types of information. I have developed a computer program, OxCal, running under Windows 3.1 (for IBM PCs), that will perform both 14C calibration and calculate what extra information can be gained from stratigraphic evidence. The program can perform automatic wiggle matches and calculate probability distributions for samples in sequences and phases. The program is written in C++ and uses Bayesian statistics and Gibbs sampling for the calculations. The program is very easy to use, both for simple calibration and complex site analysis, and will produce graphical output from virtually any printer.

Peatlands occupy a significant portion of the land surface of the Earth and form a large carbon store. Most peat-forming systems have two layers. The upper layer, the acrotelm, fixes carbon by photosynthesis, loses some of it by decay and passes the remainder on to the lower layer, the catotelm. In the catotelm, decay continues at a slower rate. Mathematical models of the growth of the catotelm have been proposed which relate the cumulative mass of peat above a particular depth to calendar age of the peat at that depth. We demonstrate how 14C dating and the Bayesian approach to data analysis can be used to make inferences about the relation between calendar age and cumulative mass, and to estimate the accumulation and decay rates.

Prehistoric tell stratigraphies, like deep-sea sediments or peat deposits, store information about past atmospheric 14C variations. By matching the 14C ages on charcoal samples from settlement deposits with the tree-ring calibration curve, estimates for the time span covered by successive stratigraphic phases can be derived. This method is applied to 14C data from the tell mounds at Troy, Turkey and Ezero, Bulgaria. I compare the derived chronologies with the results of pottery shape seriation using correspondence analysis.

The British Nuclear Fuels plc reprocessing plant at Sellafield in Cumbria, Northwest England, discharges between 1 and 3 TBq of 14C per annum as low-level waste via a pipeline into the Eastern Irish Sea. Our results demonstrate 14C activities in excess of the current ambient level (i.e., ca. 260 Bq kg−1 carbon) in a range of both biotic and abiotic samples with evidence that a significant percentage of the discharges are in the form of dissolved inorganic carbon (DIC). There also appears to be a significant seasonal influence on the uptake of the DIC by the phytoplankton, most likely linked to their bloom periods. Results from different aspects of the research are integrated to provide an opportunity to consider the overall distribution and behavior of 14C in the Eastern Irish Sea. We also discuss the results in terms of existing sediment chronologies and sedimentation rates and consider the possible consequences of phytoplankton blooms on the cycling of the other radionuclides.

We compared 14C levels in annual growth rings of pine trees around the Tomsk nuclear fuel reprocessing plant (NFRP) and the Chernobyl nuclear power plant (NPP). At the Chernobyl site, samples were taken from the control area (within a 30-km radius zone around the site) to a distance of 80 km. In Tomsk, we collected samples between 8–10 km and 10–15 km from the site, taking into account prevailing wind directions. Background samples were collected 200 km from the plant. Samples were converted to benzene and counted in an ultra-low-level LSC Quantulus 1220™. Because of the Chernobyl accident, a signal can be detected in the background of routine plant operation. Comparison with the Tomsk data suggests that the routine discharges from Tomsk are more significant than the discharge from the Chernobyl accident. We estimated Tomsk NFRP annual discharge level at up to 30–45 TBq 14C from 1985 to 1988.

We measured the 14C content of atmospheric methane at a 200-m-high sampling station in The Netherlands. Combined with trajectories and a transport model, it is possible to estimate the 14CH4 emissions from nuclear power plants in northwestern Europe. We demonstrate here two different methods of analyzing the data: forward modeling and an inverse method. Our data suggest that the emissions from pressurized water reactors are 260 ± 50 GBq per GW installed power per year, ca. 1.6 ± 0.4 times higher than generally assumed. We also find that, in addition to the known nuclear sources of 14CH4 (pressurized and boiling water reactors), there are two very strong sources of 14CH4 (520 ± 200 and 1850 ± 450 GBq yr−1, respectively), probably two test reactors near the sampling station.

Canadian deuterium uranium (CANDU) pressurized heavy-water reactors produce 14C by neutron activation of trace quantities of nitrogen in annular gas and reactor components (14N(n,p)14C), and from 17O in the heavy water moderator by (17O(n,α)14C). The radiocarbon produced in the moderator is removed on ion exchange resins incorporated in the water purification systems; however, a much smaller gaseous portion is vented from reactor stacks at activity levels considerably below 1% of permissible derived emission limits. Early measurements of the carbon speciation indicated that >90% of the 14C emitted was in the form of CO2. We conducted surveys of the atmospheric dispersion of 14CO2 at the Chalk River Laboratories and at the Pickering Nuclear Generating Station. We analyzed air, vegetation, soils and tree rings to add to the historical record of 14C emissions at these sites, and to gain an understanding of the relative importance of the various carbon pools that act as sources/sinks within the total 14C budget. Better model parameters than those currently available for calculating the dose to the critical group can be obtained in this manner. Global dose estimates may require the development of techniques for estimating emissions occurring outside the growing season.

We measured airborne releases of 14C from the Paks Pressurized Water Reactor (PWR) Nuclear Power Plant (NPP). Two continuous stack samplers collect 14C in 14CO2 and 14CnHm chemical forms. 14C activities were measured using two techniques; environmental air samples of lower activities were analyzed by proportional counting, stack samples were measured by liquid scintillation counting. 14C concentration of air in the stack varies between 80 and 200 Bqm−3. The average normalized yearly discharge rates for 1988–1993 were 0.74 TBqGW−1ey−1 for hydrocarbons and 0.06 TBqGW−1ey−1 for CO2. The discharge rate from Paks Nuclear Power Plant is about four times higher than the mean discharge value of a typical Western European PWR NPP. The higher 14C production may be apportioned to the higher level of nitrogen impurities in the primary coolant. Monitoring the long-term average excess from the NPP gave D14C = 3.5‰ for CO2 and D14C = 20‰ for hydrocarbons. We determined 14C activity concentration in the primary coolant to be ca. 4 kBq liter−1. The 14C activity concentrations of spent mixed bed ion exchange resins vary between 1.2 and 5.3 MBqkg−1 dry weight.

As an industrialized city, Hong Kong annually consumes a large amount of fossil fuel. In addition, the Daya Bay Nuclear Power Plant in Shenzhen, Guangdong Province, has just begun operation 20 km from Hong Kong. These factors suggest that it may be appropriate and significant to examine the variation of atmospheric 14C levels in Hong Kong. We have collected and tested a variety of samples from different parts of Hong Kong: terrestrial annual grasses, marine plants and atmospheric CO2. We measured their 14C activity and compared it with that of cassia oil samples from Guangxi Province, China. The values obtained indicate that environmental 14C levels in the Hong Kong region agree with those found in Guangxi, both of which are significantly higher than the levels predicted by Povinec, Chudý and Šivo (1986).

We present here first results of 14CO2 monitoring at two sampling sites in the equatorial region of the South American continent (station Aychapicho, Ecuador and station Llano del Hato, Venezuela). We also include the data for two other stations representing undisturbed marine atmosphere at mid-latitudes of both hemispheres, far from large continental sources and sinks of CO2 (station Izaña, Tenerife, Spain and station Cape Grim, Tasmania). Between 1991 and 1993, 14CO2 levels in the tropical troposphere were generally higher by 2–5‰ when compared to mid-latitudes of both hemispheres. This apparent maximum of 14C in the tropics can be explained by two major factors: 1) emissions of 14C-free fossil fuel CO2, restricted mainly to mid-latitudes of the northern hemisphere; and 2) 14C depletion due to gas exchange with circumpolar Antarctic upwelling water, influencing mainly mid- and high southern latitudes. The δ14C record so far available for the Aychapicho station provides direct evidence for a regional reduction of atmospheric 14CO2 levels due to gas exchange with 14C-depleted equatorial surface ocean in the upwelling regions and dilution with the 14C-depleted CO2 released in these areas. Recurrent ENSO events, turning on and off the 14C-depleted CO2 source in the tropical Pacific, lead to relatively large temporal variations of the atmospheric 14C level in this region.

High-precision ∆14C and stable isotope (∆18O and ∆13C) records are reported for post-bomb corals from three sites off the eastern Australian coast. We observe that ∆14C values increased from ca. −50′ in the early 1950s to +130‰ by 1974, then decreased to 110‰ by 1991. There is general agreement between the coral results and ∆14C of dissolved inorganic carbon (DIC) in seawater measured previously for locations in the South Pacific. ∆14C values at our southern hemisphere sites increased at a slower rate than those observed previously in the northern hemisphere. Small variations in the ∆14C records among our three sites are likely due to differences in circulation between the shallow coastal waters and the open ocean influenced by seasonal upwelling. Low ∆14C is associated with most El Niño/Southern Oscillation (ENSO) events after 1970, indicating input of low 14C waters from the southern-shifted South Equatorial Current. The exception is the severe ENSO event of 1982–1983 when upwelling in the South Equatorial Current could have ceased, causing normal ∆14C values in the corals during this time.

Samples of living terrestrial plants comprising a moss (Calliergon sp.), Carex spp. and Alchemilla spp. were collected from the surface of the mire at Engimýri in Mýrdalur, southern Iceland, 10 km from the crater rim of the central complex of the Katla volcano. This area is 16 km from the fissures active in ad 1918 and was directly affected by the tephra fall. Although there is no hot-spring or fumerole activity in the area, sufficient volcanic activity during the weeks preceding sample collection produced a strong sulphurous odor in the streams. As part of a large-scale dating program, we analyzed the modern vegetation to determine whether anomalies caused by the uptake of “old” volcanic CO2 were apparent. The results showed 14C values for the Calliergon sp., Carex spp. and Alchemilla spp. of 113.2 ± 0.6 pMC, 113.03 ± 0.52 pMC and 113.10 ± 0.6 pMC, respectively. The δ13CPDB values were −28.7‰, −28.0‰ and −27.0‰, respectively. Similar vegetation, i.e., terrestrial plants from a marsh environment in southern Scotland, were also analyzed as a comparison and gave 14C values of 113.16 ± 0.55 and 112.98 ± 0.59 pMC. The implication is that Icelandic vegetation at Engimýri is not affected by “old” carbon from volcanic emissions and dates obtained for this Icelandic peat are acceptable and directly comparable with Scottish peat.

The Choushuichi alluvial-fan delta is the largest groundwater system in Taiwan. The coastal area of this fan delta has recently suffered severe land subsidence and sea-water intrusion due to overpumping of groundwater. To study the hydrogeological character of proper water-resource management, an intensive program of geological and water monitoring through well drilling and sampling has been undertaken. At present, radiocarbon dating and/or tritium analysis for samples from >70 boreholes has been completed. The clearly defined position of “bomb-tritium front” allows the groundwater flow and recharge rates of the confined aquifer to be calculated with confidence at <10 m yr−1 and 9.0 × 108 m3 yr−1, respectively. The flow velocities of the confined aquifer have also accelerated remarkably due to much pumping during the last decades.

Karst springs are abundant in Hungary, and many are thermal (temperatures >30°). As thermal springs are a significant part of Hungary's water resources, it is important to quantify their travel times in the karst systems. Thus, we chose to measure T and δ18O in the water and δ13C and 14C in dissolved inorganic carbon (DIC) in water from 50 thermal and nonthermal springs and wells in the Bükk Mountains, northeastern Hungary. Environmental isotope data confirm the karst water-flow pattern implied by earlier studies. We found the water in warm springs and boreholes to be mixtures of cold young and old thermal water. We also determined short mean-residence times for some large cold springs. The 14C activities measured in these springs indicate that the recharge area of the karst aquifer is open to the atmosphere, and atmospheric CO2 contributes to the 14C activity of these groundwaters. We observed good correlation between 14C and 3H activities and we determined negative correlations between 14C concentration and δ13C values and temperature. From the δ18O values of the oldest thermal waters, we attribute their origin to precipitation during colder temperatures than at present.

We report the first attempt to date Icelandic groundwater by the radiocarbon dating technique. We propose that boron concentrations of the water samples can be taken as a measure of the amount of dead CO2 in the groundwater leached from rock. By assuming that the ratio of boron to CO2 is the same in the rock and groundwater it is possible to correct the 14C concentration for the contribution of rock-derived carbon in the groundwater and thus calculate the initial (undiluted) 14C concentration. When this correction is applied, the values for the cold and warm waters lie in the range 60 to 110 pMC, which is in accordance with the estimated residence time for the groundwater. We have also calculated the fractions of atmospheric and organic carbon in the water, based on the δ13C measurements, and the results agree well with the general hydrology and vegetational cover of the areas. Results from geothermal systems indicate that it is possible to use 14C to estimate relative age differences of the water in a particular geothermal area.

We used moss increment counting to obtain well-defined samples of the topmost peat layers of two Sphagnum fuscum hummocks. The two ombrotrophic bogs, Lakkasuo in central Finland and Korvinsuo in eastern Finland, are of different ages, covering 3 and 9 ka, respectively. Using AMS dating, we traced bomb-produced 14C through the topmost parts of the two peat profiles. A well-defined 14C activity peak was found in both sequences dating the corresponding layer to ad 1965. A comparison between the maximum peat activities and the corresponding atmospheric values for the period of interest provides an opportunity to evaluate the amount of CO2 emanating from the decaying peat bog, and taken up by the living sphagnum plants.

Considerable variations in δ13C values were also observed. These variations indicate, at least partly, annual variations in the emission rate of CO2 from decomposition of older peat in the bog, and are connected with climatic factors such as temperature and precipitation.

We used direct radiocarbon dates of peat samples, pollen dates and land-uplift chronology from >1300 cores comprising all or most of the Holocene in the boreal region of Finland, Estonia and Maine (USA) to obtain long-term carbon accumulation rates for boreal peatlands. The “apparent” long-term rate of carbon accumulation (LORCA; g C m−2 a−1) ranged from 4.6 to 85.8 (mean 19.9 ± 10.7), depending on the geographical location, wetland type and the age of the mire. The “true” or “actual” rate of carbon accumulation (ARCA), as derived from models for peatbog growth, was usually ca. 70% of LORCA. We studied the raised bog Reksuo more intensely in terms of growth dynamics, and we report preliminary results of the “three-dimensional” or spatial carbon accumulation rates. These results strongly contradict the concept of constant input and constant decay throughout the millennia. The study emphasizes the importance of exploring the formation and dynamics of entire mire ecosystems, and the role of carbon in these systems, in addition to studying single cores.